Journal of General Virology (2012), 93, 2518–2527
DOI 10.1099/vir.0.045849-0
Infectious titres of sheep scrapie and bovine
spongiform encephalopathy agents cannot be
accurately predicted from quantitative laboratory
test results
Lorenzo González,1 Leigh Thorne,2 Martin Jeffrey,1 Stuart Martin,1
John Spiropoulos,2 Katy E. Beck,2 Richard W. Lockey,23
Christopher M. Vickery,2 Thomas Holder2 and Linda Terry24
Correspondence
Lorenzo González
[email protected]
Received 6 July 2012
Accepted 16 August 2012
1
Animal Health and Veterinary Laboratories Agency (AHVLA), AHVLA-Lasswade,
Pentlands Science Park, Penicuick, Midlothian EH26 0PZ, UK
2
Animal Health and Veterinary Laboratories Agency (AHVLA), AHVLA-Weybridge, Addlestone,
Surrey KT15 3NB, UK
It is widely accepted that abnormal forms of the prion protein (PrP) are the best surrogate marker
for the infectious agent of prion diseases and, in practice, the detection of such diseaseassociated (PrPd) and/or protease-resistant (PrPres) forms of PrP is the cornerstone of diagnosis
and surveillance of the transmissible spongiform encephalopathies (TSEs). Nevertheless, some
studies question the consistent association between infectivity and abnormal PrP detection. To
address this discrepancy, 11 brain samples of sheep affected with natural scrapie or experimental
bovine spongiform encephalopathy were selected on the basis of the magnitude and predominant
types of PrPd accumulation, as shown by immunohistochemical (IHC) examination; contra-lateral
hemi-brain samples were inoculated at three different dilutions into transgenic mice
overexpressing ovine PrP and were also subjected to quantitative analysis by three biochemical
tests (BCTs). Six samples gave ‘low’ infectious titres (106.5 to 106.7 LD50 g”1) and five gave ‘high
titres’ (108.1 to ¢108.7 LD50 g”1) and, with the exception of the Western blot analysis, those two
groups tended to correspond with samples with lower PrPd/PrPres results by IHC/BCTs.
However, no statistical association could be confirmed due to high individual sample variability. It
is concluded that although detection of abnormal forms of PrP by laboratory methods remains
useful to confirm TSE infection, infectivity titres cannot be predicted from quantitative test results,
at least for the TSE sources and host PRNP genotypes used in this study. Furthermore, the near
inverse correlation between infectious titres and Western blot results (high protease pretreatment) argues for a dissociation between infectivity and PrPres.
INTRODUCTION
The detection of abnormal prion protein (PrP) accumulating in the brains of infected animals is the basis of the
diagnostic tests for prion diseases or transmissible spongiform encephalopathies (TSEs). Whether by visualization of
disease-associated PrP (PrPd, also designated PrPSc)
aggregates by immunohistochemistry (IHC) or by detection of proteinase K (PK)-resistant PrP (PrPres) by
biochemical tests (BCTs), demonstrating the presence of
3Present address: University of Southampton, University Road,
Southampton SO17 1BJ, UK.
4Present address: Department of Pathology and Immunology, CMU,
University of Geneva, 1 Rue Miguel Servet, 1211 Geneva 4, Switzerland.
A supplementary table is available with the online version of this paper.
2518
abnormal PrP remains the cornerstone of TSE diagnosis,
and PrPd/PrPres is widely considered a surrogate marker of
the infectious agent(s). However, early evidence from a
variety of experiments (reviewed by Rohwer, 1991)
suggests that relationships between infectivity and positive
test results for PrPd/PrPres are not simple. More recently, in
certain TSE models where infectivity has been demonstrated, little or no PrPres can be detected by conventional
laboratory methods (Lasmezas et al., 1997; Barron &
Manson, 2003; Barron et al., 2007; Balkema-Buschmann
et al., 2011), while in other models the magnitude of PrPd
revealed by IHC or histoblotting lacks correlation with
infectivity (Jeffrey et al., 2001a). Such studies suggest the
existence of infectious and pathological conformations of
PrP that are not readily detectable by current diagnostic
tests. Conversely, not all PrPd/PrPres forms that are
Downloaded from www.microbiologyresearch.org by
045849 G 2012 Crown copyright
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
Printed in Great Britain
TSE infectivity versus laboratory tests
detected by laboratory tests methods may necessarily be
infectious (Piccardo et al., 2007; Jeffrey et al., 2012), and
dissociation between abundance of PrPres and infectivity
has also been reported as one of the arguments against the
prion hypothesis (Miyazawa et al., 2011).
