| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Veterinary Laboratories Agency (VLA), Woodham Lane, New Haw, Addlestone, Surrey KT15 3NB, UK
2 Canadian Food Inspection Agency, Winnipeg Laboratory, Winnipeg, MB R3E 3M4, Canada
3 Health Canada, Health Products and Food Branch, Tunney's Pasture, Ottawa, ON K1A 0K9, Canada
Correspondence
M. E. Arnold
m.arnold{at}vla.defra.gsi.gov.uk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Present address: National and OIE BSE Reference Laboratories, Animal Diseases Research Institute, Canadian Food Inspection Agency, Township Road 9-1, Lethbridge, AB T1J 3Z4, Canada. 
Present address: Pathology, Infectious Disease and Biosecurity, School of Veterinary Science, University of Queensland, St Lucia, QLD 4072, Australia.
Present address: Oak Farm, Harpsden Bottom, Henley-on-Thames, Oxon RG9 4HY, UK.
Appendices A (including Supplementary Table S1 and Supplementary Fig. S1), B and C are available with the online version of this paper.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
) is an epidemic, feed-borne prion disease of domestic cattle (Wilesmith et al., 1988) that, via food, is considered the origin of the human variant form of Creutzfeldt–Jakob disease. In natural cases of BSE (Fraser & Foster, 1994; MAFF, 1995; Buschmann & Groschup, 2005) and in experimentally orally infected cattle (Wells et al., 1998, 1999, 2005; Terry et al., 2003; Espinosa et al., 2007; Hoffmann et al., 2007; Masujin et al., 2007), the infectivity and the disease-associated isoform of the prion protein (PrPSc) have a relatively restricted tissue distribution. As in other orally acquired prion diseases, the central nervous system (CNS) becomes infected only after part of the incubation period (IP) has elapsed, and thereafter, particularly in BSE, contains most of the infectivity in the animal. Crucial to the control policy for BSE has been assessment of the age at which certain tissues, the specified risk materials (SRM), should be removed from cattle at slaughter for the protection of human and animal health. This assessment with regard to the CNS and associated structures has previously been based on available experimental data from pathogenesis studies and on the epidemiology of BSE with respect to age at infection and age at detection by clinical and active surveillance (European Food Safety Authority, 2005).
Experimental, sequential-kill studies using an oral-exposure dose of 100 g BSE-affected brainstem have provided data on tissue infectivity and/or PrPSc relative to time post-exposure (p.e.) (Wells et al., 1996, 1998, 2005; Grassi et al., 2001; Espinosa et al., 2007; Hoffmann et al., 2007). However, in none of these studies can the observations of tissue infectivity and time of clinical onset be compared directly because, given the sequential-kill protocol of the studies, the IPs of the asymptomatic recipients killed cannot be determined.
Also, a high-dose exposure cannot provide information on the dynamics of involvement of different tissues relative to the much lower doses responsible for the BSE epidemic (Wilesmith et al., 1988; Wells et al., 2007). This quantitative relationship requires consideration of additional data on low-dose exposure and attack rates.
The primary aim of this study was, therefore, to provide estimates of the timing of initial PrPSc detection in the CNS and closely associated peripheral nervous system ganglia, relative to clinical onset, in cattle exposed to 100 or 1 g oral doses of a titrated homogenate of BSE-affected brainstem. The components of the study were threefold. Firstly, a panel of diagnostic methods was applied to tissues from the initial (Wells et al., 1996, 1998, 2005) and subsequent UK sequential-kill, oral-exposure experiments in cattle for the detection of PrPSc relative to time p.e. Methods used were in current use for active surveillance for the detection of BSE cases at the time of the initiation of the study. Secondly, IP distributions were obtained from experimental oral-titration studies to determine attack rate and dose–IP response (Wells et al., 2007) and from survival data in the sequential-kill, oral-exposure experiments in cattle. Thirdly, the relationship between PrPSc detection in tissues from sequential-kill studies and the period before clinical onset was facilitated by the development of a statistical model, which accounted for the differences in the IP distribution and probability of infection between the different dose groups.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 1998, 2005). All procedures described involving animals were approved under the Animal (Scientific Procedures) Act 1986.
Source animals and inoculum.
Material was examined from a previously reported experimental oral exposure of cattle to the BSE agent (experiment 1) (Wells et al., 1996, 1998, 2005), which utilized an inoculum consisting of a homogenate of a pool of brainstems from 75 cases of BSE (BBP12/91) sourced in 1991. End-point titration of the homogenate in RIII mice gave a titre of 103.46 mouse intracerebral+intraperitoneal (i.c.+i.p.) ID50 g–1.
