Transmissible spongiform encephalopathies (TSE) are a group of fatal neurodegenerative diseases affecting a wide range of hosts including scrapie in sheep and goats, chronic wasting disease in cervidae as well as Creutzfeldt-Jakob disease in humans. The hallmark of these diseases is the accumulation of a disease-associated partially Protease-resistant isoform (PrP^Sc) resulting from the conversion of the host-encoded membrane-bound glycoprotein, cellular prion protein (PrP^c).

Bovine specified risk materials (SRM) are tissues which are considered to possibly contain bovine spongiform encephalopathy (BSE) infectivity in incubating animals. These tissues are banned for the use in human food and health products because of the potential risks of transmission of BSE to consumers and the attendant/further development of variant Creutzfeldt-Jakob disease. Currently the list of bovine SRM varies between countries, including those of the European Union and North America. No international SRM regulations exist. For example, the small intestine from cattle of all ages is banned for use in the EU whereas in North America only the distal ileum is banned. As a consequence products such as beef casings, which are produced from the jejunum, are banned from human consumption in the EU but not in North America.

The scientific basis for the current SRM regulations concerning the gut is based on the wide distribution of infectivity in the intestine of TSE infected sheep and on BSE studies that have had a limited scope. In these studies BSE infectivity and the detection of PrP^Sc was limited to the distal ileum of the small intestine. In the distal ileum of experimentally infected cattle both PrP^Sc, beginning at 10 months post infection (mpi), and BSE infectivity, at 6 mpi, were detected [1–4]. The distal ileum of naturally occurring cases has also been shown to contain PrP^Sc and/or BSE infectivity [3, 5, 6]. Immunohistochemical examinations of the distal ileum revealed PrP^Sc accumulations in Peyer's Patches (PP) as well as in the enteric nervous system (ENS) in an age dependent manner [2, 3]. Younger, orally infected animals up to 36 mpi showed only a staining reaction in a small proportion of their follicles in the Peyer's Patches of the distal ileum [2, 3]. Therein the accumulation of PrP^Sc is initially confined to tingible body macrophages but in the later clinical disease phase a reaction pattern resembling follicular dendritic cells (FDC) can be seen. A positive staining reaction of the ENS is restricted to clinical cases and is the only accumulation detectable in the distal ileum of natural occurring infections [3, 6]. However, evidence for BSE infectivity or PrP^Sc accumulations in other parts of the bovine small intestine, other than the distal ileum, has not been convincingly demonstrated for either experimentally or naturally occurring cases.

Up to a year of age, the ileal PP represent the major gut-associated lymphoid tissue in ruminants possessing an extensive bed of follicular dendritic cells and a specialised epithelium actively engaged in the uptake and transcytosis of macromolecules from the gut, explaining the restriction of PrP^Sc to this region of the gut [7]. How TSE agents cross the epithelium is not exactly known but several mechanisms have been proposed. The first is via the M-cells, a cell type which is associated with the epithelium of the gut and capable to transcytose the scrapie agent in vitro [8]. A recent scrapie study showed that PrP^Sc was transported across the absorptive epithelium of villi into lacteals in vivo [9]. The transmission route using a direct uptake by dendritic cells that can acquire antigens directly from the intestinal lumen cannot be ruled out [10, 11]. After crossing the mucosal barrier TSE infectivity and PrP^Sc first accumulates in the PP and this replication is thought to facilitate further neuro-invasion [12, 13]. In this model neuro-invasion would take place between the submucosal plexus of the ENS and PP though pathways which have yet to be ascertained [14, 15]. However neuro-invasion can occur without apparent involvement of the lympho-reticular system (LRS) in different species suggesting a direct route of infection via the submucosal network of nerve fibres [16] which contains nerves that are present directly adjacent to the basement membrane of villous epithelium [9].

The study presented here concerned cattle of all age groups but focused on the small intestines of pre-clinical BSE cattle and in particular on younger animals between 4 to 24 months of age, because the majority of consumer beef and beef by-products come from that age group. The study determined when BSE infectivity and/or accumulations of PrP^Sc can be found in the small intestine of cattle between the ages of 4 to 6 months which were orally challenged with BSE. To our knowledge this is the first study investigating extensively not only the ileum/ileocaecal-junction but also the jejunum of cattle in preclinical stages of BSE. The results presented provide data for inclusion to qualitative and quantitative risk assessments and give new insights into the early gut associated pathogenesis of BSE.

We examined the gut associated pathogenesis of classical BSE by mapping the exact temporal and spatial emergence and distribution of PrP^Sc in the gut associated lymphoid tissues (GALT) of the small intestines of pre-clinical cattle.

The histopathological examination revealed for all samples, including the controls an eosinophilic enteritis, varying in degrees from moderate to severe. The cause was most likely due to a mild coccidiosis, as shown by light microscopy in some animals.