Determining the association or dissociation between the
TSE infectious agent(s) and its surrogate markers is of
importance in terms of possible underdiagnosis, not only
at the level of the animal but also in the identification of
tissues that contain infectivity. Such determination is
therefore of relevance for food safety, animal health and
disease eradication policies.
bioassay, and levels of abnormal PrP, as detected by standard
IHC and BCTs, and (ii) whether the levels of infectivity and
test outputs from different PrP detection methods correlate
with specific IHC types of PrPd accumulating in the brains of
clinically affected, TSE-infected sheep.
RESULTS
Correlation between BCTs
For the 40 scrapie samples tested (see details in Methods),
the three BCTs showed high correlation values, with
Spearman r values of 0.70 for enzyme immunoassay (EIA)
versus Western blot (WB), 0.72 for enzyme-linked immunosorbent assay (ELISA) versus WB and 0.72 for ELISA
versus EIA (Fig. 1a). The correlations were also high for BSE
samples, with r values between 0.69 (EIA vs WB) and 0.95
(ELISA vs EIA) for ARQ/ARQ sheep samples and between
0.74 (ELISA vs EIA) and 0.92 (ELISA vs WB) for ARR/ARR
sheep samples (Fig. 1b). It is worth noting that ARR/ARR
sheep BSE samples provided much lower values in all three
BCTs than those from ARQ/ARQ sheep (Fig. 1b).
In previous studies we have shown that, depending on the
TSE strain and model, sheep diagnosed with clinical disease
display different patterns of deposition of PrPd in the brain
as demonstrated by IHC. For example, in the brains of
CH1641 experimentally infected sheep, PrPd deposits are
almost exclusively found within the cytoplasm of neurons
and microglial cells (Jeffrey et al., 2006), while in a model
of natural scrapie in Suffolk sheep, the majority of PrPd is
extracellular and predominantly associated with astrocytes
(González et al., 2002, 2003, 2010b; Sisó et al., 2010); by
ultrastructural examination such extracellular PrPd is
shown to aggregate at the cell membrane (Jeffrey et al.,
2011). In further contrast, sheep experimentally infected
with the bovine spongiform encephalopathy (BSE) agent
display conspicuous intra- and extracellular PrPd deposits
(González et al., 2005a). These observations exemplify
differences in morphology and cell-association of PrPd
aggregates among different ovine TSEs. Differences in
magnitude or levels of PrPd accumulation in the brain of
clinically affected sheep are also demonstrable in different
disease situations. Thus, sheep dosed orally with BSE have
lower amounts of PrPd in the forebrain compared with
sheep inoculated intracerebrally with the same inoculum
and ARR/ARR sheep (A, alanine and R, arginine at codons
136, 154 and 171 of ovine PrP) challenged intracerebrally
with BSE have little PrPd when compared with ARQ/ARQ
sheep (Q, glutamine) challenged with the same inoculum
and by the same route and at the same stage of clinical
disease (González et al., 2005a). Similarly, VRQ/ARR sheep
(V, valine) experimentally infected with the SSBP/1 scrapie
source (González et al., 2002) and some naturally infected
goats (González et al., 2010a; Konold et al., 2010), despite
showing clinical signs of scrapie, display low magnitudes of
brain PrPd. Indeed, when the IHC abundance of PrPd is
determined for different models of TSE-infected sheep,
there is no correlation between the amount of PrPd present,
the incubation period and the presence of clinical disease.
These observations raise questions about the nature of the
infectious form of PrP and the efficiency of detection of
different abnormal PrP forms and types by different test
methods.
