########  Bovine Spongiform Encephalopathy   #########

Journal of Virological Methods 117 (2004) 2736
Atypical scrapie cases in Germany and France are identified
by discrepant reaction patterns in BSE rapid tests
A. Buschmanna, A.-G. Biacabe b, U. Ziegler a, A. Bencsik b, J.-Y. Madecb,
G. Erhardt c, G. Lühken c, T. Baron b, M.H. Groschup a,?
a Federal Research Centre for Virus Diseases of Animals, Institute for
Novel and Emerging Infectious Diseases,
Boddenblick 5a, 17493 Greifswald-Insel Riems, Germany
b AFSSA-Lyon Unité Virologie-ATNC, 31 Avenue Tony Garnier, 69364 Lyon
Cedex 07, France
c Department of Animal Breeding and Genetics, Justus-Liebig University
Giessen, Ludwigstr. 21B, 35390 Giessen, Germany
Received 27 August 2003; received in revised form 13 November 2003;
accepted 18 November 2003
Abstract
The intensified surveillance of scrapie in small ruminants in the
European Union (EU) has resulted in a substantial increase of the number
of diagnosed cases. Four rapid tests which have passed the EU evaluation
for BSE testing of cattle are also recommended currently and used
for the testing of small ruminants by the EU authorities. These tests
include an indirect ELISA (cELISA), a colorimetric sandwich ELISA
(sELISA I), a chemiluminescent sandwich ELISA (sELISA II), and a Western
blot (WB). To this point, the majority of samples have been
screened by using either sELISA I (predominantly in Germany) or WB
(predominantly in France). In this study, it is shown that a number
of the German and French scrapie cases show inconsistent results using
rapid and confirmatory test methods. Forty-eight German sheep,
209 French sheep and 19 French goat transmissible spongiform
encephalopathy (TSE) cases were tested. All cases were recognised by the
sELISA I and either one of the confirmatory methods (scrapie-associated
fibrils (SAF)-immunoblot or immunohistochemistry). Surprisingly,
three rapid tests failed to detect a significant number of scrapie cases
(29 in France and 24 in Germany). The possible reasons for these
inconsistent reaction patterns of scrapie cases are discussed. Similar
discrepancies have not been observed during rapid testing of cattle for
BSE, the disease for which all diagnostic methods applied have been
evaluated.
© 2003 Elsevier B.V. All rights reserved.
Keywords: Scrapie; Prion protein; Rapid test; Confirmatory method
1. Introduction
Scrapie in sheep and goats is the longest known transmissible
spongiform encephalopathy (TSE) and has first
been described in the 18th century (McGowan, 1922).
As no obvious clinical or epidemiological connection to
human disease has been revealed to date, scrapie is considered
non-pathogenic for humans, at least under natural
conditions. However, an epidemic of bovine spongiform
encephalopathy (BSE) in cattle emerged during the last two
decades of the 20th century. Following this extensive exposure,
a variant form of CreutzfeldtJakob disease in humans
was discovered in 1996, which is linked directly to BSE, the
bovine form of transmissible spongiform encephalopathy
? Corresponding author. Tel.: +49-38351-7163; fax: +49-38351-7191.
E-mail address: martin.groschup@rie.bfav.de (M.H. Groschup).
(Will et al., 1996; Collinge et al., 1996; Bruce et al., 1997).
Under experimental conditions, sheep are easily infected
orally by the BSE agent and suffer from clinical signs indistinguishable
from scrapie. In contrast to cattle, sheep
carry abundant amounts of infectivity throughout most body
tissues even early after infection (Foster et al., 1993, 2001).
Although scrapie is a notifiable disease, the actual number
of scrapie cases in the EU member states during the
last years remained unclear. To identify the true incidence
of this disease, EU regulations for an active surveillance
have been implemented in the year 2002 (Moynagh and
Schimmel, 1999). These regulations (EC regulation 999/
2001 and amendments) require large scale TSE testing of
small ruminants using EU approved rapid tests that have
been successfully evaluated for BSE testing in cattle. Sample
numbers that need to be tested in each member state were
set according to the numbers of slaughtered and fallen animals
in a member state and comprise up to 66,000 tests (for
0166-0934/$  see front matter © 2003 Elsevier B.V. All rights reserved.
doi:10.1016/j.jviromet.2003.11.017
28 A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736
Germany and France). As a result of the obligatory active
surveillance program, numbers of scrapie cases diagnosed
increased considerably throughout Europe. For example, in
Germany no or only single scrapie cases per year were reported
until the year 2001, which added up to 12 cases in total
(Junghans et al., 1998). In contrast, after implementation
of the monitoring scheme, 35 cases were observed during
2002 followed by 13 cases during the first 7 months of 2003.
Rapid tests are based on the detection of pathological
prion protein (designated PrPSc). In contrast to its cellular
counterpart, PrPSc is partially proteinase K (PK) resistant,
and due to its high hydrophobicity forms scrapie-associated
fibrils (SAF) (Oesch et al., 1994; Lehmann and Harris,
1995). The four rapid tests used commonly well as the con-
firmatory methods that have been approved by the Office
International des Epizooties (OIE; SAF-immunoblot and
immunohistochemistry) apply polyclonal or monoclonal
antibodies to detect the proteinase K-treated PrPSc that is
accumulated in the brains of TSE-affected animals.
As a natural BSE transmission to sheep cannot be excluded
and would suggest a high risk of exposure for humans,
the scientific and public attention towards scrapie and
its discrimination from BSE has increased. Strain typing
in mice is used to differentiate BSE and scrapie infection
in sheep. It compares the incubation times and scores the
pathological lesions that the transmitted sheep brain samples
cause in inbred mouse lines (Fraser and Dickinson, 1968;
Bruce et al., 1996). However, this method is time-consuming
and the results may be ambiguous and therefore difficult to
interpret. Recently, alternative methods for the strain differentiation
have therefore been developed such as the interpretation
of the molecular weight and the glycosylation pattern
of the pathological and partially proteinase K-resistant form
of the prion protein (PrPSc) accumulated in the brain of the
infected host (Hill et al., 1998; Hope et al., 1999; Kascsak
et al., 1986; Kuczius et al., 1998; Sweeney et al., 2000; Baron
et al., 1999; Groschup et al., 2000; Zanusso et al., 2003).