In a second experiment (experiment 2), two groups of 100 cattle from UK farms with no history of BSE were dosed at 4–6 months of age with either 100 or 1 g of a pool of BSE-affected brainstem tissue, derived from 254 cases of BSE sourced in 1996 and 1997. End-point titration of the brain-pool homogenate (SE1736 : BBP1) in RIII mice gave a titre of 103.10 mouse i.c.+i.p. ID50 g–1. A further 100 cattle, sourced similarly to the test animals, served as undosed controls. From the time of dosing, six challenged and three age-matched, unexposed control cattle were killed at 3-monthly intervals p.e., increasing to 6-monthly intervals after the first year p.e. in the case of the 1 g dose group. The logistics of each scheduled cull required that necropsy examinations be spread operationally over a mean of 19 days. If animals presented with clinical signs of BSE or intercurrent disease and required culling earlier than scheduled originally, their place in the timed cull would be taken up by another animal. Ten per cent of the animals in each dose/control group were females and were retained throughout the experiment until clinical disease was evident, and killed with age-matched controls. Those surviving at January 2006 did not form part of this study.
Clinical methods to determine IP.
Clinical monitoring of cattle according to defined approaches (Wells et al., 1996, 2007; Wells & Hawkins, 2004) was maintained throughout the studies and comprised regular clinical observations for behavioural signs associated with BSE (Austin et al., 1994) and neurological examinations (Konold et al., 2004). For the purpose of this study, as described previously (Wells et al., 1996, 2007), IP (expressed in months) was defined as the time between exposure and onset of definite signs of BSE.
An experimental oral titration in cattle to determine the attack rate and IP for a range of amounts of BSE-infected cattle brain, together with survival data from the two oral-exposure, sequential-kill experiments using the same clinical approaches (Wells et al., 2007; Appendix A, available in JGV Online), provided data for the estimation of the IPs according to dose, required for the calculations in the present study. For the oral titration in cattle, groups of calves were dosed with a range of doses: 3x100 g on three successive days, or single doses of 100, 10, 1, 0.1, 0.01 or 0.001 g of a pool of brain tissue from clinical cases of BSE (Wells et al., 2007).
Tissues.
From both sequential-kill experiments, samples of fresh brainstem (three different areas: medulla–obex, rostral medulla and midbrain), spinal cord, dorsal root ganglia (DRG), trigeminal ganglion (TRG), stellate ganglion and cranial cervical ganglion that had been collected aseptically and frozen (–70 °C) were retrieved (subject to availability) for unfixed-tissue methods of PrPSc detection. Adjacent or contralateral samples that had been collected into 10 % formalin were prepared for histopathological and immunohistochemical (IHC) examinations. For further anatomical detail of sampling, the reader is referred to Appendix B (available in JGV Online).
Test methods.
Three techniques were applied for the detection of PrPSc; an ELISA test, Western blotting and IHC. In addition, histological sections of CNS were stained routinely with haematoxylin and eosin (HE) for the morphological detection of spongiform changes. The total numbers of each tissue examined by each method from exposed cattle, according to experiment and dose group, and from undosed control cattle are summarized in Table 1.
|
; Grassi et al., 2001) in accordance with the manufacturer's instructions. For those tissues (ganglia) with a fibrous content compared with the consistency of brain, two additional and larger ceramic beads were added above and below the sample to facilitate homogenization. Samples with an absorbance value (450/620 nm) lower than the cut-off were considered negative, according to the manufacturer's instructions. However, results just below the cut-off value (cut-off –10 %) were treated as inconclusive, and duplicate repeat testing was initiated. Samples with an absorbance value greater than the cut-off value were considered to be initially reactive, and tested in duplicate. Samples that gave inconclusive results (within cut-off –10 %) on repeat testing were considered negative.
Western blotting was carried out by using a method that incorporates detergent extraction, proteinase K digestion and a PrPSc-enhancement step involving precipitation with sodium phosphotungstate (NaPTA) dibasic hydrate (P-6395; Sigma Aldrich) (WBNaPTA) (Stack, 2004). The monoclonal antibody (mAb) used was 6H4 (Prionics Ag), which is a mouse IgG1 antibody raised to the aa 144–152 sequence of human PrP (Korth et al., 1997). Results for these studies were recorded as positive, negative or inconclusive. A positive result was indicated by the detection of protein bands resistant to proteinase K digestion (PrPres). The protein bands comprise a disease-specific triplet pattern corresponding to the detection by mAb 6H4 of diglycosylated, monoglycosylated and unglycosylated regions of PrPres.
For IHC detection of PrPSc, two anti-PrP mAbs, both used widely for confirmatory diagnosis in BSE surveillance, were used to provide some indication of possible differences in sensitivity/specificity between antibodies. R145 (VLA) is a rat anti-PrP C-terminal-specific mAb, raised against the bovine PrP peptide sequence of aa 221–232 (Terry et al., 2003). F99 (Pullman) is a mouse anti-PrP C-terminal-specific mAb, raised against the ovine PrP peptide sequence of aa 217–231 (O'Rourke et al., 2000). To ensure within-animal consistency of IHC, all tissues from each animal were included in the same immunolabelling batch.