Significant correlations were obtained when comparing
IHC scores for total PrPd and the results of the BCTs on the
40 scrapie brain samples examined (Table 1, Fig. 2a), with
Spearman’s r coefficients of 0.41 for EIA, 0.52 for ELISA
and 0.68 for WB. The magnitude of accumulation of all the
different PrPd types recognized by IHC correlated well with
the end-point dilution values obtained in the three BCTs
with the exception of extracellular, astrocyte-associated
PrPd (ASTR: subpial, peri-vascular, peri-vacuolar and
subependymal types), which did not show correlation with
any of the three BCTs. The stellate PrPd type (STEL:
extracellular, possibly associated with resting microglia)
gave the highest correlation coefficients with all three
BCTs, followed by intracellular PrPd (ITCL: intra-neuronal, intra-microglial and intra-astrocytic types) and by
extracellular PrPd in the grey matter neuropil (NRPL:
particulate/coalescing, linear and peri-neuronal).
With these premises in mind, the present study questioned
(i) whether a correlation exists between TSE infectivity
levels in brain tissue of sheep, as determined by mouse
When the correlation between IHC and BCTs was analysed
independently for the four brain areas (Table S1), similarly
to what was observed when comparing the three BCTs, the
http://vir.sgmjournals.org
When analysing the association between BCT results
independently for each brain area examined (Fig. 1 and
details in Table S1, available in JGV Online), scrapie samples
provided significant correlation values between all three BCTs
in the cerebral cortex (0.77–0.95) and in the cerebellum
(0.76–0.95), only between EIA and ELISA in the corpus
striatum (0.65) and no significant correlations in the obex
(20.05–0.48). For ARQ/ARQ BSE samples, correlations
between the three BCTs were significant in the cerebellum
(0.78–0.93), and only between ELISA and EIA in the cerebral
cortex (0.89) and in the corpus striatum (0.41–0.86).
Correlation between IHC and BCTs
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
2519
L. González and others
(a) Scrapie
EIA
(b) BSE
20000
EIA
25000
20000
15 0 0 0
rs=0.72 ***
rs ARQ =0.95 ***
15 0 0 0
10 0 0 0
10 0 0 0
5000
rs ARR =0.74 **
5000
0
0
0
500
10 0 0
15 0 0
2000
2500
3000
0
500
10 0 0
15 0 0
2000
ELISA
WB
40000
2500
3000
ELISA
WB
10 0 0 0
8000
30000
rs=0.72 ***
rs ARQ =0.73 ***
6000
20000
4000
10 0 0 0
rs ARR =0.92 ***
2000
0
0
0
500
10 0 0
15 0 0
2000
2500
3000
0
500
10 0 0
15 0 0
2000
40000
3000
ELISA
ELISA
WB
2500
WB
10 0 0 0
8000
30000
rs=0.70 ***
rs ARQ =0.69 ***
6000
20000
4000
10 0 0 0
rs ARR =0.83 ***
2000
0
0
0
5000
10 0 0 0
15 0 0 0
20000
0
5000
EIA
10 0 0 0
15 0 0 0
20000
25000
EIA
Fig. 1. Correlation between BCTs. Results indicate end-point dilution values for each sample (scrapie, n540; ARQ/ARQ BSE,
n521; ARR/ARR BSE, n512) in the three different tests. Correlation coefficients [Spearman r (rs)] and their significance
(**, P,0.01; ***, P,0.001) are given for each comparison . Note that scales have been adjusted to actual values to provide a
better representation of the correlations. For scrapie and ARQ/ARQ BSE samples open symbols indicate: circles, cerebral
cortex; squares, corpus striatum; triangles, cerebellum; diamonds, obex (scrapie only). Black diamonds represent ARR/ARR
BSE samples, regardless of brain area.