Amino acid variations at positions 136 (valine (V) and
alanine (A)), 154 (arginine (R) and histidine (H)) and 171
(glutamine (Q), arginine (R) and histidine (H)) of the ovine
prion protein have been shown to influence the susceptibility
of sheep to natural and experimental scrapie infections
(Laplanche et al., 1993; Goldmann et al., 1994; Belt et al.,
1995; ODoherty et al., 2002). In general, the ovine alleles
PrPVRQ and PrPARQ seem to be related to a higher susceptibility
while the alleles PrPARR and PrPAHQ coincide
with a low susceptibility to a scrapie infection. Therefore,
the PrP alleles of the sheep TSE cases described here were
also taken into consideration.
To date, no independent comparative study has been performed
to evaluate the rapid tests performances with sheep
or goat scrapie samples. We report that a substantial number
of scrapie cases which have been diagnosed using the
approved confirmatory methods, e.g. SAF-immunoblotting
and immunohistochemistry (OIE manual) could not be detected
with three of the four rapid tests used currently.
2. Materials and methods
2.1. Sampling of sheep brains
Brainstem tissue pieces (up to 10 g) were collected at abattoirs
and rendering plants by sampling through the foramen
magnum. Depending on their origin, sample quality varied
from fresh to autolytic.
2.2. BSE rapid tests
BSE rapid tests were performed following the manufacturers
instructions. As each rapid test requires a specific homogenisation
procedure and buffer developed solely for that
test, homogenates that were prepared for one test method
could not be exchanged with other tests. Moreover, due to
lack of brainstem material, it was impossible in many cases
to prepare a macerate of a larger amount of brainstem material
for the preparation of comparable homogenates. In most
cases, the initial rapid testing was repeated at the national
reference laboratory using the original homogenate. For
further testing using the same or another rapid test method,
new homogenates from another piece of brain needed to be
prepared.
A colorimetric sandwich ELISA (sELISA I, Platelia, Biorad,
Munich) was used for primary screening of most of the
samples. Approximately 0.35 g brainstem tissues from the
obex region were homogenised and digested with PK. After
precipitation, samples were resolubilised and diluted before
pipetting them on a microtitre plate that was coated with
a PrP-specific monoclonal antibody. After intensive washing
of the plate and subsequent incubation with the conjugate,
the colorimetric signal was developed and the reaction
stopped prior to measurement of the absorbance at 450 and
620 nm. The cut-off value to determine positive results was
calculated by adding a fixed value of 0.21 to the mean optical
density of the negative controls.
For the indirect ELISA (cELISA, Enfer TSE kit, version
2.0, Enfer Scientific, Dublin), 0.51 g brainstem material
was homogenised, a small amount of each sample was centrifuged
to pellet residual detritus before samples were adsorbed
on a microtitre plate and digested with PK. After intensive
washing, samples were incubated with a PrP-specific
polyclonal serum, washed again and incubated with conjugate,
washed again before the chemiluminescence substrate
was added. Emitted light units (LU) were measured and all
samples with LU above 5.5 were considered as reactive.
For the rapid Western blot (rapid WB, Prionics Check
Testkit, Prionics, Zurich), 10% (w/v) homogenates were
prepared from 0.5 g brainstem tissue and digested with
PK for 40 min. The reaction was stopped by adding a
protease inhibitor and samples were heat denatured. This
was followed by sodium dodecylsulphate-polyacrylamide
(SDS-PAGE), transfer to a PVDF membrane and incubation
with a PrP-specific monoclonal antibody. After
incubation with the conjugate, signals were visualised by
A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736 29
chemiluminescence and detected using either film or a CCD
camera system. Samples that gave signals of the molecular
weight 1529 kDa and showed the typical triple banding
pattern, or at least the uppermost diglycosylated PrPSc band
at 29 kDa were considered reactive.
For the luminescence sandwich ELISA (sELISA II, LIA,
Prionics), 10% (w/v) homogenates of 0.5 g brainstem tissue
were digested with PK, incubated first with assay buffer
and then with pre-incubation buffer before the PrP-specific
monoclonal antibody was added to the samples. The samples
were transferred to the detection microtitre plate that
was coated with another PrP-specific monoclonal antibody
and incubated at room temperature to allow the binding of
PrP-detection antibody complexes to the microtitre plate.
After intensive washing, a chemiluminescence substrate solution
was added and light units were measured. The mean
of eight negative control values multiplied by 10 was taken
as negative control cut-off (NCC). The mean of all samples
whose values lie under the NCC were multiplied by 10 to
give the real sample cut-off (RSC). All samples with values
higher than the RSC were considered as reactive.
2.3. Exchange of single rapid WB components
All relevant test components were exchanged individually
to examine their influence on the test results. To
exchange the homogenisation buffer without having to prepare
a fresh homogenate from the limited sample amount,
5% (w/v) brain homogenates were made by diluting the
10% (w/v) homogenates for the rapid test WB in a 1:2
ratio with either the same homogenisation buffer or with a
0.64M sucrose buffer containing 1% (w/v) desoxycholate
and 1% (w/v) Nonidet-P 40 (NP40; ICN, USA). In the
following steps, the proteinase K solution supplied with
the testkit was exchanged with 50 
g/ml PK from another
supplier (Roche, Mannheim), or the primary antibody L42
was applied instead of the antibody included in the testkit,
or the secondary antibody GAM-AP, Dianova was used to
replace the conjugate from the testkit. We then replaced the
solution as well as the primary and secondary antibodies
at the same time (modified rapid Western blot). Finally,
all relevant test components (homogenisation buffer, PK,
primary and secondary antibodies) were replaced by the
reagents used in the in-house BFAV rapid Western blot.
2.4. Proteinase K digestion of crude 10% brain
homogenates (BFAV rapid Western blot)
Some selected samples were also tested using a simple
protease degradation protocol. Homogenates (10%, w/v)
were prepared in a 0.32M sucrose buffer containing 0.5%
(w/v) deoxycholic acid sodium salt (DOC; Serva, Heidelberg)
and 0.5% (w/v) NP40. These homogenates were
incubated with 50 
g/ml PK (Roche, Mannheim) for 1 h
at 37 ?C. Sample buffer containing mercaptoethanol was
added and the reaction was stopped by adding 10mM
PMSF and heating to 95 ?C for 5 min. These samples were
analysed in a SDS-PAGE as described further, using mab
L42 that specifically binds to aa 144166 of ruminant PrP
(Harmeyer et al., 1998).