Tissue sections were dewaxed and rehydrated routinely. Epitope demasking was performed by immersion of sections for 30 min in undiluted formic acid, then rinsing under running tap water for 15 min, followed by autoclaving at 121 °C in 2 l citrate buffer, pH 6.1 (8.8 mM trisodium citrate dihydrate and 1.3 mM citric acid in purified water) for 30 min. Primary antibodies were applied at dilutions of 1/500 (R145) or 1/1000 (F99) for 1 h at room temperature, with immunodetection performed by using an avidin–biotin–peroxidase complex (Vector Elite) technique using diaminobenzidine chromogen. Sections were counterstained with haematoxylin and examined by light microscopy.
Statistical model of the timing of detected PrPSc relative to clinical onset.
The experimental sequential-kill study data give the timing of the first detection of PrPSc relative to period p.e. However, the timing of the presence of detected PrPSc is likely to depend on either the number of months before clinical onset or the proportion of the IP completed. This means that the probability of detecting PrPSc in any tissue at a given kill time will depend on the IP in the animal from which the tissue was taken. The IP of each non-clinically diseased animal is unknown. However, the statistical distribution of the IP can be estimated from the experimental study of the oral attack rate of BSE infection in cattle (described above and in Appendix A, available in JGV Online) and sequential-kill studies and can be used to derive the statistical distribution for the number of months before onset for subclinical cattle; it gives the IP distribution with the kill time subtracted and truncated to be >0. This distribution enables maximum-likelihood methods to be used to estimate the probability of detecting PrPSc in terms of the number of months before clinical onset, denoted 
(details are given in Appendix C, available in JGV Online). The estimation can also be done in terms of the proportion of the IP completed (Appendix C).
The key output of the statistical model was the probability of detected PrPSc in an infected animal as the number of months before clinical onset increased, denoted (). The parameter () was given by the following logistic-regression curve, commonly used to fit binary data:
|
| (equation 1) |
, 
were parameters estimated from the data.
Using the above model for the timing of first detected PrPSc, the following analyses were performed. (i) The timing of the presence of detected PrPSc was compared for each tissue for the most sensitive of the tests, i.e. parameters , in equation (1) were estimated separately for each tissue for the most sensitive test. (ii) The uncertainty in the predicted test sensitivity of IHC applied to the medulla–obex was investigated. Ninety-five per cent confidence intervals for the ID50 were calculated by non-parametric bootstrap (Efron, 1985), with 10 000 samples and taking into account the uncertainty in the dose-dependent attack rate and IP distribution. (iii) A likelihood-ratio test was performed to determine whether there was a significant difference between the best-fitting parameters for the 1 g- and 100 g-dosed animals (with the 100 g-dosed animals from each experiment pooled due to the small sample size of the 100 g group from experiment 1).
The above models all consider the probability of detecting PrPSc given the number of months before clinical onset. In addition, the timing of first detected PrPSc was estimated in terms of the proportion of the IP completed for IHC applied to the medulla–obex, rostral medulla and midbrain (details are given in Appendix C, available in JGV Online). The log-likelihood was then compared with the model fitted in terms of the number of months before clinical onset to see which provided the better fit to the data.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). The aggregated results for Bio-Rad TeSeE, WBNaPTA, IHC and HE are shown for each CNS tissue at each time point (Table 3).
|
|
Estimating the relationship between CNS detection and IP
The earliest tissue to be detected as positive was the medulla–obex by IHC (Table 3
). Of the three positives in experiment 2, 100 g dose group killed 30–32 months p.e., one was positive at 30 months and two at 32 months. For all other CNS tissues, the timing of first detected PrPSc was at 32 months.
Applying the statistical model (equation 1) to the data in Table 3 to estimate the correlation between IP and detected PrPSc in the medulla–obex for IHC separately for each dose group (i.e. experiment 2, 1 g dose group and the 100 g dose groups of experiments 1 and 2) gave a significant improvement in model fit compared with fitting common parameters for all of the groups together (P<0.01). The parameter values , [from equation (1)] for the 1 g and 100 g groups were respectively 5.94, –3.40 and 8.44, –0.88. These estimates result in 50 % of infected animals being detected (with 95 % bootstrap confidence intervals) at 1.7 (0.2, 4.0) and 9.6 (4.6, 15.7) months before clinical onset for the 1 g and 100 g dose groups, respectively. The 50 % point of detection was 7.0 months before clinical onset when only the lower of the two 100 g dose titres (experiment 2) was considered. Results indicate that, whilst the confidence intervals are wide for each of the dose groups, there is a clear separation between the 1 g and 100 g dose groups in the timing of detected PrPSc relative to clinical onset. The results for the 1 g group are most relevant to the field situation, as a 1 g dose of brain homogenate results in a mean IP consistent with that estimated from data on reported cases, i.e. approximately 5 years (Wilesmith et al., 1988). This implies that PrPSc in cattle may only be detectable with high probability approximately 1.5 months before clinical onset.