samples of obex did not provide significant correlation
with either total PrPd or with the different types
considered. Correlations were best in the cerebral cortex,
where all four PrPd types considered correlated with all
three BCT results, and where correlation with total PrPd
was at its highest (0.84 for EIA to 0.94 for ELISA). Samples
of cerebellum also provided high correlation coefficients,
with the exception of ITCL PrPd, whereas in the corpus
2520
striatum, only WB provided significant correlation coefficients, again with the exception of ITCL PrPd. Notably, the
highest correlation coefficients in all these brain areas were
for the STEL PrPd type, in line with the values obtained in
the overall analysis. In contrast, ASTR PrPd, which gave
non-significant correlation in the global analysis, showed
high correlation with the BCTs in the cerebral cortex, and
less so in the cerebellum and corpus striatum.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
Journal of General Virology 93
TSE infectivity versus laboratory tests
Table 1. Correlation analyses between IHC and BCTs in brain
samples
P-values of correlation analyses are indicated as: NS, not significant; *,
P,0.05; **, P,0.01; ***, P,0.001. ITCL, Intracellular; STEL, stellate;
ASTR, astrocyte-associated extracellular; NRPL, neuropil-associated
extracellular (for extended description see text). Note that in the case
of ARR/ARR BSE samples data are not shown for another type of
PrPd, non-vascular plaques, which are nevertheless considered to
provide the total PrPd values.
ITCL
STEL
ASTR
NRPL
Total
PrPd
r
P
r
P
r
P
r
P
r
P
ELISA 0.57
***
0.73
***
20.02
NS
0.60
***
0.52
***
EIA
0.79
***
0.69
***
20.27
NS
0.47
**
0.41
**
WB
0.60
***
0.80
***
0.11
NS
0.63
***
0.68
***
Scrapie (n540)
BSE ARQ (n521)
ELISA 20.02
NS
0.59
**
20.17
NS
0.09
NS
0.09
NS
EIA
20.09
NS
0.47
*
20.23
NS
0.10
NS
20.01
NS
WB
0.00
NS
0.41
NS
20.33
NS
0.22
NS
0.07
NS
gave lower ELISA (mean 466) and EIA (mean 1130) values
than those with high infectious titres (mean values of 9.6,
779 and 2860 for IHC, ELISA and EIA, respectively); their
WB results, however, were higher (mean 1821) than those
of samples with high infectious titre (mean 781). None of
those differences was significant in the Mann–Whitney,
non-parametric test due to high individual variability of
test results within each group (Figs 3 and 4). For IHC for
example, while the three samples with the lowest magnitude of total PrPd [BSE1 Cc, Scra1 St (Fig. 4) and BSE2 Cc]
were in the group with low infectious titre and the two with
the highest PrPd scores [BSE6 Cc (Fig. 4) and BSE6 Cb]
were in the group of high titre, samples with intermediate
IHC scores could be in either group (Table 2 and Fig. 4).
A further analysis was performed to determine the
relationship between infectivity and the scores of different
PrPd types determined by IHC examination. Although the
mean scores of all PrPd types considered were higher in the
group of high than of low infectious titre, the only
significant difference was with respect of the STEL PrPd
type (Table 2 and Fig. 3).
BSE ARR (n512)
ELISA 0.47
NS
0.42
NS
20.26
NS
0.68
*
0.59
*
EIA
0.73
**
0.48
NS
20.18
NS
0.71
**
0.71
**
WB
0.66
*
0.34
NS
20.32
NS
0.70
*
0.60
*
For the 33 BSE brain samples, the correlation obtained
between IHC and BCTs was much lower than that
observed for the scrapie samples (Table 1, Fig. 2b). The
only correlations observed were between STEL PrPd and
ELISA and EIA in ARQ/ARQ sheep, and between NRPL
PrPd and all three BCTs and EIA and WB with ITCL PrPd
in ARR/ARR sheep.
The analysis of ARQ/ARQ BSE samples by brain area
confirmed the overall analysis insofar as the STEL type was
the one showing the highest correlation coefficients in all
three brain areas, although these were only significant for
ELISA and EIA in the cerebral cortex and for EIA in the
cerebellum (Supplementary data).