2.5. Confirmatory testing
2.5.1. SAF-immunoblot
For preparation of the scrapie-associated fibrils, a 10%
(w/v) homogenate in a 0.01M sodium phosphate buffer
(pH 7.4) containing 10% (w/v) sarcosine, 0.5mM phenylmethylsulphonylfluoride
and 0.5mM N-ethylmaleimide
was prepared with brainstem material from the obex region.
The amount of material used depended on the initial rapid
test result and on the amount of brainstem available (i.e. 2 g
sample material or less was used if sELISA I absorbance
was >1). After a first centrifugation for 30 min at 20,000×g
to pellet residual detritus, the supernatant was transferred
into a new centrifuge tube and centrifuged for 2 h and
15 min at 220,000 × g. Pellets were resuspended in 3ml of
0.015M Tris (pH 7.4), incubated for 15 min at 37 ?C, and
the double volume of 15% potassium iodide-high salt buffer
containing 60mM sodium pentahydrate, 10mM TrisHCl,
and 1% N-lauroylsarcosine was added. Samples were incubated
at 37 ?C for another 30 min and then split into equal
parts before 45 
g PK was added to one of the aliquots
and incubated for 1 h at 37 ?C. Afterwards, 4.5 ml of 10%
potassium iodide-high salt buffer containing 60mM sodium
pentahydrate, 10mM TrisHCl, and 1% N-lauroylsarcosine
was added to the digested and non-digested aliquots. Finally,
samples were centrifuged through a gradient of
20% sucrose in 10% potassium iodide-high salt buffer
for 1 h at 280,000 × g and the pellets were resuspended
in a sample buffer (pH 6.8) containing 0.1 g/ml sodium
dodecylsulphate, 25mM TrisHCl (pH 7.4), 0.5% mercaptoethanol
and 0.001% bromphenol blue, heat denatured for
5 min at 95 ?C and loaded on SDS-PAGE gels containing
13% bis-acrylamide. After electrophoresis, proteins were
transferred on a PVDF membrane in a semi-dry chamber.
Membranes were then blocked in I-Block (Tropix, Bedford,
USA) for 30 min and incubated with the PrP-specific
monoclonal detection antibody (mab) P4 (Harmeyer et al.,
1998) for 1 h and 30 min at room temperature. In cases with
only a weak positive rapid test result, mab 6H4 was added
to mab P4 as a second detection antibody. Membranes were
washed three times for 10 min with phosphate buffered
saline (PBS) containing 0.1% Tween 20 and then incubated
with a secondary antibody bound to alkaline phosphatase
(goat anti-mouse AP, Dianova) for 1 h at room temperature.
After washing, the chemiluminescence substrate CDP-Star
(Tropix) was applied and incubated on the membrane for
5 min before the light signals were detected in a camera.
2.5.2. Modified SAF-immunoblot
Immunoblot analysis was carried out as described previously
(Madec et al., 2000). Briefly, a 10% (w/v) homogenate
30 A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736
in 5% (w/v) glucose was prepared from 0.35 g of tissue.
Proteinase K was added to a concentration of 10 
g/ml and
samples were incubated at 37 ?C for 1 h. N-Laurylsarcosyl
was added to a final concentration of 10% (w/v). The samples
were then centrifuged at 465,000 × g for 2 h over a
10% (w/v) sucrose cushion. The pellet was resuspended in
sample buffer (4% (w/v) SDS, 2% (w/v) -mercaptoethanol,
192mM glycine, 25mM Tris, 5% (w/v) sucrose) and subjected
to immunoblot analysis as described earlier, using
mab SAF 84 as a detection antibody.
2.5.3. Immunohistochemical PrPSc detection
Samples were processed as described previously (Hardt
et al., 2000). Briefly, 3mm sections of the obex region were
fixed in 3.5% natrium buffered formalin (NBF) for at least
48 h. After a 1 h incubation in 98% formic acid, samples
were dehydrated automatically using pressure and vacuum
at 35 ?C through a series of ethanol solutions and embedded
in paraffin blocks. Sections (3 
m) were then prepared and
immunohistochemical staining using the PrP-specific monoclonal
antibody L42 was carried out in an automated stainer.
This procedure included a pre-treatment for 15 min in 98%
formic acid followed by an incubation for 5 min in tap water
and for 30 min in NBF. The sections were then washed twice
in PBS for 5 min before entering them into the autostainer.
The automated staining protocol included a protease treatment
for 12 min at 42 ?C. Signals were visualised using the
Fast Red detection system. For weak positive samples, the
same procedure was also applied using mab SAF 70, SAF 84
(Demart et al., 1999), and mab F89 (ORourke et al., 1998).
French samples were analysed by immunohistochemistry
as described previously (Debeer et al., 2001, 2002) using
SAF 84 monoclonal antibody.
2.6. Determination of PrP genotypes
PrP genotypes of the diseased sheep were determined
by sequencing as described earlier (Junghans et al., 1998)
and/or by PCR-RFLP. Briefly, genomic DNA was extracted
from brain samples by using a commercial kit (QiaAmp
DNA kit) followed by PCR amplification of the open reading
frame of the PrP gene. The PCR fragments were directly
used in sequencing reactions or restriction enzyme digestions
for determination of the DNA codons at positions 136,
154, and 171 of the ovine PrP.
2.7. Statistics
Fishers exact test of association was used for the
non-parametric assessment of association between genotype
and rapid WB results.