Ileum

One cow out of the eight mpi (IT39) group and all cows from the 12 mpi group revealed positive and/or inconclusive results in the biochemical tests applied. However, in all samples examined there were no positive reactions by using the BioRad TeSeE rapid test. With the IDEXX HerdChek rapid test, clear reactive samples were detected in the distal ileum of two cows (IT01 and IT57), both at 12 mpi. Inconclusive samples were seen with three cattle at 8 mpi (IT39) and 12 mpi (IT16, IT06). Furthermore four of the cows (IT39, IT01, IT06, IT57) had distinct PrP^Sc accumulation using PTA-WB. After Proteinase-K treatment one sample (IT16), had an ambiguous band pattern. This sample was therefore defined as inconclusive.

There were differences in the apparent level of infectivity between the different parts of the small intestine (Table 3). The ileal samples showed the highest level of infectivity with the shortest incubation period (mean value 320 days) and the highest transmission rates (more than 2/3 of mice were affected) as compared to the jejunal and ileocaecal-junction samples. Infectivity in the distal ileum was detected in 11 of the 16 infected cattle examined by bioassay. The younger cattle, at 8 mpi and 12 mpi, showed a wider and a more constant distribution of infectivity than the older animals. Consequently the highest amount of infectivity was seen in samples examined at 12 mpi, and the lowest at 16 mpi.

Months p.i.	Cow	Localization
		Jejunum A	Jejunum B	Ileum	Ileocaec.
	IT 14	0/10 --	0/14 --	9/13 280 ± 41	0/15 --
8	IT 20	0/14* --	0/13 --	8/11 334 ± 61	0/15* --
	IT 39	1/14* 502	0/12* --	7/11* 310 ± 49	0/12 --
	IT 55	2/10* 379, 522	0/12 --	7/9 332 ± 47	3/13 442, 517, 582
12	IT 01	0/14* --	3/10 407, 432, 600	12/13 305 ± 45	14/15 315 ± 40
	IT 16	0/13 --	0/12 --	6/11 412 ± 31	1/10 395
	IT 06	0/12 --	1/13 375	10/10 279 ± 47	12/12 277 ± 39
	IT 57	0/15 --	2/10 393, 428	11/11 260 ± 55	9/11 297 ± 64
	IT 07	0/14* --	0/15 --	0/12* --	0/9 --
16	IT 28	0/11 --	0/11 --	0/11 --	0/12 --
	IT 46	0/11 --	0/12 --	0/10 --	0/12 --
	IT 65	0/12 --	0/11 --	11/14* 319 ± 41	13/13* 349 ± 53
20	IT 10	0/13 --	2/15 363, 408	0/11* --	0/9 --
	IT 17	0/12* --	0/10 --	0/15* --	12/12 311 ± 33
	IT 50	0/12* --	0/14* --	11/13 353 ± 57	0/12* --
	IT 60	1/12 391	0/9 --	11/12 345 ± 37	0/12 --

The first line indicates the number of positive tested mice/total number of mice, the second line indicates the mean incubation period in days post infection

* augmented groups - additional inoculations after heating the inoculum to 70°C for 10 min

Detectable amounts of PrP^Sc were found in the Ileum of 37 out of the 46 cattle examined by immunohistochemistry. In total 31 cattle showed an accumulation of PrP^Sc in the follicles of the ileal PP. Thereby clear quantitative and qualitative age-related variations in the PrP^Sc accumulations were visible. From 4 mpi to 36 mpi the PrP^Sc was visible in most of the cattle but in the later stages of the incubation period only single animals revealed a detectable amount of PrP^Sc. An obvious undulant pattern was found when counting the follicles with a positive staining reaction as compared to the total numbers of follicles examined. In doing so the first traces of PrP^Sc (less than 1% of the follicles examined) was observed in the 4 mpi group, followed by an abrupt rise in the numbers of positive follicles (up to 20%) at 12 mpi. Within the 16 mpi and 20 mpi group only traces of PrP^Sc were seen again (less than 1% of the follicles examined). However, as seen in the 12 mpi group a high number of positive follicles, up to 25%, were seen at 24 mpi. In contrast animals within the group 28 mpi which only revealed approximately 1% positive follicles. Later stages of the incubation period did not show such clear variations, but higher numbers of positive reactive follicles were observed in single animals at 32, 36, and 40 mpi. No clustering of positive stained follicles was seen in any of the samples.