The timing of detected PrPSc with respect to the proportion of the IP completed was also significantly different for each of the dose groups (P<0.01). The model predicted that PrPSc in 50 % of infected animals would have been detected at 97 and 79 % of the IP for the 1 g and 100 g groups, respectively (Fig. 1). Interestingly, in six mouse i.p. inoculation models of scrapie, Kimberlin & Walker (1988, 1989) showed that initial detection of replication of agent in the brain (medulla) occurred after a fixed proportion of the IP. However, the fit of the model to the data (as measured by log-likelihood) was poorer than the fit of the models giving the timing of detected PrPSc in number of months before clinical onset.
|
) indicate that PrPSc is likely to be detected earliest in the medulla–obex followed by, with a fractional lag, the rostral medulla. PrPSc is estimated to be detected, on average, in the thoracic and cervical spinal cord approximately 1 month after detection in the medulla–obex, and in the midbrain and lumbar spinal cord approximately 1.3 months after detection in the medulla–obex.
|
), suggesting a possible simultaneous timing of entry of infectivity in these tissues, which would be an important difference from the 100 g group. However, this finding should be treated with caution, as the small number of positives in the 1 g group could mean that a difference in the timing of entry in the different tissues could have been missed.
Sequence of DRG and sympathetic ganglion involvement relative to the CNS
The timing of detection of PrPSc in the TRG and the DRG was concurrent with or after all CNS tissues were positive. In particular, the DRG were not positive until the corresponding level of spinal cord was positive, although PrPSc was detected in the cervical DRG later than in the thoracic DRG, despite the respective spinal-cord levels having similar results. On average, the timing of detection of the TRG and DRG was very late in the IP (Table 4; Fig. 2). This is an important result in terms of consumer protection, as infectivity in the DRG has been identified as a significant contributor to human exposure to BSE infectivity in beef products (Comer & Huntly, 2004). An important assumption in previous risk assessments of potential human exposure to the BSE agent via the food chain was that DRG had levels of infectivity similar to those of the CNS (Comer & Huntly, 2004). The results of the present study call this into question, as PrPSc was detected several times in CNS, but not in DRG. In no case was PrPSc detected in the stellate or cranial cervical ganglia by either of the tests applied (Bio-Rad TeSeE or IHC).
|
). For example, for the 100 g group in experiment 2 for the rostral medulla, there were a total of 28 positives for IHC and 25 for both WBNaPTA and Bio-Rad TeSeE, but only 20 for HE, and the impact of these results on the estimates of timing of detectable PrPSc by each of the tests can be seen in Fig. 3. Whilst there was agreement between the results of the WBNaPTA and Bio-Rad TeSeE for the rostral medulla and midbrain, there was variable agreement for the spinal-cord tissues, with WBNaPTA detecting one, five and seven more positives than the Bio-Rad TeSeE for the lumbar, thoracic and cervical spinal cord, respectively (Table 3).
|
There was complete agreement between the results for WBNaPTA, Bio-Rad ELISA and IHC for the brain tissues taken from 1 g-dosed cattle. This is a consequence of the fact that there were no marginal positives for the 1 g group, i.e. those animals detected as positive in the 1 g group had widespread immunolabelling. This is in contrast to the results of the 100 g-dosed group, where IHC detected cases with minimal immunolabelling that were not detected by WBNaPTA or Bio-Rad TeSeE (Table 3; Fig. 3).
There were no significant differences between the IHC results for the R145 and F99 antibodies applied to the CNS tissues, whereas such differences were observed for the DRG and TRG, with the F99 antibody detecting more positives than R145 (Table 4). There were no positive results for peripheral nerve ganglia by the Bio-Rad TeSeE.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
) and the age of detection for healthy slaughter positives from active surveillance; the similarity of the age at detection of healthy slaughter to the confirmed field cases suggests that detection is late in the IP for healthy slaughter positive animals (M. E. Arnold, unpublished data). The estimates of the timing of detected PrPSc relative to clinical onset are dependent on estimates of the IP distribution from the attack-rate studies and survival data in the sequential-kill, oral-exposure experiments. The onset of definite clinical disease status in these studies was assessed against set criteria from routinely monitored animals. This is not the case for field animals and, therefore, it is possible that the IPs in the field could differ from those in the experimental studies, particularly in countries with a low BSE prevalence where farmers and veterinarians are not familiar with the clinical signs of BSE. Furthermore, the attack rate and IP distributions are themselves subject to uncertainty, as they have been obtained from experimental studies with limited numbers of animals. Therefore, whilst the analysis shows that the estimated timing of detection is significantly different between the dose groups, even when the uncertainty in the IP distribution and attack rate is allowed for, there are wide confidence intervals for the point at 50 % detection.