3. Results
Similar to other observations after the introduction of
mass screening for BSE, rapid testing of small ruminant
samples led to a substantial increase in the number of reported
scrapie cases in sheep and goats from Germany and
France. Since 2002, 48 cases of scrapie have been found in
Germany and 228 in France by active surveillance (as of 31
July 2003). Thirty-eight of the German scrapie cases were
initially detected by using the sELISA I, 9 by using the
rapid WB, and 1 by immunohistochemistry, while 133 of the
French cases (117 in sheep and 16 in goats) were initially
detected by using the rapid WB and 95 (92 in sheep and 3
in goats) by using the sELISA I testing. All cases were con-
firmed by either one of the two standard diagnostic methods
recommended by the OIE which are SAF-immunoblotting
including a proteinase K treatment and ultracentrifugation,
or immunohistochemistry. Both methods gave no
PrPSc-specific false positive signals when brain material
from a negative sheep was examined. SAF-immunoblotting
of the sheep samples gave clear and unequivocal reactions
with the typical PrPSc banding pattern (Fig. 1A). Some of
Fig. 1. OIE immunoblot, rapid Western blot, and immunohistochemistry
for selected sheep samples. Note that samples A8G and A21G are not
detected by rapid Western blot; sample A21G is also negative in
immunohistochemistry.
(A) Preparation of scrapie-associated fibrils following the
OIE-approved protocol (for details see Section 2). Negative control sample
and positive sample T22G were tested both with and without PK
digestion, for other samples, only PK digestion could be performed due
to low amount of available brain material. Mab P4 was applied as detection
antibody. Amount of prepared brain material is given for each sample,
three-fourths of each probe was loaded per lane. (B) Rapid Western
blot of sheep samples as under (A). Test was performed following the
manufacturers instructions. Mab 6H4 was used as detection antibody. (C)
Immunohistochemistry for sheep samples as under (A) using mab L42
which specifically detects ruminant PrP. Signal detection using Fast Red
system.
A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736 31
Fig. 1. (Continued ).
the samples also displayed an additional band of a lower
molecular mass around 12 kDa. When applied during diagnostic
routine, immunohistochemistry revealed intra- and
extraneuronal PrPSc deposition at the level of the obex and
(where additional samples were available) at other sites
with varying signal intensities (Fig. 1C and Table 1).
We then retested the confirmed scrapie cases using the
other available rapid tests. All eight German scrapie cases
that had been initially detected by the rapid WB were also
clearly recognised using the sELISA I rapid test. In contrast,
24 out of the 38 scrapie cases that had been detected by
sELISA I with results varying between absorbance values
of 0.5 and 2.3 were not detected by the rapid WB (Fig. 1B).
These samples were therefore categorised as atypical scrapie
cases. These samples were also negative when the sELISA
II was used. Due to a lack of sample material, not all of
these cases could be retested using the cELISA, but when
11 out of the 24 samples were analysed, all but 1 gave a
negative result. In contrast, one out of the eight samples
that had been initially recognised in the rapid WB and that
could be retested using other methods, was also positive
using the cELISA test (Table 1). As the cELISA has only
been approved for Germany recently, no samples that had
first been detected by using this method were available for
a comparison with other rapid tests.
Atypical scrapie cases were also found in a number of
the French samples (Table 2). From the 92 cases that had
first been detected using the sELISA I test with absorbance
values of up to 3.5 and that were confirmed by immunohistochemistry,
29 (28 sheep and 1 goat) samples were not
recognised in the rapid WB. In contrast, samples that were
initially detected using the rapid WB with signal intensities
varying between weak and strong could be confirmed using
the sELISA I and the other rapid tests whenever sufficient
sample amounts were available.
To follow-up the discrepancies between the various tests,
PrPSc deposition patterns of all German cases with incongruent
results were reassessed by immunohistochemical
examination. Depending on the quality and freshness of
the samples that had been sent to the national reference
centre for confirmation, histological preparations could
not in all cases be taken from the obex region. Although
PrPSc staining of most scrapie cases was easily detectable,
atypical cases also included three samples that were hardly
recognised, and another six samples for which PrPSc deposition
could not been detected at all. These results remained
32 A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736
Table 1
Selected German scrapie cases displaying two different immunochemical
PrPSc recognition patterns: while typical samples are confirmed positive
with
all methods (such as T3G, T5G, T6G, T10G, T11G, T12G, and T22G),
atypical cases remain negative using the rapid WB, cELISA, and sELISA II
test
(such as A4G, A16G, and A18G)
Sample PrP
genotype
sELISA Ia
(cut-off range
OD 0.220.29)
sELISA II
(cut-off range
OD 6459)
Rapid
WB
cELISA
(cut-off OD 5.5)
IHC SAFimmunoblot
Material used
for preparation
(g)
T2G ARQ/ARQ 3.3 295789/126259 ++ n.d. +++ ++ 1.5
T5G ARQ/ARQ 3.2 131465/156239 ++b 4720.3/5907.0 ++ ++ 1.5
T6G ARQ/ARQ 3.3 455848/866856 ++b n.d. +++ ++ 1.5
T10G ARQ/ARQ 3.2/2.9 234241/280448 ++ 1745.0/1685.3 ++ ++ 1.5
T11G ARQ/ARQ 3.6 488483/470383 ++b 1967.6/1901.1 + ++ 1.0
T12G ARQ/ARQ 3.5 187268/210202 ++b n.d. ++ ++ 1.0
T22G ARQ/ARQ 3.3/3.3 n.d. ++ n.d. ++ +++ 1.0
A2G AHQ/AHQ 1.2 68/99 b n.d. + + 1.5
A3G ARQ/ARQ 1.0 213/205  148.9/139.9 ++ ++ 2.0
A4G ARR/AHQ 1.5/1.4 108/103  1.0/0.9 - ++ 1.5
A5G ARQ/ARQ 0.9 268/285 b n.d. - ++ 2.0
A6G ARQ/ARQ 1.2/0.15 125/132  3.5/6.5 + ++ 1.5
A7G ARQ/ARQ 1.2/0.3 145/146  n.d. + ++ 1.5
A8G ARQ/ARQ 0.8 116/132  n.d. +++ ++ 2.0
A9G AHQ/AHQ 1.0/0.6 104/110  n.d. + + 1.5
A10G ARQ/ARQ 0.5 195/186 b n.d. ++ ++ 2.0
A11G AHQ/ARQ 1.4 131/132  1.1/1.5 - ++ 1.5
A12G ARQ/ARQ 2.2 111/98  1.2/1.2 + + 1.0
A13G AHQ/ARR 1.4 121/128  0.9/1.2 + + 1.5
A14G AHQ/ARQ 0.5 109/104 b n.d. - + 1.5
A15G AHQ/AHQ 0.4/0.3 90/90  0.9/1.2 - + 2.0
A16G ARQ/ARQ 2.3 110/118  0.6/0.7 + n.d. 
