Does the Presence of Scrapie Affect the Ability of Current Statutory Discriminatory Tests To Detect the Presence of Bovine Spongiform Encephalopathy?

Abstract

Current European Commission (EC) surveillance regulations require discriminatory testing of all transmissible spongiform encephalopathy (TSE)-positive small ruminant (SR) samples in order to classify them as bovine spongiform encephalopathy (BSE) or non-BSE. This requires a range of tests, including characterization by bioassay in mouse models. Since 2005, naturally occurring BSE has been identified in two goats. It has also been demonstrated that more than one distinct TSE strain can coinfect a single animal in natural field situations. This study assesses the ability of the statutory methods as listed in the regulation to identify BSE in a blinded series of brain samples, in which ovine BSE and distinct isolates of scrapie are mixed at various ratios ranging from 99% to 1%. Additionally, these current statutory tests were compared with a new in vitro discriminatory method, which uses serial protein misfolding cyclic amplification (sPMCA). Western blotting consistently detected 50% BSE within a mixture, but at higher dilutions it had variable success. The enzyme-linked immunosorbent assay (ELISA) method consistently detected BSE only when it was present as 99% of the mixture, with variable success at higher dilutions. Bioassay and sPMCA reported BSE in all samples where it was present, down to 1%. sPMCA also consistently detected the presence of BSE in mixtures at 0.1%. While bioassay is the only validated method that allows comprehensive phenotypic characterization of an unknown TSE isolate, the sPMCA assay appears to offer a fast and cost-effective alternative for the screening of unknown isolates when the purpose of the investigation was solely to determine the presence or absence of BSE.

INTRODUCTION

The transmissible spongiform encephalopathies (TSEs) are fatal neurodegenerative diseases of animals, of which scrapie in small ruminants (SRs) is the archetype. Scrapie has been recognized as a disease in sheep and goats for almost 300 years although many aspects of the disease are still poorly understood. Despite the relative uniformity of the clinical signs, scrapie can be caused by “strains” with differing biological and biochemical features (1, 2). Polymorphisms in the host PRNP gene, which encodes the cellular protein PrP^C, also influence scrapie susceptibility, strain selection, and the ultimate disease phenotype displayed by the host (3, 4). Historically, the diversity of the scrapie agent has been demonstrated by the serial passaging of natural isolates to a panel of inbred mouse lines, but this does not provide a comprehensive and reliable picture of the diversity of the TSE agents in small ruminants. However, transgenic mice are proving to be susceptible to a wider range of TSEs, enabling more comprehensive characterization of strains (2). In 1998, the definition of ovine TSEs was extended by the discovery, in Norway, of an experimentally transmissible neurological disease of sheep that was clearly distinguishable by all phenotypic parameters from the classical cases that had been reported so far. It was therefore considered to be an atypical form of scrapie (4).

Despite similar diseases occurring in humans (e.g., see reference 5), the animal TSEs were not regarded as zoonotic until the emergence in 1996 of variant Creutzfeldt-Jakob disease (vCJD) linked to bovine spongiform encephalopathy (BSE) (6), which was first described in cattle in the 1980s (7).

Experimental studies in food animal species showed the potential transmissibility of BSE to a range of alternative hosts (8), and it became clear that the small ruminant population was potentially exposed to infection by dissemination through concentrate feed, which may have contained contaminated meat and bone meal, implicated as the origin of the BSE epidemic in cattle (9).

Experimental studies in sheep demonstrated that the disease can result from oral challenge with cattle BSE (10) and, once established, can transmit naturally (11). In addition, the biological properties of the resulting ovine BSE in laboratory models indicate a potentially enhanced virulence for other species including humans (12, 13).

Although ovine BSE has not yet been identified in the field, two naturally occurring cases of caprine BSE have been reported (14–16). As a consequence of this potential risk of BSE in the small ruminant (SR) population, the current European Commission (EC) regulations (999/2001 as amended 36/2005) require the discriminatory testing of all TSE-positive SR surveillance samples to enable the discrimination of BSE from classical scrapie in these samples.

The phenotype of experimental ovine BSE (17) bears a clear resemblance to naturally occurring scrapie, which is endemic in many sheep populations. However, there are some subtle differences in the biochemical signatures of these diseases. The disease-specific isoform (PrP^Sc) of the normal host protein (PrP^C) is the target for all current TSE biochemical diagnostic tests. Depending on the TSE isolate, e.g., BSE or one of the various forms of scrapie, there are differences in the molecular location of protease K (PK) cleavage sites and/or relative PK sensitivity. These differences, as visualized by comparing the relative binding of antibodies against various epitopes around these PK cleavage sites, form the basis of the discriminatory tests currently listed in the regulations, which use either immunohistochemistry (IHC) (18), Western blot (WB) analysis (19, 20), or enzyme-linked immunosorbent assay (ELISA) (21) formats.