The study shows a clear ordering of the relative timings at which each part of the CNS has detectable PrPSc and, whilst there are only small differences in the timings (Fig. 2), this feature has not been reported previously in the pathogenesis of BSE. The initial site relative to time p.e. in which PrPSc was detected in the CNS of 100 g-dosed cattle was the medulla–obex (Fig. 2). In the 1 g-dosed group, PrPSc was first detected simultaneously in all brainstem areas examined and this preceded detection in the spinal cord. The selection of the medulla–obex region for sampling for statutory diagnosis, originally based on morphological changes (Wells et al., 1989), therefore remains optimal for immunochemical diagnostic approaches.
The p.e. sequence of different CNS areas testing positive for both dose levels (Fig. 2) is consistent with the notional pathogenetic routeing of agent from the intestine to the CNS via sympathetic and parasympathetic components of the autonomic nervous system (Wells & Wilesmith, 2004; Hoffmann et al., 2007). Detection of PrPSc in the brainstem prior to the spinal cord suggests that, in this model at least, the parasympathetic pathway, involving the vagus nerve, provides the faster route. This routeing of the spread of agent has also been shown in experimental rodent models after oral exposure to scrapie agents (Kimberlin & Walker, 1989; Baldauf et al., 1997; Beekes et al., 1998; McBride & Beekes, 1999; McBride et al., 2001) and in sheep with scrapie (van Keulen et al., 2000). The sequence of PrPSc detection by IHC, illustrated in the 100 g dose group (Fig. 2b), i.e. medulla–obex, rostral medulla and then midbrain, and thoracic and cervical spinal cord, followed by lumbar spinal cord, is entirely in keeping with the view that the thoracic spinal cord and the medulla are primary sites of infection of the CNS, with the brainstem marginally preceding spread from the periphery to the spinal cord.
Detection of PrPSc in sensory ganglia, inconsistently late in the IP and mostly later than PrPSc was detected in the corresponding level of the CNS neuraxis, is consistent with secondary spread of infectivity to the ganglion from the CNS rather than their involvement in any primary pathogenetic routeing from peripherally infected tissues. The apparent absence of PrPSc in ganglia of the sympathetic chain is of interest in that, even if, consistent with the current hypothesis of the peripheral pathogenesis of BSE, spread of infectivity from the intestine to the CNS involves, at least in part, the sympathetic efferents of the autonomic nervous system, subsequent spread throughout the sympathetic ganglion chain may be limited or confined to cases with clinical disease.
The results of this study give information on the timing of detected PrPSc in the CNS by methods used for diagnosis in the course of targeted active surveillance of BSE, but caution is required in the interpretation and relevance of these results for the field situation with respect to diagnostic sensitivities. It was recognized from the outset that, in common with all such studies, methods applicable to unfixed tissues cannot be performed on the same tissue sample as those requiring fixed tissue. Furthermore, with respect to unfixed tissues, there are important differences in sampling in these experiments from sampling for routine diagnosis. These include different anatomical sites to those used diagnostically and differences in the sampling approach at given sites. In particular, the tissue used for Bio-Rad TeSeE testing for active surveillance is the medulla at the level of the obex, which was not the tissue level used for the Bio-Rad TeSeE in this study, and it has been established that spatial separation of sampling sites along the medulla neuraxis affects rapid test sensitivities (Schimmel et al., 2002). It is also apparent that sampling and subsampling for rapid test methods in the course of surveillance are also subject to variation in approach and efficiency (M. M. Simmons, unpublished data). However, whilst this study was not able to apply the Bio-Rad TeSeE to the medulla–obex and, hence, we cannot infer the timing of detectable infectivity directly for field cases, there was no difference in test results between any of the tests applied to the 1 g group and, therefore, it is likely that these results are representative of the sensitivity of Bio-Rad TeSeE for the medulla–obex for 1 g-dosed cattle.
In the 100 g dose group, IHC detected positives where WBNaPTA and Bio-Rad TeSeE were negative. The improved performance of IHC is probably due to the nature of the sampling, giving it an ability in sections to detect cases with minimal focal or localized immunolabelling, possibly missed by the subsampling of tissue for WBNaPTA and Bio-Rad TeSeE. It is unclear why such cases only occurred with the 100 g dose group. It is possible that the small number of positives in the 1 g group was such that cases with minimal immunolabelling did not arise by chance or that there is a real difference in the dynamics of pathogenesis between the 100 g and 1 g groups. If it is assumed that, at the clinical stage of disease, cattle with BSE, irrespective of dose, have similar levels of infectivity in the CNS, the results of the 100 g dose group could be explained by a spread and accumulation of PrPSc in the CNS over a longer preclinical period than in the 1 g dose group, whereas, paradoxically, in the latter, accumulation of PrPSc to detectable levels has been a relatively sudden event, late in the IP.