Correlation between laboratory test results and
infectivity
Table 2 provides details of attack rates and incubation periods
of the bioassay carried out in TgshpXI mice with the 11 brain
samples selected on the basis of their PrPd profile and total
magnitude of PrPd accumulation. Based on attack rates at the
different dilutions the infectious titres were calculated, and it
became clear that the sheep inocula could be divided in two
groups: those that killed 90–100 % of mice at the 1023 and
0 % at the 1025 dilutions (n56, titres 106.5–106.7LD50 g21,
‘low’), and those that killed 70–100 % of mice at the 1025
dilution (n55, titres 108.1–¢108.7LD50 g21, ‘high’).
Sheep samples of low infectious titres had an average lower
magnitude of IHC-detectable PrPd (mean 5.4), and also
http://vir.sgmjournals.org
DISCUSSION
A clear correlation between laboratory test results, that is
levels of abnormal PrP detectable by different methods, and
infectious titre, has not been demonstrated within this
study. It is however true that, with the exception of WB,
sheep brain samples containing higher levels of abnormal
PrP tended to provide higher infectious titres, although the
individual variability in laboratory test results made differences to be non-significant. In addition to the small
sample size, a number of other considerations need to be
made in order to discuss and interpret this somewhat poor
correlation: the assessment of infectivity in TgshpXI mice
was done at 100-fold dilutions, the highest being at 1025;
this created three problems that could have affected the
discriminatory capability of the comparative analysis: (i)
the bimodal rather than sequential distribution of infectious titres, which prevented the realization of correlation
analyses, (ii) the close proximity (in some cases identity) of
the titres of some samples, and (iii) the inability to
calculate precise infectious titres in the two samples that
resulted in 100 % attack rate at the highest dilution.
However, despite the fact that the infectious titres obtained
partitioned into two distinct groups [considered to be of
statistically significant different infectivity since titres
differed by more than 0.8 log units (Mould et al., 1967)],
no clear correlation was found with corresponding test
results. Therefore, it is also likely that a more accurate
calculation of titres would not have fundamentally changed
their correlation with test results.
Because the selection of samples for the different analyses was
done from archival material, the samples scored for PrPd by
IHC were from one hemi-brain while those for biochemical
analysis and bioassay were from the other hemi-brain.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
2521
L. González and others
(b) BSE
(a) Scrapie
PrP d
PrP d
12
20
9
15
6
10
rs ARQ =0.09
3
rs ARR =0.59 *
5
rs=0.52 ***
0
0
0
500
10 0 0
15 0 0
2000
2500
0
3000
500
10 0 0
15 0 0
2000
ELISA
PrP d
PrP d
20
9
15
6
10
rs ARQ =-0.01
rs ARR =0.71 **
5
rs=0.41 **
0
0
0
5000
10 0 0 0
15 0 0 0
20000
0
5000
10 0 0 0
15 0 0 0
20000
EIA
PrP d
3000
ELISA
12
3
2500
EIA
PrP d
12
20
9
15
6
10
rs=0.68 ***
3
25000
rs ARQ =0.07
rs ARR =0.60 *
5
0
0
0
7000
14 0 0 0
2 10 0 0
28000
35000
0
2000
WB
4000
6000
8000
10 0 0 0
WB
Fig. 2. Correlation between IHC and BCT results. Results indicate total PrPd scores (IHC) for each sample (scrapie, n540;
ARQ/ARQ BSE, n521; ARR/ARR BSE, n512) and their corresponding end-point dilution value in the three different BCTs.
Correlation coefficients [Spearman r (rs)] and their significance (no symbol, not significant; *, P,0.05; **, P,0.01; ***,
P,0.001) are given for each comparison. Note that scales have been adjusted to actual values to provide a better
representation of the correlations. For scrapie and ARQ/ARQ BSE samples open symbols indicate: circles, cerebral cortex;
squares, corpus striatum; triangles, cerebellum; diamonds, obex (scrapie only). Black diamonds represent ARR/ARR BSE
samples, regardless of brain area.