In the absence of identified naturally occurring ovine BSE, the development and evaluation of the discriminatory tests that form the basis of the current European Union statutory requirements was based on panels of samples comprising naturally occurring classical scrapie, experimentally induced ovine BSE, and bovine BSE, all of which were demonstrated to be readily distinguishable by these tests (22).

The only experimental study that has been undertaken (23) suggests that the situation with coinfection in sheep is complicated; the WB and IHC data from central nervous system (CNS) tissues resemble classical scrapie, while in the lymphoreticular system (LRS) they may resemble either classical scrapie or BSE. A subsequent bioassay using two transgenic models (tg110, a line which overexpresses bovine PrP on a null murine PrP background [24], and tg338, a line that overexpresses a VRQ allele of the ovine PrP on a null murine PrP background [25]), has shown that BSE and scrapie can be identified by bioassay from the brains of these sheep despite only a classical scrapie signal being detectable on initial screening in WB and IHC (23). These two transgenic mouse lines, used in combination, are widely accepted to be a robust approach to biological discrimination because the ovinized line would preferentially propagate scrapie isolates and the bovinized line would preferentially propagate BSE. They are endorsed by the European Union reference laboratory (EURL) strain typing group in the guidance document for discriminatory testing in the context of European Union regulation 36/2005 (http://www.tse-lab-net.eu/documents/tse-oie-rl-handbook.pdf).

Although not a widely reported occurrence, there is evidence that animals can be naturally coinfected with atypical and classical scrapie (26) and experimentally coinfected with BSE and scrapie (23, 27). There are concerns that current in vitro tests would not be able to provide reliable discrimination in situations where a sheep or goat was coinfected with BSE and scrapie. This uncertainty applies to a lesser extent to bioassay models using wild-type mice (27, 28), although transgenic models offer greater potential through differing strain susceptibility (29–31).

Any in vitro mixing study cannot replicate a natural host coinfection situation, and results must not be extrapolated in that way. However, a preliminary estimate of how tests might perform can be sought through the testing of scrapie samples spiked with BSE and vice versa. Even this in vitro approach cannot cover all the potential variables relating to the scrapie isolates from donors and issues such as the timing and route of infection (e.g., temporally separated or not, same route or not, compatibility of donor and recipient genotypes, age at challenge, etc.) without becoming unworkably complicated, so the study reported here must be viewed in this context.

Recently, a new in vitro method was developed (32) that exploits the differential amplification of PrP^Sc from BSE and classical scrapie sources in different substrates using serial protein misfolding cyclic amplification (sPMCA), with retention of strain-specific biochemical characteristics. Although presently sPMCA is not a statutory test, it was included retrospectively in this blinded assessment to enable a preliminary direct comparison of this new discriminatory approach with the current statutory discriminatory methods, including the bioassay gold standard.

MATERIALS AND METHODS

Materials.

Experimentally generated ovine classical BSE (17) and four different naturally occurring scrapie isolates, which were characterized pathologically, biochemically, and biologically, were sourced (see Table 1 for details) and prepared as a 10% (wt/vol) homogenate in 0.85% sterile saline (33, 34). For each scrapie source selected, the ovine BSE was mixed with it at 1%, 10%, 25%, 50%, 75%, 90%, or 99% based on volume. These mixtures, together with the neat ovine BSE and scrapie samples, were subdivided into aliquots, blinded, and stored frozen as test panels comprising 33 samples (panel 1).

TABLE 1.

Estimation of PrP^res concentrations in the brains of the sources that contributed to the mixtures

TSE source	PrPres (μg/ml)	Normalized ratio for BSE	Log10
BSEa	2.2	1	0.00
VRQ/VRQ classical scrapieb	1.0	0.45	−0.34
ARQ/ARQ classical scrapiec	12.0	5.45	0.74
1-4-7 ARQ/ARQ classical scrapied	9.3	4.23	0.63
Atypical scrapie (AHQ/ARQ)e	4.1	1.86	0.27

^aExperimental ovine BSE was produced and characterized at APHA.

^bClassical scrapie case supplied by Olivier Andreoletti and characterized in tg338 mice (PG 127 classical scrapie [39]).

^cClassical scrapie field case characterized at APHA (scrapie 67 [40]).

^dClassical scrapie field case characterized at APHA (scrapie 19 [40]). The designation 1-4-7 indicates a lesion profile in RIII mouse bioassay with peaks in areas 1, 4, and 7, which is the same as that seen with BSE isolates.