The cellular reaction patterns of PrP^Sc (Table 1) observed was in a time-dependent manner. With the increase in the number of positive follicles there was a simultaneous increase in the number of cells accumulating PrP^Sc. The first detectable amounts of PrP^Sc were confined to large mononuclear cells, resembling tingible body macrophages (TBM). Initially a few intra-cytoplasmatic granules were only detectable in a few cells. These single punctas within the TBM's (Figure 3A) increased over the incubation period showing, at first, a fine multigranular intracellular appearance followed by a more globular accumulation (Figure 3B), seen in several cells of one follicle. The positive stained cells were mainly located in the light central zones of the follicles and less commonly within the dark zones. A curvilinear staining reaction, mostly seen with the mab 6C2, resembling the follicular dendritic cell (FDC) was observed in single follicles and was associated with an increase of positive stained cells. Initially a very weak linear reaction pattern was observed at 12 mpi and a more distinct net-like staining reaction later in the incubation period with a peak at 24 mpi (Figure 3B).

Accumulation of PrP^Sc in the ENS was observed in 18 of the 46 cattle examined. It was first detected at 16 mpi (IT28). Only a few animals showed detectable amounts of PrP^Sc up to an incubation period of 32 mpi, with a peak at 24 mpi. In these age groups mainly individual plexuses are concerned (Table 1). In contrast, later in the incubation period most of the animals which stained positive in the ENS showed a wider distribution of PrP^Sc within several multifocal distributed plexus. In all cattle, the submucosal layer, as well as the myenteric plexuses, was involved but the latter to a much higher degree (about three fold). An intra-glial, intra-neuronal, peri-neuronal and more rarely a linear reaction pattern can be seen in all plexuses independent of the age of the animals (Figure 4A). In one cow at 12 mpi (Figures 4B and 4C) a clear intracellular staining reaction of some cells located in the submucosa beneath follicles of the PP was observed. An exact identification by morphological characteristics was not possible. Clustering of PrP^Sc in specific localizations/areas was rarely seen as the affected myenteric plexuses were in most cases not associated with adjacent positive follicles. Most of the older animals lacked a simultaneous follicular accumulation of PrP^Sc.

Jejunum

Two animals had inconclusive results using the biochemical tests, IT39 (8 mpi) with both the IDEXX HerdChek rapid test and PTA-Immunoblot and IT47 (24 mpi) using the PTA-Immunoblot.

Positive bioassays were seen in seven cattle examined, mainly in younger cows at 8 and 12 mpi (Figure 5). The jejunal samples revealed the lowest levels of infectivity shown in the present study with distinct longer incubation periods (about 100 days) and low transmission rates with fewer mice per group affected.

An immunohistochemical staining reaction in the jejunal samples was never seen.

The test results showed that nearly all of the cattle (40/46) from 1 mpi to 44 mpi carried BSE infectivity and/or detectable amounts of PrP^Sc in their small intestines. We were able to demonstrate infectivity not only in the ileum, but also in the ileocaecal junction and in the jejunum and in some animals in all three anatomical locations simultaneously.

Samples for biochemical examinations and the bioassay were taken from one PP allowing a direct comparison of these methods. Clear differences in sensitivities were observed. While the BioRad TeSeE failed to detect any amount of PrP^Sc, both the IDEXX HerdChek and the PTA-immunoblot were positive in a number of samples from the ileum/ileocaecal-junction. The most sensitive test, however, was the bioassay, using bovine PrP over-expressing transgenic mice (Tgbov XV), showing infectivity in all parts of the small intestine. The highest amounts of infectivity (high transmission rates, short incubation periods) were found in the ileum, a moderate degree in the ileocaecal junction and traces of infectivity in the jejunum. A high sensitivity was also achieved through IHC, using a second PP. Nearly all animals examined (40/46) revealed detectable amounts of PrP^Sc mainly in the ileum but also in the ileocaecal-junction. In an earlier study only approximately one third of the cattle examined showed positive results in the distal ileum [3]. These differences in sensitivity are most probably due to the larger number of sections prepared in our study. For example the total number of follicles examined here is 3-5 times higher than in comparable studies and the same is true for the plexus numbers. Due to the higher number of samples examined and therefore a higher sensitivity, this allowed us not only the very early detection of PrP^Sc in the ENS of a cow (IT28) at 16 mpi, but also we were able to demonstrate PrP^Sc in the ENS of one cow (IT26) which we reported to be negative two years before [2].