Although it is widely assumed that both infectivity and
abnormal PrP accumulation in the brain occur symmetrically, specific and detailed studies to verify these assumptions
have not been carried out to the best of our knowledge. It is
therefore possible that some asymmetry in the distribution of
PrPd or infectivity might have contributed to a lack of clear
correlation between IHC and infectious titres and between
2522
IHC and BCTs. This caveat, however, should not apply to the
comparison between BCT results and infectivity, as all were
carried out on the same ipsi-lateral brain samples. As the
correlation between these tests and infectious scores was even
lower than between infectivity and total amount of PrPd
detected by IHC, it appears unlikely that poor correlations
can be attributed to variations between tissue samples.
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
Journal of General Virology 93
TSE infectivity versus laboratory tests
Table 2. Individual results of infectious titre and laboratory test results of brain samples taken for bioassay
Cc, Cerebral cortex; St, corpus striatum; Ob, obex; Cb, cerebellum. AR, attack rate (%); IP, incubation period in mean days±SEM (superscripts
indicate dilution). Titres expressed as log10 of mouse (intracerebral) LD50 g21. ARs and IPs are not given for the 1021 dilution of the different
samples (ARs of 100 % in all cases and mean IPs between minimum 248 days for BSE5 and maximum 474 days for Scra2). ITCL, Intracellular (in
parentheses intraneuronal PrPd scores); STEL, stellate; ASTR, astrocyte-associated extracellular; NRPL, neuropil-associated extracellular (for
extended description see text).
Titration of infectivity in 11 brain samples by bioassay in TgshpXI mice
Sheep
171
Area
AR”3
IP”3
AR”5
BSE1
Scra1
BSE2
Scra2
BSE3
Scra3
BSE4
BSE5
BSE6
BSE1
BSE6
RR
QQ
QQ
QQ
RR
QQ
QQ
QQ
QQ
RR
QQ
Cc
St
Cc
Ob
Ob
Cc
Cc
Cb
Cb
St
Cc
90
100
100
100
100
100
100
100
100
100
100
418±21
565±41
419±14
504±37
394±11
479±32
327±10
350±6
324±8
306±11
295±18
0
0
0
0
0
0
71
86
86
100
100
IP”5
Titre
IHC
ELISA
EIA
WB
492±27
461±41
404±20
469±21
455±7
6.5
6.7
6.7
6.7
6.7
6.7
8.1
8.4
8.4
¢8.7
¢8.7
1.2
1.1
3.9
7.7
9.1
9.7
4.1
7.2
16.7
9.5
10.4
60
125
7
2048
128
425
1200
320
750
125
1500
190
370
70
3500
2000
650
3000
1100
1800
900
7500
32
128
16
8192
512
2048
1024
256
2048
64
512
IHC scores for different PrPd types in brain samples used for titration of infectivity in TgshpXI mice
Sheep
171
Area
Titre
ITCL
BSE1
Scra1
BSE2
Scra2
BSE3
Scra3
BSE4
BSE5
BSE6
BSE1
BSE6
RR
QQ
QQ
QQ
RR
QQ
QQ
QQ
QQ
RR
QQ
Cc
St
Cc
Ob
Ob
Cc
Cc
Cb
Cb
St
Cc
6.5
6.7
6.7
6.7
6.7
6.7
8.1
8.4
8.4
¢8.7
¢8.7
0.7
0.1
0.6
3.2
6
0.5
1.3
2.8
5.5
3.6
3
A possible explanation for the relatively discrepant results
between infectivity and BCTs is that these tests, which
detect either protease-resistant PrP fragments (ELISA and,
particularly, WB) or abnormal PrPd in a particular state of
aggregation (EIA), fail to detect, to a greater or lesser
extent, infectious PrP conformers (if such are actually the
infectious agent). The dissociation between infectivity and
WB results argues for a separation between highly PKresistant forms of PrP and infectivity. This is in contrast
with reports from early studies (McKinley et al., 1983), but
would be in agreement with others showing that partially
disaggregated and PK-sensitive (Rohwer, 1991; Caughey et
al., 1997), non-fibrillar PrP particles (Silveira et al., 2005)
are associated with increased infectivity in hamsters. Along
similar lines would be the reports of some cases of sporadic
Creutzfeldt–Jakob disease, in which most or all the diseaseassociated PrP is PK-sensitive (Safar et al., 2005; Gambetti
et al., 2008; Head et al., 2009). The better association
between the ELISA (low PK concentration) or the EIA (no
PK treatment) results and infectious titres would support
http://vir.sgmjournals.org
(0)
(0.1)
(0.3)
(2)
(1.5)
(0)
(0)
(1)
(1)
(1)
(0.2)
STEL
ASTR
NRPL
0
0.2
0.7
2
0.2
1.2
1.2
1.5
3
1.1
2.2
0.1
0.8
2.5
0.5
0.2
6.9
1.6
1.9
3.2
0.6
5.1
0
0
0.1
2
2.2
1.1
0
1
5
3.2
0.1
the notion of infectivity being related with partially
PK-sensitive, not highly aggregated forms of abnormal
PrP. Furthermore, the fact that our IHC protocol did not
use enzyme digestion and is able to detect a wider range of
PrPd forms in different aggregation states and of different
PK resistance (González et al., 2005b), could have contributed to the better association found between IHC and
infectivity.