^eAtypical scrapie field case (United Kingdom active surveillance) characterized and titrated (10^6.92 50% lethal dose [LD₅₀]/g) at APHA (M. M. Simmons, J. Spiropoulos, C. M. Vickery, and M. J. O'Connor, unpublished data).

Estimations of the PrP^res (the proteinase K [PK]-resistant moiety of PrP^Sc, which is usually detected by biochemical tests such as Western blotting and ELISA) present in each “neat” sample were obtained so that estimates of the relative proportions of PrP^res contributed by the two components of the mixture may be calculated retrospectively, but this did not affect the choice of material, since known biological phenotype criteria (i.e., distinct bioassay characteristics) were considered more relevant for this study.

The sample panels were provided blinded to teams in the Animal and Plant Health Agency (APHA) (for discriminatory Western blotting) and CEA laboratories (for discriminatory ELISA), and the two laboratories were asked to provide an initial interpretation of BSE-like or not BSE-like before samples were decoded; samples were also provided to ADAS/University of Nottingham for sPMCA analysis and were similarly interpreted before decoding.

All the neat sources used to prepare the panel were selected for bioassay in transgenic mice, together with the mixture with the highest undetectable percentage of BSE and the mixture containing 1% BSE from each scrapie-BSE combination. For animal bioassays, bovinized (tg110) and ovinized (tg338) mouse lines were used. A total of 24 assays were performed.

Discriminatory Western blotting.

The samples were subjected to discriminatory Western immunoblotting using the APHA Bio-Rad hybrid Western blot method, as described in detail in the European Union discriminatory testing handbook (http://www.tse-lab-net.eu/documents/tse-oie-rl-handbook.pdf). TSE strains can be characterized as classical or atypical based on the profile of the protein bands detected by different antibodies following Western immunoblotting. Within this study, the PrP forms in the sample panel tested originated from either classical scrapie, classical ovine BSE, or atypical scrapie.

Classical forms of TSE typically present with a 3-band profile consisting of a diglycosylated (top), monoglycosylated (middle), and diglycosylated (bottom) pattern, and a high molecular mass migration of the unglycosylated PrP band with the core antibody (SHA31) and a similar or stronger intensity of overall signal with the N-terminal antibody (P4) is observed for classical scrapie. A low molecular mass migration of the unglycosylated PrP band with the core antibody (SHA31) and a much reduced, or lack of, intensity of overall signal with the N-terminal antibody (P4) is observed for classical ovine BSE. Atypical scrapie typically presents with a 4-band profile where all bands give a distinctive downward shift with the lowest band at ≤15 kDa.

The resulting band profile for each sample was visually assessed and categorized based on the above criteria (Fig. 1). The initial WB results were reported as BSE, scrapie, or a mixture in which the BSE and each of the scrapie sources were provisionally classified

FIG 1.

Representative Western blots for blinded samples, using MAbs SHa31 (upper) and P4 (lower). This is an original diagnostic blot, selected to illustrate the nature of the blots on which the initial blind interpretation was made. The samples represented in lanes 1 to 14 contained either 100% of a pure form of each TSE strain used in the study (lanes 1 to 5) or various mixtures of the TSE strains (lanes 6 to 14). Mixed samples exhibit multiple-band patterns dependent on the actual TSE strains contained in each mix. Interpretation of each sample was based on the observation of the main characteristics described in Discriminatory Western blotting. Detailed description of the samples and their interpretations in lanes 1 to 14 can be found in Table 2. B, classical bovine BSE control; O, classical ovine scrapie control; M, molecular mass marker.

Discriminatory ELISA.

The samples were tested in duplicate using the discriminatory ELISA method that has been described in detail elsewhere (21). This method treats each sample with one of two different PK digestion protocols (mild and stringent) and expresses the subsequent differences in antibody binding as a ratio. This ratio is further normalized against the BSE control sample in each assay run. Three internal controls were included, one classical scrapie sample (highly PK-resistant normalized ratio inferior to 0.3), an unusual scrapie sample previously reported (21) to give an intermediate result (PK-resistant normalized ratio comprising between 0.3 and 0.7), and an experimental BSE sample (PK-sensitive normalized ratio comprising between 0.7 and 1.3). According to these values, the blinded samples were categorized as scrapie, intermediate scrapie, or BSE. Samples with a normalized ratio above 1.3 were classified as atypical scrapie.

PrP^res estimation.