The estimate of the number of months before onset that PrPSc in naturally infected cattle could be detected was lower than the most pessimistic of scenarios considered in a review of the age at which cattle should be allowed to enter the food chain in the UK (Arnold & Wilesmith, 2003; Ferguson & Donnelly, 2003). This lower estimate of test sensitivity would result in higher estimated human exposure in terms of the levels of cattle oral ID50 entering the food chain, given the same assumed estimated levels of cattle infectivity with age. Conversely, however, a later point of detectable infectivity implies a possible lower level of infectiousness. In other words, it may be that infectivity is occurring in the CNS tissues later than was estimated in previous risk assessments of potential human exposure to the BSE agent (Comer & Huntly, 2004), where infectivity in the CNS was 3 logs lower than the clinical value at 70 % of the IP, and 4.5 logs lower at 50 % of the IP. Therefore, the interpretation of the results of the present study in terms of quantitative estimates of human exposure is unclear. What is clear is that lower estimates of test sensitivity will affect back-calculation estimates of cattle infection prevalence, where data on post-mortem testing are a key component. Lower estimates of test sensitivity result in higher estimates of infection prevalence, as a higher proportion of true positives will be missed by the testing.
Dependent upon surveillance data on age-specific incidence and stage of a BSE epidemic within a country, the results for the 1 g dose group would indicate that, given the minimum IP for this dose from these and the attack-rate data, detection of PrPSc in the CNS and DRG in the majority of animals is likely only after 42 months p.e. Given that calfhood exposures to contaminated feed resulted in most cattle becoming infected in the first 6 months of life (Wilesmith et al., 1992; Arnold & Wilesmith, 2004), this corresponds to an age range at detection of approximately 42–48 months. Infectivity, detected by assay in transgenic mice overexpressing the bovine PrP gene, found in brainstem material pooled from three animals killed at 27 months p.e. from experiment 2 (Espinosa et al., 2007) and the finding of PrPSc in the CNS of an animal killed at 24 months p.e. after oral exposure to a 100 g dose of BSE-affected brain, in a separate study (Hoffmann et al., 2007), are consistent with the results of the modelling reported here, given a minimum IP of 31 months for a 100 g dose in the attack-rate studies (Wells et al., 2007) and assuming possible detection some 9 months before clinical signs.
However, the estimated mean IP of cattle infected in the field, which is in the range of 5.0–5.5 years (Wilesmith et al., 1988; Arnold & Wilesmith, 2004), calls into question the practical relevance of a 100 g exposure dose. Clearly, a precautionary approach to SRM controls must allow a margin for differences in the sensitivities of detection of PrPSc and infectivity, particularly with respect to the apparently anomalous dose-dependent timing of PrPSc detection relative to clinical onset, but these findings, taken in concert with epidemiological data, offer considerable scope for modulation of current regulations.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Arnold, M. E. & Wilesmith, J. W. (2004). Estimation of the age-dependent risk of infection to BSE of dairy cattle in Great Britain. Prev Vet Med 66, 35–47.[CrossRef][Medline]
Austin, A. R., Hawkins, S. A. C., Kelay, N. S. & Simmons, M. M. (1994). New observations on the clinical signs of BSE and scrapie. In A Consultation on BSE with the Scientific Veterinary Committee of the Commission of the European Communities, held in Brussels 14–15 September 1993, pp. 277–287. Edited by R. Bradley & B. Marchant. Document VI/4131/94-EN. Brussels: European Commission.
Baldauf, E., Beekes, M. & Diringer, H. (1997). Evidence for an alternative direct route of access for the scrapie agent to the brain bypassing the spinal cord. J Gen Virol 78, 1187–1197.[Abstract]
Beekes, M., McBride, P. A. & Baldauf, E. (1998). Cerebral targeting indicates vagal spread of infection in hamsters fed with scrapie. J Gen Virol 79, 601–607.[Abstract]
Buschmann, A. & Groschup, M. H. (2005). Highly bovine spongiform encephalopathy-sensitive transgenic mice confirm the essential restriction of infectivity to the nervous system in clinically diseased cattle. J Infect Dis 192, 934–942.[CrossRef][Medline]
Comer, P. J. & Huntly, P. J. (2004). Exposure of the human population to BSE infectivity over the course of the BSE epidemic in Great Britain and the impact of changes to the Over Thirty Month Rule. J Risk Res 7, 523–543.[CrossRef]
Deslys, J. P., Comoy, E., Hawkins, S., Simon, S., Schimmell, H., Wells, G., Grassi, J. & Moynagh, J. (2001). Screening slaughtered cattle for BSE. Nature 409, 476–478.[Medline]
Efron, B. (1985). Bootstrap confidence intervals for a class of parametric problems. Biometrika 72, 45–58.
Espinosa, J. C., Morales, M., Castilla, J., Rogers, M. & Torres, J. M. (2007). Progression of prion infectivity in asymptomatic cattle after oral bovine spongiform encephalopathy challenge. J Gen Virol 88, 1379–1383.