With regard to the relationship between different morphological PrPd types recognized by IHC and infectivity,
the only type that showed significant correlation was the
stellate type and, to a lesser, non-statistically significant
extent, the intracellular PrPd. When these two types are
combined, the difference between the low infectious score
group (n56, mean±SEM of combined ITCL+STEL
PrPd52.4±0.81) and the high infectious score group
(n55, 5.1±0.97) is not statistically significant (P50.063 in
the t-test). The apparent correlation between infectivity and
stellate PrPd may not be biologically relevant; while
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
2523
L. González and others
12
4
10
8
3
6
2
4
1
2
1200
our study, however, are inconsistent with the second
possibility since sheep brain samples with low infectious titres
could either show conspicuous intraneuronal PrPd deposits
[e.g. BSE3 Ob (Fig. 4e) and Scra2 Ob] or none at all [e.g. Scra3
Cc (Fig. 4f) and BSE1 Cc (Fig. 4a)], and vice-versa (Table 2).
5
14
IHC
ITCL
3
1000
800
2
*
600
400
1
200
4500
4000
3500
3000
2500
2000
1500
1000
500
ELISA
4
STEL
3
2
1
EIA
ASTR
3
3500
3000
2500
2
2000
1500
1
1000
500
WB
NRPL
Fig. 3. Relationship between infectivity and laboratory test results.
Results represent the mean values (and SEM) for the different forms
of disease-associated PrP in sheep brain samples grouped by
their infectious titres: low (open bars, n56, titres between 6.5 and
6.7) or high (grey bars, n55, titres between 8.1 and ¢8.7). In the
left column note that, except for WB, samples with high infectious
scores tend to provide higher laboratory test results; however,
individual variability, particularly for the BCTs, makes differences
not to be statistically significant. Similarly in the right column,
where high infectious titre tends to correlate with higher IHC
scores of different PrPd types; however, the only significant
difference (P,0.05) is for the stellate (STEL) PrPd type. ITCL,
Intracellular; ASTR, extracellular astrocyte-associated; NRPL,
extracellular in neuropil.
intracellular, or at least intraneuronal, PrPd is commonly
found in sheep BSE and classical scrapie cases, the stellate
pattern is not a universal feature of sheep TSEs. For instance,
experimental CH1641 scrapie infection lacks stellate PrPd
(Jeffrey et al., 2006), indicating that this PrPd type can at best
provide only partial correlation with infectivity.
It has been recently reported that hamsters injected with the
‘hyper’ strain of transmissible mink encephalopathy show
short survival times and presence of intraneuronal PrPSc,
while those challenged with the ‘drowsy’ strain of the same
agent have considerably longer incubation periods and
absence of intraneuronal PrPSc deposits (Ayers et al., 2011).