Aliquots of the unmixed 10% brain homogenates (in normal saline) of ovine BSE and scrapie were pelleted out by high-speed centrifugation and rehomogenized in Bio-Rad TeSeE ELISA kit proprietary buffer to give 20% wt/vol as specified in the kit instructions. A dilution series of each isolate was prepared, and samples were analyzed in accordance with the manufacturer's instructions. The endpoint dilution assays were used to generate PrP^res protein estimations for each homogenate (Table 1), and these were normalized for the ovine BSE sample.

Animal bioassays.

All intracerebral inoculations were carried out under general anesthesia and in accordance with the United Kingdom Animal (Scientific Procedures) Act 1986, under License from the United Kingdom Government Home Office (project license 70/7159). Such licenses are only granted following approval by the internal APHA ethical review process as mandated by the home office.

Each inoculum (10% wt/vol brain homogenate in normal saline) was used to challenge 10 tg110 and 10 tg338 mice intracerebrally (20 μl per mouse). The mice were allowed to develop TSE disease and were euthanized when they reached terminal disease stage or due to other welfare reasons. At postmortem, each brain was sectioned parasagittally; two-thirds was fixed and subsequently processed for histology and immunohistochemistry while the remaining one-third was kept frozen.

All samples were examined for the presence of TSE-specific vacuolation in hematoxylin and eosin (H&E) slides and for PrP^Sc detection using immunohistochemistry with the polyclonal antibody Rb486 according to standard methodology as previously described (34).

Discriminatory sPMCA.

Following completion of this ring trial and bioassay, a potentially discriminatory sPMCA method became available. This method, which is described in detail elsewhere (32), uses five sPMCA rounds with AHQ/AHQ and VRQ/VRQ sheep brain homogenates as the substrates being used in alternate rounds, followed by PK digestion and visualization in the WB using the monoclonal antibody (MAb) SHa31. This method selects for the amplification of BSE but not scrapie.

Amplified products that are detected in the WB are then additionally probed separately by P4 and SHa31 antibodies in order to confirm BSE status.

A panel of original aliquots from this comparative study was supplied and blinded for testing (panel 1). Then, following initial results, panel 2, which was generated from the original sources and extended the dilution range of the ovine BSE to 1:10,000, was also tested blind.

RESULTS

Estimated PrP^res concentrations in the brains of the sources that contributed to the mixtures.

The PrP^res concentrations in the brains of the sources that contributed to the mixtures are shown in Table 1. There was less than 1-log difference between the concentration of PrP^res in the ovine BSE source and each of the scrapie sources. Assuming that PrP^res is a reliable indicator for infectivity, the ratios of BSE relative to scrapie in the series of mixtures used in this study is accurate within 0.27 to 0.63 logs depending on the classical scrapie source. With the exception of the VRQ/VRQ classical scrapie source, the PrP^res concentration in the BSE source was lower than the other scrapie sources. Therefore, with the exception mentioned above, in the mixtures of scrapie sources with BSE the concentration of PrP^res attributed to scrapie was higher than that indicated by the percentage ratio of the scrapie source in the mixture.

Discriminatory Western blotting.

The WB results for each sample were recorded using the following criteria: high or low molecular migration with the core antibody (SHa31); strong, medium, weak, or negative with the N-terminal antibody (P4); and a description of either classic 3 band or atypical profile for each antibody. Using the combination of results, each sample was assigned a concluding result of BSE, scrapie, atypical scrapie, or a description of combined TSE types. Representative descriptions for 14 blinded samples are presented in Table 2, with corresponding Western blots shown in Fig. 1.

TABLE 2.

Representative descriptions for 14 blinded samples^a

WB lane	MAb SHa31			MAb P4		Blind interpretation	Sample details
	Detection affinity	Banding pattern	Molecular massb	Detection affinityc	Banding pattern
1	++	Atypical	NAd	+	Atypical	Atypical scrapie	100% Atypical scrapie
2	+++	Classicale	High	+++	Classical	Classical scrapie	100% ARQ 147 scrapie
3	+++	Classical	Low	+	NAf	BSE	100% ovBSEg
4	+++	Classical	High	+++	Classical	Classical scrapie	100% VRQ scrapie
5	+++	Classical	High	+++	Classical	Classical scrapie	100% ARQ scrapie
6	+++	Classical	Low	−	NA	BSE	99% ovBSE, 1% atypical scrapie
7	+++	Classical	Low	+++	Classical	Scrapie with BSE	50% ovBSE, 50% ARQ147 scrapie
8	++	Classical	Low	++	Atypical	BSE with atypical	10% ovBSE, 90% atypical scrapie
9	+++	Classical	High	+++	Classical	Scrapie	10%ovBSE, 90% VRQ scrapie
10	+++	Classical	Low	+	Classical	BSE with low dilution of scrapie or CH1641	99% ovBSE, 1%ARQ147 scrapie
11	+++	Classical	Low	+	Classical	BSE with low dilution of scrapie or CH1641	75%ovBSE, 25% atypical scrapie
12	+++	Classical	High	+++	Classical	Scrapie	25%ovBSE, 75%VRQ scrapie
13	+++	Classical	Low	++	Classical	BSE with scrapie	90% ovBSE, 10% ARQ scrapie
14	++	Atypical	NA	++	Atypical	Atypical	1% ovBSE, 99% atypical scrapie

^aIllustrated in Fig. 1.