A17G AHQ/AHQ 0.3/1.9 133/125  n.d. +/- + 2.0
A18G AHQ/ARQ 0.5 91/92  0.7/0.7 ++ + 2.0
A21G AHQ/ARQ 0.8/0.4 n.d.  n.d. - + 1.5
n.d.: not determined.
a Repetitive sELISA I readings obtained in the NRL are italicised.
b Less than 0.5 g brain sample was used to prepare a 10% homogenate as
the availability of these brain samples was limited.
Table 2
Reactivity patterns of selected French scrapie cases in sheep and one goat
in sELISA I, rapid WB and immunohistochemistry
Sample PrP
genotype
IHC
(obex)
sELISA I
(cut-off OD 0.23)
Rapid
WB
S16F ARQ/AHQ ++ 2.907 
S1F ARR/ARQ ++ 3.469 
S2F ARR/ARQ ++ 3.041 
S3F ARR/ARQ ++ 0.705 
S4F n.d. ++ >3.500 
S5F ARR/ARQ ++ 0.925 
S6F ARQ/ARQ ++ >3.500 
S7F ARR/ARQ ++ 3.081 
S8F n.d. ++ 0.474 
S9F (goat)  ++ 3.409 
S10F ARQ/ARQ ++ 0.821 
S11F n.d. ++ 0.888 
S12F ARQ/ARQ ++ 1.276 
S13F ARQ/ARQ ++ 3.137 
S14F n.d. ++ 1.200 
S15F n.d. ++ 2.397 
S17F n.d. ++ 2.517 
S18F n.d. ++ 2.086 
S19F n.d. ++ 2.942 
unchanged when different detection antibodies, such as
SAF 70, SAF 84, and F89, were applied.
To examine further the failure to detect a significant number
of scrapie samples using the rapid WB, single reagents
or procedures within the protocol were exchanged. Supplementation
of the kit included homogenisation buffer with
0.32M sucrose containing 0.5% (w/v) DOC and 0.5% (w/v)
NP40 resulted in a minor increase of the signal strength of
the pathological PrP. A similar slight increase of the detectable
PrPSc signal was achieved when the homogenate
was digested using 50 
g/ml proteinase K instead of the PK
solution included in the kit with an undisclosed concentration.
The same minor effects occurred when the in-house
monoclonal antibody mab L42 (instead of mab 6H4 in the
kit) or a different secondary antibody were used. However,
when proteinase K and the first and secondary antibodies
were exchanged at the same time, most of the deviant samples
were recognised in the so modified WB assay (Fig. 2A
and B). When samples were homogenised in 0.32M sucrose
(containing 0.5% (w/v) DOC and 0.5% (w/v) NP40)
and proteinase K as well as the first and secondary antibodies
were exchanged (referred to as BFAV rapid Western
blot), the recognition of deviant samples was again improved
(Fig. 2C).
A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736 33
Fig. 2. Rapid Western blot of scrapie-affected sheep using different
preparation and detection protocols (for details see Section 2). (A)
Rapid WB kit
(kit-included sample homogenisation buffer, digestion with kit-included
proteinase K for 40 min at 48 ?C, signal detection using mab 6H4,
kit-included
goat anti-mouse conjugate). (B) modified rapid WB protocol where
proteinase K, first antibody mab 6H4, and conjugate have been exchanged with
1 mg/ml proteinase K, mab L42, and GAM-AP. (C) BFAV rapid Western blot
protocol where the sample homogenisation buffer, PK, blocking solution
as well as first and secondary antibodies have been exchanged with 0.32M
Sucrose containing 0.5% DOC and 0.5%, 1 mg/ml proteinase K, mab L42
as first antibody and GAM-AP as secondary antibody. Note that samples
A3G, A7G, and A10G are negative in the rapid Western blot protocol and
positive using the modified rapid WB protocol and the BFAV rapid Western
blot protocol.
To exclude any insufficient proteolytic degradation as a
reason for the inconsistent rapid test results and to explore
the PK sensitivity of atypical scrapie cases, different PK
concentrations were applied to the samples. In our hands,
2.5
g PK/ml were sufficient to digest reliably all PrPC from
a negative sheep control sample and therefore abolish any
banding signal in the SAF-immunoblot. On the other hand,
a positive scrapie sample still gave a strong PrPSc signal
even after exposure to 500 
g/ml PK, while the PrPSc signal
of an atypical scrapie case was only clearly detectable after
incubation with 50 
g/ml (Fig. 3).
French scrapie cases were also tested using a modified
SAF-immunoblotting protocol, in which the first ultracentrifugation
step was omitted and the homogenate was treated
directly with proteinase K. Clear PrPSc signals were seen
in 66 samples initially detected using the sELISA I test as
well as in 131 samples detected using the rapid WB. However,
in 29 atypical samples no PrPSc banding pattern was
Fig. 3. To compare the PK resistance of PrPSc derived from typical
scrapie samples with positive results in all tests to that of samples
with non-uniform
results, we performed SAF preparations with PK concentrations from 2.5
to 500 
g/ml. While PrPC from an uninfected sheep control was completely
digested by 2.5 
g/ml, PrPSc from a typical scrapie sample (T2G) still
gave a strong signal after incubation with 500 
g/ml for 1 h at 37 ?C.
PrPSc from
an atypical scrapie sample (A24G) was almost completely digested after
incubation with 250 
g/ml. L42 was used as a detection antibody.
observed. Selected samples were subsequently tested by the
original SAF-immunoblot and distinct PrPSc patterns were
demonstrated (Fig. 4).
The PrP genotypes of all German and 63 French scrapie
cases (wherever suitable material was available) were determined.