European Food Safety Authority (2005). Annex to the Opinion. Report of the Working Group on the assessment of the age limit in cattle for the removal of certain specified risk materials (SRM). http://www.efsa.europa.eu/etc/medialib/efsa/science/biohaz/biohaz_opinions/opinion_annexes/933.Par.0001.File.dat/biohaz_report_ej220_srmremove_en1.pdf. EFSA J 220, 1–21.
Ferguson, N. M. & Donnelly, C. A. (2003). Assessment of risk posed by bovine spongiform encephalopathy in cattle in Great Britain and the impact of changes to current control measures. Proc Biol Sci 270, 1579–1584.
Fraser, H. & Foster, J. (1994). Transmission to mice, sheep and goats and bioassay of bovine tissues. In Transmissible Spongiform Encephalopathies: A Consultation on BSE with the Scientific Veterinary Committee of the Commission of the European Communities, held in Brussels, 14–15 September 1993, pp. 145–159. Edited by R. Bradley & B. Marchant. Document VI/4131/94-EN. Brussels: European Commission Agriculture.
Grassi, J., Simon, S., Créminon, C., Frobert, Y., Comoy, E., Trapmann, S., Schimmel, H., Hawkins, S. A. C., Wells, G. A. H. & other authors (2001). Rapid test for the preclinical post-mortem diagnosis of BSE in central nervous system tissue. Vet Rec 149, 577–582.
Hoffmann, C., Ziegler, U., Buschmann, A., Weber, A., Kupfer, L., Oelschlegel, A., Hammerschmidt, B. & Groschup, M. H. (2007). Prions spread via the autonomic nervous system from the gut to the central nervous system in cattle incubating bovine spongiform encephalopathy. J Gen Virol 88, 1048–1055.
Kimberlin, R. H. & Walker, C. A. (1988). Incubation periods in six models of intraperitoneally injected scrapie depend mainly on the dynamics of agent replication within the nervous system and not the lymphoreticular system. J Gen Virol 69, 2953–2960.
Kimberlin, R. H. & Walker, C. A. (1989). Pathogenesis of scrapie in mice after intragastric infection. Virus Res 12, 213–220.[CrossRef][Medline]
Konold, T., Bone, G., Ryder, S., Hawkins, S. A. C., Courtin, F. & Berthelin-Baker, C. (2004). Clinical findings in 78 suspected cases of bovine spongiform encephalopathy in Great Britain. Vet Rec 155, 659–666.
Korth, C., Stierli, B., Streit, P., Moser, M., Schaller, O., Fischer, R., Schulz-Schaeffer, W., Kretzschmar, H., Raeber, A. & other authors (1997). Prion (PrPSc)-specific epitope defined by a monoclonal-antibody. Nature 390, 74–77.[CrossRef][Medline]
MAFF (1995). Bovine Spongiform Encephalopathy in Great Britain: a Progress Report. London: Ministry of Agriculture, Fisheries and Food.
Masujin, K., Matthews, D., Wells, G. A. H., Mohri, S. & Yokoyama, T. (2007). Prions in the peripheral nerves of bovine spongiform encephalopathy-affected cattle. J Gen Virol 88, 1850–1858.
McBride, P. A. & Beekes, M. (1999). Pathological PrP is abundant in sympathetic and sensory ganglia of hamsters fed with scrapie. Neurosci Lett 265, 135–138.[CrossRef][Medline]
McBride, P. A., Schulz-Schaeffer, W. J., Donaldson, M., Bruce, M., Diringer, H., Kretzschmar, H. A. & Beekes, M. (2001). Early spread of scrapie from the gastrointestinal tract to the central nervous system involves autonomic fibers of the splanchnic and vagus nerves. J Virol 75, 9320–9327.
O'Rourke, K. I., Baszler, T. V., Besser, T. E., Miller, J. M., Cutlip, R. C., Wells, G. A., Ryder, S. J., Parish, S. M., Hamir, A. N. & other authors (2000). Preclinical diagnosis of scrapie by immunohistochemistry of third eyelid lymphoid tissue. J Clin Microbiol 38, 3254–3259.
Schimmel, H., Catalani, P., Le Guern, L., Prokisch, J., Philipp, W., Trapmann, W., Zeleny, R. & Moynagh, J. (2002). The Evaluation of Five Rapid Tests for the Diagnosis of Transmissible Spongiform Encephalopathy in Bovines (2nd Study). http://ec.europa.eu/food/food/biosafety/bse/bse42_en.pdf
Stack, M. J. (2004). Western immunoblotting techniques for the study of transmissible spongiform encephalopathies. In Techniques in Prion Research, pp. 97–116. Edited by S. Lehmann & J. Grassi. Berlin: Birkhäuser Verlag.