The results presented here add new insights into the gut-associated pathogenesis of BSE in cattle which shows a clear time-dependent pattern. It has been shown for some sheep scrapie infections that a latent period of at least about one month after exposure occurs during which no detectable infection is present [9, 22, 23]. This time lag is supposedly due to insufficient amounts of PrP^Sc for detection by immunohistochemistry [9, 22]. A similar time-dependent pattern for the detection of PrP^Sc is seen in the cattle examined here. However, the time lag between replication of the agent and accumulation of PrP^Sc in detectable amounts by IHC seems to be approximately four months. A shorter time-lag seems unlikely due to the only very weak amounts of PrP^Sc in ileal follicles at 4 mpi. During the first 8 mpi of BSE infection in cattle there seems to be an equilibrium between replication, accumulation and degradation as indicated by weak amounts of PrP^Sc represented by a punctate reaction pattern in TBM's. As the incubation time increased, the balance changed in favour of a massive accumulation of PrP^Sc in follicles of the PP's as seen at 12 mpi. Corresponding results are seen with the bioassay since younger animals up to 12 mpi showed a wide distribution of infectivity affecting all localizations which indicates a multifocal increase of the BSE agent. Infectivity data for the distal ileum presented here are compatible with those reported in UK studies [3, 4, 24]. For the distal ileum a peak at 12 mpi concerned both the number of follicles involved and the amount of PrP^Sc detectable in the individual follicles. Most interestingly, PrP^Sc is detectable in FDC as well as in TBM, indicating increased clearance activities by the latter, as described already for scrapie [13, 25]. The very low (if any) PrP^Sc accumulation and infectivity loads seen at time points 16 mpi and 20 mpi also support this clearance hypothesis. Such a decrease in the infectivity loads in ileal PP's of BSE-infected cattle in 20 mpi animals has also been described before [26].

A second peak of PrP^Sc accumulation, involving both TBM's and FDC was seen at 24 mpi, suggesting a replication cycle (i.e. a shift of the balance between accumulation and degradation in favour of accumulation) for BSE about 12 mpi. Infectivity studies done previously using BoPrP-Tg110 mice showed/revealed a comparable low incubation period of 24 mpi for 1 mouse out of 5 mice [1]. Therefore infectivity data from that time point would be of interest, and will be done in future studies. A third peak of PrP^Sc accumulation in the ileal PP's can be demonstrated, after a clear decrease at 28 mpi, in single cows at 32, 36 and 40 mpi, reflecting an individual variability in older cows.

The exact route of infection for the ENS, which is thought to be the entry point to the peripheral nervous system, still remains unclear [27]. In particular during the pathogenesis of BSE an accumulation of PrP^Sc in the ENS has been rarely reported and was confined to clinical cases [3, 6]. In contrast, in sheep scrapie a wide distribution of PrP^Sc in the ENS occurs, often associated with abundant deposits of PrP^Sc in the adjacent PP's [14, 28] indicating an infection of the ENS via the GALT. However, a direct route via nerve fibres underneath the villous epithelium is also discussed [9, 15]. In the study presented here evidence for both routes of infection can be seen. We observed one cow at 12 mpi which had a slight staining reaction in the submucosa directly adjacent to a follicle, which might reflect an early neuro-invasion of a submucosal plexus. Unfortunately we were not able to clearly characterize the cells involved and macrophages penetrating the submucosa can not completely be ruled out. The first clear evidence for an ENS infection is seen in a single myenteric plexus at 16 mpi without adjacent follicles staining. Moreover, it is striking that in younger cows the accumulations of PrP^Sc were confined to single plexuses, whereas later in the incubation period, in particular associated with end-stage BSE, the number of affected plexuses increased. However, in most of the samples examined a randomly distributed pattern of affected myenteric plexuses is obvious, a clear association to affected follicles is not seen suggesting a direct neuro-invasion without an involvement of the GALT. Nevertheless these results clearly indicate that the ENS is involved in the early propagation of PrP^Sc, but to a much lesser extend as compared to scrapie. Furthermore, the marginally higher number of affected plexus seen in clinical cases in combination with their multifocal distribution pattern could be due to a local accumulation over time, but a limited spread along the ENS cannot be completely ruled out.

Our presented data are relevant for a risk assessment based SRM definition. At least in younger animals up to 12 mpi the jejunum carries BSE infectivity. A recently published quantitative study reported a considerable amount of lymphoid and neural tissues in natural sausage casing produced from cattle small intestine after the cleaning procedure [29]. Cross-contaminations between animals of various ages during the production process as well as during food preparation cannot be ruled out. Moreover, it has to be emphasized that only a very small proportion of the jejunum, (which has an overall length of 25-30 m and contains up to 40 PP) was examined in this current study.

In summary our study shows that nearly all of the BSE-infected cattle that were examined had BSE infectivity and/or detectable amounts of PrP^Sc in their small intestines. The highest amounts of BSE infectivity and/or PrP^Sc were seen in the ileum and ileocaecal junction with lower amounts in the jejunum. Clear time-dependent variations in the detectable amounts of PrP^Sc were visible. Younger cattle killed 8 mpi to 12 mpi carried higher PrP^Sc levels and had a more widespread distribution of PrP^Sc than animals at 16 mpi to 20 mpi. The data from this study show a more widespread distribution of the BSE agent in the small intestines in particular of pre-clinical cattle and are the first describing infectivity in jejunal samples of cattle. Therefore the data presented here are important for the definition of SRM and the implementation of their removal as part of functional public health protection measures against BSE.