Without employing a full statistical analysis on
relative allele frequencies, it seemed that atypical cases are
not linked to one particular PrP genotype or allele but derive
from a variety of genotypes: ARQ/ARQ, ARQ/AHQ,
AHQ/AHQ, ARQ/ARR, and AHQ/ARR (in decreasing order
of assumed scrapie susceptibilities). Fishers exact test
revealed no significant association between a distinct PrP
genotype and the result of rapid WB. A straight linkage between
the non-uniform diagnostic recognition and genetic
effects of a particular allele can therefore be excluded. However,
it should be noted that 6 out of 10 deviant French and
10 out of 20 deviant German sheep carried an allele assumed
to be linked to a higher resistance to scrapie (ARR or
34 A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736
Fig. 4. Confirmation of six French scrapie cases by OIE approved
SAF-immunoblotting according to the OIE approved protocol. Sample
amounts of 0.7 g were purified and extracts loaded per lane. Mab P4 was
used as a detection antibody and the light emission of CDP-Star was
recorded using an imaging system. Note strong PrPSc banding patterns
which were found in the atypical samples. As detection antibodies, a
combination of mabs P4 and 6H4 was applied, conjugate GAM-AP was
used.
AHQ) whereas such genotypes were found in only 4 out of
53 French and in no other German case that gave positive
results in all tests. Similarly, incongruent results were found
in sheep belonging to several breeds (e.g. Suffolk, Merino,
Texel) so that a correlation to specific breeds may also be
ruled out.
4. Discussion
Since the spring of 2002, rapid tests are being used for
active surveillance of scrapie in the national sheep herds of
the EU member states. These tests have been approved by
the European commission and they include a colorimetric
sandwich ELISA, a luminescence sandwich ELISA, an indirect
ELISA using chemiluminescence as well as a rapid
Western blotting assay. According to EU legislation (EC
regulation 999/2001 and amendments), rapid tests are only
approved for screening brain samples. In case of a reactive
result, the sample has to be examined in the national reference
laboratory using one of the OIE approved confirmatory
methods which are SAF-immunoblotting and immunohistochemistry.
The introduction of this active surveillance
programme in the EU using BSE rapid tests demonstrated
that scrapie in small ruminants is much more prevalent than
had been previously estimated. Moreover, we found that not
all scrapie cases are detected equally well by the four applied
rapid tests as a significant portion of the samples were
found positive with the sELISA I test but were not reactive
with the rapid WB and sELISA II. The cELISA could only
be used on a rather small number of samples, however, it
must be assumed from the available results that it shows a
performance similar to that of the rapid WB and sELISA
II. Although some of the sELISA I results were weak positive,
all cases that were initially detected using this method
were confirmed by using OIE-recommended methods. It is
therefore concluded that the sELISA I results are true positive,
whereas the rapid WB, sELISA II and cELISA results
of the same samples must be considered as false negative.
This of course presupposes that the currently applied con-
firmatory methods do not produce any false positive results.
All samples that were first detected by using the rapid WB
also gave positive results using the other rapid test methods.
This observation supports the hypothesis that the rapid WB
fails to detect certain positive sheep. Regrettably, it was not
feasible during this study to retest small ruminant field samples
that had initially been negative in the rapid WB with
the sELISA I test in order to check if any positive samples
had been ignored during the first screening. Non-uniform
rapid test results were reproducible when selected coded
samples were exchanged between the German and French
national TSE reference laboratories and the samples were
repeatedly positive in the sELISA I. Furthermore, it should
be emphasised that the reactivity using this test was generally
high (more than four times the cut-off level in most
cases), showing that negative results obtained with other
methods cannot be explained by threshold levels of protease
resistant prion protein in these particular samples alone.
Brainstem samples collected in abattoirs and rendering
plants do not always fulfill the desired diagnostic quality
standards in terms of freshness and in sampling localisation.
In the beginning, we therefore could not exclude the possibility
that single incongruent test results were due to varying
PrPSc concentrations in the brainstems. However, the large
number of such samples and the variety of the sample histories
argue against such artefacts. However, such effects may
explain why a few cases were negative by immunohistochemistry,
but positive by SAF-immunoblotting, a diagnostic
method where PrPSc is concentrated from larger brain
areas.
These 53 atypical scrapie cases in France and Germany
out of a set of 276 may represent a novel strain of this disease
in the field. Similar observations have also been made by
the Norwegian National TSE Reference Laboratory, where
some sheep scrapie samples were not reactive using the rapid
WB (Benestad et al., 2003). However, not all characteristics
described for those Nor98 designated cases match the
atypical scrapie cases reported here. In particular, all Nor98
cases display a strong signal at 12 kDa that seems to be absent
in a number of the German and French cases. We therefore
postulate that at least three different scrapie phenotypes
A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736 35
(typical scrapie and two atypical strains) exist within the
European sheep flock. To determine critical factors affecting
the rapid WB detection of PrPSc, we undertook a series
of exchange experiments. The detection of atypical cases
was improved after replacing the rapid WB homogenisation
buffer supplied within the testkit of which the composition is
undisclosed by an in-house homogenisation buffer containing
desoxycholate and NP40 as detergents. Attempts were
also made to replace the proteinase K solution, the detection
antibody (by mab L42 that is directed to the same epitope as
mab 6H4), and the conjugate antibody. Although no single
critical step was revealed that alone enabled the detection
of samples, these modifications altogether led to positive results
for almost all atypical samples. It became also evident
that minor changes in the sample treatment may have major
effects when the modified SAF-immunoblotting technique
was applied: while some of the atypical samples were negative
with this method in which the first ultracentrifugation
step is omitted and PK digestion is performed prior to SAF
preparation, the same samples became clearly positive when
the other SAF preparation protocol was applied.
However, the physiochemical characteristics of ovine
PrPSc derived from different scrapie isolates that are the basis
for the observed effects still need further research efforts.