^bMolecular mass of the unglycosylated band of a classical 3-band pattern.

^c−, absent; +, weak; ++, moderate; +++, strong.

^dNA, not available.

^eThree-band pattern associated with classical scrapie and BSE.

^fSample too weak to determine the banding pattern.

^govBSE, ovine BSE.

Western blotting consistently detected BSE within a mixture with scrapie when it was present as 50% of the mixture (Table 3).

TABLE 3.

Ability of biochemical tests to identify BSE when mixed with different scrapie sources (panel 1)

^aRelative to the source. Shaded cells indicate the limit below which WB and ELISA failed to identify the BSE component in the mixture.

^b+, BSE detected; −, BSE not detected; ±, inconclusive result.

^cOne-hundred percent BSE. The percentages in the other cells in the same column indicate the percentages of BSE in the mixtures. The difference reflects the percentage of scrapie material in the mixture.

^dMixture selected for bioassay in transgenic mice.

Discriminatory ELISA.

The ELISA method consistently detected BSE when mixed with scrapie when it was present as 99% of the mixture (Table 3). However, the results varied depending on the scrapie strains. BSE was detected when it was present as 75% of the mixture (for classical scrapie VRQ/VRQ and ARQ/ARQ), 90% (classical scrapie 1-4-7 ARQ/ARQ), or 99% (atypical scrapie). When the data are unblinded and the ELISA results are grouped by scrapie type (Fig. 2), it can be seen that this apparent inability to detect BSE is partly due to the restrictions of having a numerical result and cutoffs. There is no qualitative data to aid interpretation of intermediate cases.

FIG 2.

Discriminatory ELISA. The mixed samples were tested in duplicate in blinded conditions. Ovine BSE was mixed with atypical scrapie (A), classical scrapie (VRQ/VRQ) (B), classical scrapie (1-4-7 ARQ/ARQ) (C), or classical scrapie (ARQ/ARQ) (D). The normalized ratio for classical scrapie samples, which are highly PK resistant, is less than 0.3, intermediate scrapie samples present a normalized ratio comprised between 0.3 and 0.7, and experimental ovine BSE samples have a normalized ratio between 0.7 and 1.3. Atypical scrapie samples have a ratio greater than 1.3. According to these values, the blinded samples were categorized as scrapie (gray), intermediate scrapie (hatched), BSE (black), or atypical scrapie (white).

A summary of the ability of the biochemical tests to discriminate BSE in the presence of scrapie at a ratio of 1% to 99% (panel 1) when the samples are blinded is presented in Table 3. Different isolates resulted in different discriminatory thresholds, which were different between the tests. For example, BSE was detected at very low levels against a background of atypical scrapie in the WB, but the BSE signal was masked by small amounts of atypical scrapie when present as a mixture in the ELISA.

Bioassay. (i) Mixtures of BSE with VRQ/VRQ or ARQ/ARQ classical scrapie.

The two tests, WB and ELISA, failed to identify the presence of BSE in the dilution series when its ratio in the mixture with VRQ/VRQ classical scrapie dropped below 50% (Table 3). Therefore, the inoculum just below the cutoff point (25% BSE ratio relative to scrapie) and the inoculum with the lowest BSE ratio relative to scrapie (1% BSE) were subjected to bioassays. Western blotting and ELISA also failed to identify the presence of BSE when the ratio of BSE to ARQ/ARQ classical scrapie was below 25% (Table 3). Therefore, the inocula with 10% and 1% BSE concentrations relative to scrapie were subjected to bioassays.

In tg338 mice, the scrapie agents isolated from the two classical scrapie sources were indistinguishable, with very short incubation periods of 69.5 and 75.5 mean days postinoculation (dpi) for VRQ/VRQ and ARQ/ARQ scrapie isolates, respectively (Fig. 3A and 4A), and similar lesion profiles (Fig. 3B and 4B) and brain distribution of PrP^Sc types as assessed by IHC (data not shown).