Terry, L. A., Marsh, S., Ryder, S. J., Hawkins, S. A. C., Wells, G. A. H. & Spencer, Y. I. (2003). Detection of disease-specific PrP in the distal ileum of cattle orally exposed to the BSE agent. Vet Rec 152, 387–392.
van Keulen, L. J. M., Schreuder, B. E. C., Vromans, M. E. W., Langeveld, J. P. M. & Smiths, M. A. (2000). Pathogenesis of natural scrapie in sheep. Arch Virol Suppl 16, 57–71.[Medline]
Wells, G. A. H. & Hawkins, S. A. C. (2004). Animal models of transmissible bovine spongiform encephalopathies: experimental infection, observation and tissue collection. In Techniques in Prion Research, pp. 37–71. Edited by S. Lehmann & J. Grassi. Basel: Birkhäuser Verlag.
Wells, G. A. H. & Wilesmith, J. W. (2004). Bovine spongiform encephalopathy and related diseases. In Prion Biology and Diseases, 2nd edn, pp. 595–628. Edited by S. Prusiner. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.
Wells, G. A. H., Scott, A. C., Johnson, C. T., Gunning, R. F., Hancock, R. D., Jeffrey, M., Dawson, M. & Bradley, R. (1987). A novel progressive spongiform encephalopathy in cattle. Vet Rec 121, 419–420.[Medline]
Wells, G. A. H., Hancock, R. D., Cooley, W. A., Richards, M. S., Higgins, R. J. & David, G. P. (1989). Bovine spongiform encephalopathy: diagnostic significance of vacuolar change in selected nuclei of the medulla oblongata. Vet Rec 125, 521–524.[Abstract]
Wells, G. A. H., Dawson, M., Hawkins, S. A. C., Austin, A. R., Green, R. B., Dexter, I., Horigan, M. W. & Simmons, M. M. (1996). Preliminary observations on the pathogenesis of experimental bovine spongiform encephalopathy. In Bovine Spongiform Encephalopathy – The BSE Dilemma, pp. 28–44. Edited by C. J. Gibbs, Jr. New York: Springer.
Wells, G. A. H., Hawkins, S. A. C., Green, R. B., Austin, A. R., Dexter, I., Spencer, Y. I., Chaplin, M. J., Stack, M. J. & Dawson, M. (1998). Preliminary observations on the pathogenesis of experimental bovine spongiform encephalopathy (BSE): an update. Vet Rec 142, 103–106.
Wells, G. A. H., Hawkins, S. A. C., Green, R. B., Spencer, Y. I., Dexter, I. & Dawson, M. (1999). Limited detection of sternal bone marrow infectivity in the clinical phase of experimental bovine spongiform encephalopathy (BSE). Vet Rec 144, 292–294.
Wells, G. A. H., Spiropoulos, J., Hawkins, S. A. C. & Ryder, S. J. (2005). Pathogenesis of experimental bovine spongiform encephalopathy: preclinical infectivity in tonsil and observations on the distribution of lingual tonsil in slaughtered cattle. Vet Rec 156, 401–407.
Wells, G. A. H., Konold, T., Arnold, M. E., Austin, A. R., Hawkins, S. A. C., Stack, M., Simmons, M. M., Lee, Y. H., Gavier-Widén, D. & other authors (2007). Bovine spongiform encephalopathy: the effect of oral exposure dose on attack rate and incubation period in cattle. J Gen Virol 88, 1363–1373.
Wilesmith, J. W., Wells, G. A. H., Cranwell, M. P. & Ryan, J. B. M. (1988). Bovine spongiform encephalopathy: epidemiological studies. Vet Rec 123, 638–644.[Abstract]
Wilesmith, J. W., Ryan, J. B. M. & Hueston, W. D. (1992). Bovine spongiform encephalopathy: case control studies of calf feeding practices and meat and bone meal inclusion in proprietary concentrates. Res Vet Sci 52, 325–331.[Medline]
Received 8 March 2007;
accepted 18 July 2007.
This article has been cited by other articles:
|
|
 |
|
  Y. Tang, W. Xiang, S. A. C. Hawkins, H. A. Kretzschmar, and O. Windl Transcriptional Changes in the Brains of Cattle Orally Infected with the Bovine Spongiform Encephalopathy Agent Precede Detection of Infectivity J. Virol., September 15, 2009; 83(18): 9464 - 9473. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. J. Sohn, Y. H. Lee, R. B. Green, Y. I. Spencer, S. A. C. Hawkins, M. J. Stack, T. Konold, G. A. H. Wells, D. Matthews, I. S. Cho, et al. Bone marrow infectivity in cattle exposed to the bovine spongiform encephalopathy agent Vet Rec., February 28, 2009; 164(9): 272 - 273. [Full Text] [PDF]
|
|
|
|
 |
|
 Y. H. Lee, M. M. Simmons, S. A. C. Hawkins, Y. I. Spencer, P. Webb, M. J. Stack, and G. A. H. Wells Detection of Pathologic Prion Protein in the Olfactory Bulb of Natural and Experimental Bovine Spongiform Encephalopathy Affected Cattle in Great Britain Vet. Pathol., January 1, 2009; 46(1): 59 - 62. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
| J MED MICROBIOL | ALL SGM JOURNALS | |