PrP molecules derived from animals affected with different
TSE strains vary in their cleavage sites for proteinase K
digestion and therefore display different molecular weights
when analysed in immunoblot (Hill et al., 1998; Baron et al.,
2000; Stack et al., 2002). For example, the N-terminus of
PrP derived from BSE-affected sheep is digested further by
proteinase K than PrP derived from scrapie-affected sheep
and therefore leads to a residual PrP of a lower molecular
mass. This effect has been proposed as a possible diagnostic
marker to differentiate between BSE and scrapie infections
in sheep. Our experiments showed that this variation in the
PK cleavage site between scrapie and BSE in sheep has no
impact on the performance of the rapid tests on ovine BSE
PrPSc since all commercial rapid tests detect PrPSc derived
from experimentally BSE infected sheep of the PrPARQ/ARQ
genotype (data not shown). Moreover, WB and sELISA II
use the same monoclonal antibody (mab 6H4; Korth et al.,
1997) which binds to an epitope far away from the PK cleavage
site (aa 144148) and would therefore not be expected to
react differently with scrapie- and BSE-derived PrPSc. The
same is the case with the French immunoblot test that uses
mab SAF 84 (Demart et al., 1999), binding to an epitope in
the same region of the protein (aa 125163). No information
is available on the PrP epitopes that are targeted in the
sELISA I and cELISA.
TSE infectivity can only be confined reliably, to date, by
transmission experiments to an appropriate host. As it cannot
be ruled out completely at this stage that non-infectious
PrPC may also form intracellular aggregates with increased
protease resistance and hydrophobicity that may lead to false
positive results in diagnostic tests, the level of infectivity
of such atypical cases is currently being examined by inoculation
into RIII, C57B1 and VM95 mice. In case of a
transmission to these mouse strains, the lesion profile scores
will be determined and PrPSc will be analysed concerning
its glycotype and PK resistance.
The use of rapid tests for small ruminants was introduced
by the EU Commission on the basis of their successful evaluation
for BSE testing in cattle (Moynagh and Schimmel,
1999) in order to achieve an overview of the scrapie prevalence
in the EU. Unfortunately, no independent evaluation
has been performed in the EU to date to reveal the individual
rapid tests performance on small ruminant scrapie cases.
Therefore, their specificity and sensitivity for this use can
only be estimated by the results of samples selected randomly
that have been tested individually by the manufacturers.
Inconsistencies in the ability of rapid tests to identify
positive cases would question the current efforts to intensify
and standardise the scrapie surveillance in the EU member
states. Our data show that the actual numbers of scrapie
cases and the prevalence of scrapie may be seriously underestimated
in countries where rapid tests that may produce
false-negative results are used. In the German epidemiosurveillance
scheme for scrapie, the sELISA I is applied for
testing of more than 80% of the samples, while in France
60% of the samples are tested with the rapid WB. Therefore,
it must be accepted that the current EU-wide epidemiosurveillance
programme can only give a general impression
of the scrapie situation but may miss on average up to 12%
of the true number of German scrapie cases and up to 16%
of the French cases (estimated numbers take into account
the applied test methods and the numbers of atypical cases
since 2002). This must be kept in mind when scrapie prevalence
data obtained by BSE rapid testing are interpreted.
Acknowledgements
We wish to acknowledge Matthias Kramer and Sandra Göbel
(FRCVDA-Wusterhausen, Germany) and Didier Calavas
(AFSSA-Lyon, France) for epidemiological data, Bertrand
Bedhom (LABOGENA, France) for genotype analysis of
French scrapie cases and J. Grassi (C.E.A.-Saclay, France)
for supply of SAF 70 and SAF 84 antibody. This work was
partly funded by the German Ministry of Consumer Protection,
Nutrition and Agriculture (BMVEL).
References
Baron, T.G., Madec, J.Y., Calavas, D., 1999. Similar protein in natural
sheep scrapie and bovine spongiform encephalopathy-linked diseases.
J. Clin. Microbiol. 37, 37013704.
Baron, T.G., Madec, J.Y., Calavas, D., Richard, Y., Barillet, F., 2000.
Comparison of French natural scrapie isolates with bovine spongiform
encephalopathy and experimental scrapie infected sheep. Neurosci.
Lett. 284, 175178.
Belt, P.B., Muileman, I.H., Schreuder, B.E., Bos-de Ruijter, J., Gielkens,
A.L., Smits, M.A., 1995. Identification of five allelic variants of the
36 A. Buschmann et al. / Journal of Virological Methods 117 (2004) 2736
sheep PrP gene and their association with natural scrapie. J. Gen.
Virol. 76, 509517.
Benestad, S.L., Sarradin, P., Thu, B., Schonheit, J., Tranulis, M.A.,
Bratberg,
B., 2003. Cases of scrapie with unusual features in Norway and
designation of a new type, Nor98. Vet. Rec. 153, 202208.
Bruce, M., Chree, A., McConnell, I., Brown, K., Fraser, H., 1996.
Transmission
and strain typing studies of scrapie and bovine spongiform
encephalopathy. In: Court, L., Dodet, B. (Eds.), Transmissible Subacute
Spongiform Encephalopathies: Prion Diseases. Elsevier, Paris,
pp. 259262.
Bruce, M.E., Will, R.G., Ironside, J.W., McConnell, I., Drummond, D.,
Suttie, A., McCardle, L., Chree, A., Hope, J., Birkett, C., Cousens, S.,
Fraser, H., Bostock, C.J., 1997. Transmissions to mice indicate that
new variant CJD is caused by the BSE agent. Nature 389, 498501.
Collinge, J., Sidle, K.C., Meads, J., Ironside, J., Hill, A.F., 1996.
Molecular
analysis of prion strain variation and the aetiology of new variant
CJD. Nature 383, 685690 (see comments).
Debeer, S.O., Raron, T.G., Bencsik, A.A., 2001. Immunohistochemistry
of PrPSc within bovine spongiform encephalopathy brain samples with
graded autolysis. J. Histochem. Cytochem. 49, 15191524.
Debeer, S.O., Baron, T.G., Bencsik, A.A., 2002. Transmissible spongiform
encephalopathy diagnosis using PrPSc immunohistochemistry on fixed
but previously frozen brain samples. J. Histochem. Cytochem. 50,
611616.
Demart, S., Fournier, J.G., Cremignon, C., Frobert, Y., Lamoury, F.,
Marce, D., Lasmézas, C., Dormont, D., Grassi, J., Deslys, J.-P., 1999.
New insight into abnormal prion protein using monoclonal antibodies.
Biochem. Biophys. Res. Commun. 265, 652657.
Foster, J.D., Hope, J., Fraser, H., 1993. Transmission of bovine spongiform
encephalopathy to sheep and goats. Vet. Rec. 133, 339341.