FIG 3.

Bioassay data in tg338 and tg110 mice of 25% and 1% BSE mixtures with VRQ/VRQ classical scrapie. The original BSE and VRQ/VRQ classical scrapie sources that were used to produce the mixtures were also bioassayed. (A) Incubation periods; (B) lesion profiles in tg338 mice; (C) lesion profiles in tg110 mice. At least 5 clinically and histopathologically positive mice contributed to each lesion profile (solid lines) unless indicated (dashed line).

FIG 4.

Bioassay data in tg338 and tg110 mice of 10% and 1% BSE mixtures with ARQ/ARQ classical scrapie. The original BSE and ARQ/ARQ classical scrapie sources that were used to produce the mixtures were also bioassayed. (A) Incubation periods; (B) lesion profiles in tg338 mice; (C) lesion profiles in tg110 mice. At least 5 clinically and histopathologically positive mice contributed to each lesion profile (solid lines) unless indicated (dashed line).

All tg338 mice that were challenged with BSE succumbed to TSE 624 dpi or later (Fig. 3A). In contrast, inoculation of tg338 mice with BSE mixed with either the VRQ/VRQ or the ARQ/ARQ classical scrapie sources produced incubation periods of <90 dpi that were compatible with those produced by the respective scrapie sources alone (Fig. 3A and 4A). These data indicate that the component isolated in the mice from the BSE mixtures with VRQ/VRQ or ARQ/ARQ classical scrapie only had classical scrapie properties. The vacuolation lesion profiles alone were not conclusive because the VRQ/VRQ and the ARQ/ARQ classical scrapie and BSE profiles were not dissimilar enough to allow unequivocal interpretation, although the mixtures did align more closely with the classical scrapie profiles produced by the 100% scrapie sources (Fig. 3B and 4B).

All tg110 mice challenged with the VRQ/VRQ source were TSE negative or, in the case of the ARQ/ARQ source, showed low attack rates with the first positive animal identified 581 dpi (Fig. 3A and 4A). In contrast, mixtures of BSE with either VRQ/VRQ or ARQ/ARQ classical scrapie produced clinical-stage TSE with incubation periods of 236 to 326 dpi (Fig. 3A and 4A). These incubation periods are comparable with those generated by the original BSE source (221 to 267 dpi) albeit slightly longer, probably as a result of the slightly reduced titer of BSE in the mixtures. These data suggest that in this mouse line only the BSE component was isolated from the mixtures. The lesion profiles from the tg110 mice (Fig. 3C and 4C) further support the conclusion that the isolated agent had only BSE properties although it was not possible to construct lesion profiles from either scrapie source due to the lack of sufficient clinically positive mice diagnosed with TSE.

(ii) Classical scrapie 1-4-7 ARQ/ARQ and ovine BSE.

The two tests failed to identify the presence of BSE in the dilution series when its concentration in the mixture dropped below 50% (Table 3). Therefore, the inoculum just below the cutoff point (25% BSE ratio relative to scrapie) and the inoculum with a 1% BSE ratio relative to scrapie were subjected to bioassay.

Figure 4A shows that tg338 mice challenged with either of these mixtures succumbed to disease with incubation periods that were compatible with the incubation periods produced by the scrapie source alone, indicating that the agents isolated from the 25% and 1% BSE mixtures were scrapie components. The lesion profiles are not conclusive because the ARQ/ARQ scrapie and BSE profiles are indistinguishable (Fig. 5B).

FIG 5.

Bioassay data in tg338 and tg110 mice of 25% and 1% BSE mixtures with 1-4-7 ARQ/ARQ classical scrapie. The original BSE and 1-4-7 ARQ/ARQ classical scrapie sources that were used to produce the mixtures were also bioassayed. (A) Incubation periods; (B) lesion profiles in tg338 mice; (C) lesion profiles in tg110 mice. At least 5 clinically and histopathologically positive mice contributed to each lesion profile (solid lines) unless indicated (dashed line).

The incubation periods of tg110 mice challenged with the 1-4-7 classical scrapie isolate were relatively shorter than the incubation periods caused by BSE in this mouse line (Fig. 5A). The incubation periods produced by the 25% and 1% BSE mixture in this mouse line were aligned with the incubation periods produced by the 1-4-7 ARQ/ARQ classical scrapie source (Fig. 5A). Lesion profiles concur with this interpretation, as the lesion profiles produced by the mixtures align with the lesion profile of the 1-4-7 classical scrapie while BSE produces a distinct separate profile (Fig. 5C).