Foster, J.D., Parnham, D., Chong, A., Goldmann, W., Hunter, N., 2001.
Clinical signs, histopathology and genetics of experimental transmission
of BSE and natural scrapie to sheep and goats. Vet. Rec. 148,
165171.
Fraser, H., Dickinson, A.G., 1968. The sequential development of the
brain lesion of scrapie in three strains of mice. J. Comp. Pathol. 78,
301311.
Goldmann, W., Hunter, N., Smith, G., Foster, J.D., Hope, J., 1994. PrP
genotype and agent effects in scrapie: change in allelic interaction
with different isolates of agent in sheep, a natural host of scrapie. J.
Gen. Virol. 75, 989995.
Groschup, M.H., Kuczius, T., Junghans, F., Sweeney, T., Bodemer,
W., Buschmann, A., 2000. Characterization of BSE and scrapie
strains/isolates. Arch. Virol. 16 (Suppl.), 217226.
Hardt, M., Baron, T., Groschup, M.H., 2000. A comparative study of
immunohistochemical methods for detecting abnormal prion protein
with monoclonal and polyclonal antibodies. J. Comp. Pathol. 122,
4353.
Harmeyer, S., Pfaff, E., Groschup, M.H., 1998. Synthetic peptide vaccines
yield monoclonal antibodies to cellular and pathological prion proteins
of ruminants. J. Gen. Virol. 79, 937945.
Hill, A.F., Sidle, K.C., Joiner, S., Keyes, P., Martin, T.C., Dawson, M.,
Collinge, J., 1998. Molecular screening of sheep for bovine spongiform
encephalopathy. Neurosci. Lett. 255, 159162.
Hope, J., Wood, S.C., Birkett, C.R., Chong, A., Bruce, M.E., Cairns, D.,
Goldmann, W., Hunter, N., Bostock, C.J., 1999. Molecular analysis
of ovine prion protein identifies similarities between BSE and an
experimental isolate of natural scrapie, CH 1641. J. Gen. Virol. 80,
14.
Junghans, F., Teufel, B., Buschmann, A., Steng, G., Groschup, M.H., 1998.
Genotyping of German sheep with respect to scrapie susceptibility.
Vet. Rec. 143, 340341.
Kascsak, R.J., Rubenstein, R., Merz, P.A., Carp, R.I., Robakis, N.K.,
Wisniewski, H.M., Diringer, H., 1986. Immunological comparison
of scrapie-associated fibrils isolated from animals infected with four
different scrapie strains. J. Virol. 59, 676683.
Korth, C., Stierli, B., Streit, P., Moser, M., Schaller, O., Fischer,
R., Schulz-Schaeffer, W., Kretzschmar, H., Raeber, A., Braun, U.,
Ehrensperger, F., Wüthrich, K., Oesch, B., 1997. Prion (PrPSc)-specific
epitope defined by a monoclonal antibody. Nature 390, 7477.
Kuczius, T., Haist, I., Groschup, M.H., 1998. Molecular analysis of bovine
spongiform encephalopathy and scrapie strain variation. J. Infect. Dis.
178, 693699.
Laplanche, J.L., Chatelain, J., Westaway, D., Thomas, S., Dussaucy, M.,
Brugere-Picoux, J., Launay, J.M., 1993. PrP polymorphisms associated
with natural scrapie discovered by denaturing gradient gel electrophoresis.
Genomics 15, 3037.
Lehmann, S., Harris, D.A., 1995. A mutant prion protein displays an
aberrant membrane association when expressed in cultured cells. J.
Biol. Chem. 270, 2458924597.
Madec, J.Y., Belli, P., Calavas, D., Baron, T., 2000. Efficiency of Western
blotting for the specific immunodetection of proteinase K-resistant
prion protein in BSE diagnosis in France. Vet. Rec. 146, 7476.
McGowan, J.P., 1922. Scrapie in sheep. J. Agric. 5, 365375.
Moynagh, J., Schimmel, H., 1999. Tests for BSE evaluated. Bovine
spongiform encephalopathy. Environ. Health Perspect. 400, 105.
ODoherty, E., Healy, A., Aherne, M., Hanrahan, J.P., Weavers, E., Doherty,
M., Roche, J.F., Gunn, M., Sweeney, T., 2002. Prion protein
(PrP) gene polymorphisms associated with natural scrapie cases and
their flock-mates in Ireland. Res. Vet. Sci. 73, 243250.
ORourke, K.I., Baszler, T.V., Miller, J.M., Spraker, T.R., Sadler-
Riggleman, I., Knowles, D.P., 1998. Monoclonal antibody F89/160.1.5
defines a conserved epitope on the ruminant prion protein. J. Clin.
Micobiol. 36, 17501755.
Oesch, B., Jensen, M., Nilsson, P., Fogh, J., 1994. Properties of the
scrapie
prion protein: quantitative analysis of protease resistance. Biochemistry
33, 59265931.
Stack, M., Chaplin, M., Clark, J., 2002. Differentiation of prion protein
glycoforms from naturally occurring sheep scrapie, sheep-passaged
scrapie strains (CH1641 and SSBP1), bovine spongiform encephalopathy
(BSE) cases and Romney and Cheviot breed sheep experimentally
inoculated with BSE using two monoclonal antibodies. Acta Neuropathol.
104, 279286.
Sweeney, T., Kuczius, T., McEloy, M., Parada, M.G., Groschup, M.H.,
2000. Molecular analysis of Irish sheep scrapie cases. J. Gen. Virol.
81, 16211627.
Will, R.G., Ironside, J.W., Zeidler, M., Cousens, S.N., Estibeiro, K.,
Alperovitch, A., Poser, S., Pocchiari, M., Hofman, A., Smith, P.G.,
1996. A new variant of CreutzfeldtJakob disease in the UK. Lancet
347, 921925.
Zanusso, G., Casalone, C., Acutis, P., Bozzetta, E., Farinazzo, A., Gelati,
M., Fiorini, M., Forloni, G., Sy, M.S., Monaco, S., Caramelli, M.,
2003. Molecular analysis of iatrogenic scrapie in Italy. J. Gen. Virol.
84, 10471052...TSS

###########  http://mailhost.rz.uni-karlsruhe.de/warc/bse-l.html  ############