Tg110 mice challenged with either the scrapie or the BSE source succumbed to disease with relatively short incubation periods; therefore, the mice inoculated with the mixtures were further compared to those challenged with the original sources using immunohistochemistry (IHC) (Fig. 6).

FIG 6.

Immunohistochemistry of tg110 mice challenged with 1-4-7 ARQ/ARQ classical scrapie (A), ovine BSE (B), mixture with a relative BSE to scrapie ratio of 1:3 (25% BSE) (C), and a mixture with relative a BSE to scrapie ratio of 1:99 (1% BSE) (D). All photos show rostral medulla at the same magnification. Black rectangles in (A), (B), and (D) indicate areas that have been further magnified and presented as insets in the corresponding images. (D) The mice inoculated with the BSE-scrapie mixtures show BSE associated coalescing PrP^Sc patterns even at the lowest ratio of BSE-scrapie. Red arrowheads, examples of neurons with intraneuronal PrP^Sc deposits; blue arrowheads, patterns of coalescing PrPSc deposits; black arrowheads, examples of neurons devoid of intraneuronal PrPSc deposits.

Tg110 mice challenged with the 1-4-7 classical scrapie source showed a pattern characterized by intraneuronal and fine punctate PrP^Sc deposits in the neuropil; when aggregates were present, they were distinct, well demarcated, and ovoid (Fig. 6A). BSE challenged tg110 mice also showed intraneuronal PrP^Sc, but the neuropil was populated with diffuse granular deposits, coalescing aggregates, and plaque-like formations (Fig. 6B). In addition to the IHC attributes associated with the 1-4-7 classical scrapie pattern, tg110 mice challenged with either mixture additionally showed features that were associated with the BSE-induced pattern (Fig. 5C and B). This BSE-associated pattern also appeared to be more extensive in the mice that were challenged with the 25% BSE mixture than that in the mice that received the 1% BSE mixture.

Although the incubation periods and the lesion profiles, particularly those generated by mice that were inoculated with 25% BSE mixture, suggest that the BSE agent did not propagate selectively in the tg110 mice, it was still possible to identify the BSE component reliably in the Tg110 mice using IHC.

(iii) Atypical scrapie and ovine BSE.

The WB and ELISA failed to identify the presence of BSE in the dilution series when its concentration in the mixture dropped below 10% (Table 3). Therefore, the inoculum with 1% BSE concentration relative to scrapie was subjected to bioassays.

The incubation period data in tg338 indicate that the agent isolated from these mixtures was compatible with the agent isolated from the atypical scrapie source; in tg110 mice, the incubation period data indicate that agent isolated from the mixtures was compatible with the agent isolated from the BSE source (Fig. 7A). The lesion profiles from tg338 (Fig. 7B) and tg110 mice (Fig. 7C) provide further support to the incubation period data although it was not possible to construct lesion profiles from the scrapie source in tg110, as all of the mice challenged with this source were TSE negative.

FIG 7.

Bioassay data in tg338 and tg110 mice of a 1% BSE mixture with atypical scrapie. The original BSE and atypical scrapie sources that were used to produce the mixtures were also bioassayed. (A) Incubation periods; (B) lesion profiles in tg338 mice; (C) lesion profiles in tg110 mice. At least 5 clinically and histopathologically positive mice contributed to each lesion profile.

Discriminatory sPMCA.

The original panel of 33 samples (panel 1) was correctly reported as BSE present, with the correct exception of the unmixed scrapie samples (Table 3). The analysis of panel 2 gave identical results for the 1% to 99% mixtures. sPMCA also correctly reported the presence of BSE in all samples in which BSE was diluted to 0.1%. BSE was also successfully detected in one sample where it was diluted to 0.01% with atypical scrapie (Fig. 8).

FIG 8.

Western blots of PK digested sPMCA products. Samples of BSE brain homogenate were diluted into 4 isolates of ovine scrapie-positive brain homogenate of VRQ, ARQ, atypical, and 1-4-7 ARQ types, as indicated, at 1/10, 1/100, 1/1,000, or 1/10,000 dilutions (labeled as 1 to 4, respectively). Samples were amplified in duplicate including an equal number of scrapie only samples. Positive sPMCA samples were then further analyzed: single sPMCA sample replicates were redigested with PK and immunoblotted using the antibodies SHa31 and P4 (as shown). Blotting controls were positive sheep scrapie (+S) or BSE (+B) brain homogenates and an ovine BSE sPMCA-positive sample (+BP). Molecular mass markers (M) at 20 and 30 kDa are shown. Blots were probed with either SHa31 or P4 monoclonal antibodies. Samples were scored positive for BSE (as indicated) if over a signal threshold on the WB and using a SHa31/P4 WB ratio as previously described (32).

DISCUSSION

Coinfection studies in animals using mixtures of known infectious titer remain the hypothetical ideal for this type of study, but for several reasons such studies are not necessarily as appropriate as they might appear on initial consideration. The observed titer of a TSE isolate is not an absolute measure of the infectivity of that isolate but is also affected by the susceptibility of the host, which may differ for different isolates. For example, some scrapie strains that readily infect sheep and transgenic mouse models do not cause disease in conventional inbred mouse lines (2, 33). Therefore, it is unwise to assume that two isolates with similar observed levels of infectivity in any one model will necessarily have the same infectivity potential in other species or indeed in animals of different genotypes. Equally, PrP^res concentration cannot be considered to be a consistent proxy for the level of infectivity in an isolate (35).

The interaction of strains either in vitro or within a single host is also very poorly understood. If strain properties are conferred by tertiary molecular structure, then mixing isolates together might affect the ability of a strain to infect a host either in an inhibitory or potentiating way. This may also affect tests applied to a sample with the two isolates represented. However, the data from this study demonstrate that this is not the case, at least with the BSE and scrapie combinations used; all tests and models, except the sPMCA, correctly classified each of the strains contributing to each of the mixtures, including the successful isolation in mice of all the component isolates of each mixture, with retention of the biological phenotypes of the unmixed controls.

The bioassay data, particularly from the bovinized mice, also suggest that if sheep are exposed to BSE and scrapie, the two agents will most likely propagate as independent entities according the dynamics of titer, time of exposure to each agent, and ovine PrP genotype. Therefore, exposure to the two agents is unlikely to result in a novel agent with previously undetected biochemical or biological properties, although this possibility should always be considered when a new or unusual isolate is identified. Under these circumstances, exposure to the two agents would give rise to a mixture in which BSE can be detected with the current biochemical and biological tests, provided that the titer of scrapie is not overwhelming. The data also show that the choice of diagnostic test and which scrapie strain is present dictate the level of scrapie that “overwhelms” the detection of BSE. For BSE mixed with atypical scrapie, the presence of 10% scrapie masked BSE detection by ELISA whereas, at the other extreme, BSE mixed with atypical scrapie and detected by WB required the presence of above 90% scrapie to mask BSE.

Additional unknowns, if trying to use an in vivo challenge model to recreate possible coinfection, are the age of the animal at challenge, the order in which the challenges occur, and possibly the length of time between the challenges (36–38).

The main purpose of this study was to attempt the identification of the BSE component of the mixture by using the approved discriminatory tests (European Union regulation 36/2005), followed by discriminatory bioassays in tg338 mice to selectively propagate the scrapie strains and in tg110 mice to selectively propagate the BSE agent, in mixtures in which the identifiable biochemical signature of BSE has been lost or obscured. We conclude that the biochemical rapid tests can discriminate BSE in the presence of scrapie to various degrees depending on the scrapie source, but this variation does not appear to be attributable to the amount of PrP^res. The WB was more reliable than ELISA in discriminating BSE in the presence of scrapie, regardless of the scrapie isolate.

The bioassay was capable of resolving cases of coinfection even where BSE represented just 1% of the total TSE infectivity. In order to achieve this, bioassay systems should include a bovinized mouse line that favors propagation of BSE over scrapie and an ovinized line with complementary properties, i.e., a line that preferentially facilitates the propagation of scrapie prions over the BSE agent. In addition to the ability to identify BSE in mixed infections, bioassay continues to be the only validated method available that enables the comprehensive phenotypic characterization of an unknown isolate.

It is probably not possible to source ideal mouse lines with the above properties, particularly as classical scrapie consists of various strains with widely variable properties. However, the selected mouse lines (tg338 and tg110) are as close as possible to that ideal situation and may be used in combination to resolve coinfection cases in a surveillance context if they arise. In addition, IHC may be useful to resolve a small proportion of bioassays in which the relative mouse line susceptibility is not, by itself, conclusive.

Within this study, the new discriminatory sPMCA approach was the only in vitro method which consistently detected BSE when it was present in these mixtures, even at very low concentration (down to 0.1%). This sensitivity can potentially be exploited to screen pooled ovine TSE brain samples for the presence of BSE, greatly increasing the throughput and decreasing the costs of such screening programs in the future. When the purpose of an investigation is solely to determine the presence or absence of BSE (as opposed to characterizing whatever is in the isolate), this assay would appear to offer the potential for a fast and cost-effective alternative to bioassay and will be proposed to the EURL strain typing expert group as a useful addition to the panel of tests currently used for the screening of unknown isolates.