The authors suggested that such differences in incubation
period are due to differences in PrPd processing within and
subsequent damage to neurons. Another explanation could be
that brains with high levels of intraneuronal PrPd are more
infectious than those without such PrPd type. The results of
2524
Some unanticipated observations have arisen from the
analyses of the laboratory test results. Firstly, correlations
between BCT results appeared to be brain area-dependent, at
least for scrapie, in which such correlations did not exist in
samples of obex but were consistently significant in the
cerebral cortex and cerebellum. Also, both for scrapie and BSE
in ARQ/ARQ sheep, ELISA and EIA provided better correlation values between them than when compared to WB. This
was not the case, however, for ARR/ARR BSE samples for
which WB showed the highest correlations. Secondly, from
the comparative analysis of IHC profiles and BCTs, it would
appear that the different morphological types of PrPd that can
be identified in the brains of sheep may have different
biochemical properties, but that these are also brain area- and
agent-dependent. Thus, in ARQ/ARQ BSE sheep, ASTR
extracellular PrPd did not correlate with BCT results in any of
the brain areas examined, while it did so in the cerebral
cortex, and to a lesser extent in the cerebellum and corpus
striatum, of scrapie-infected sheep. In contrast, in sheep of the
same ARQ/ARQ genotype, STEL PrPd was the type that
provided the highest and most consistent correlations across
different brain areas regardless of the infectious agent. Taken
together, these observations suggest different morphological
and cell-associated PrPd types may have different molecular
properties in terms of aggregation states or resistance to
protease treatment and denaturation, but that these are also
influenced by a complex combination of infectious agent,
host genetics and brain topography. Since all brains
examined in this study were from animals at clinical endpoint and since the different brain areas start accumulating
PrPd/PrPres at different times after infection [obex, cerebellum, corpus striatum and cerebral cortex from earliest to
latest (Sisó et al., 2009)], it is also possible that alterations in
molecular properties of PrPd/PrPres are time-dependent.
The high correlation found between the three BCTs
employed is not surprising; these tests had been previously
evaluated in a European Food Safety Authority trial and two
of them, the ELISA and the EIA, gave similar diagnostic and
analytical sensitivity on small ruminant TSE samples (EFSA
Panel on Biological Hazards, 2009). Moreover, the three
BCTs were performed on identical tissue homogenates,
which undoubtedly contributed to the high correlation of
their results. Considering that two of them, ELISA and WB,
use PK digestion, while EIA relies on a polyanionic ligandaffinity and detects aggregated PrPd, the high correlation
between the three BCTs argues for an association between
protease resistance and aggregation state.
In conclusion, PrPd/PrPres accumulating in the brains of sheep
infected with scrapie or BSE, as detected by IHC and BCTs, is
still a useful surrogate marker for infection or diagnosis of
clinical disease. However, in quantitative terms – and at least
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
Journal of General Virology 93
TSE infectivity versus laboratory tests
(a)
(b)
(c)
(d)
(f)
(e)
Fig. 4. IHC appearance of PrPd in samples with different infectious titres. Some samples show correlation between magnitude
of PrPd and titre: (a) (BSE1 Cc) and (c) (Scra1 St) show little PrPd (1.2 and 1.0, respectively) and low infectivity (log titres of 6.5
and 6.7, respectively); (b) (BSE1 St) and (d) (BSE6 Cc) show moderate/high PrPd levels (9.5 and 10.4, respectively) and high
infectious titres (log titre ¢8.7 in both cases). Other samples, however, do not show such correlation: (e) (BSE3 Ob) and (f)
(Scra3 Cc) show similar PrPd magnitudes (9.1 and 9.7, respectively) to samples illustrated in (b) and (d) but their infectivity is
low (log titre56.7 in both cases); despite identical infectious titre, note also the difference in intraneuronal PrPd, which is notably
more prominent in BSE3 Ob (e) than in Scra3 Cc (f; for values refer to Table 2). IHC with R145 PrP antibody; magnifications:
(a), (d), (f) ¾10; (b), (c) ¾20, (e) ¾4.
within the context of the TSE sources and host PRNP
genotypes used in this study – it does not necessarily reflect the
extent of replication of the infectious agent, so that infectious
titres cannot be predicted from laboratory test results.
METHODS
Animals and samples. Tissue samples from 20 sheep from two
different, unrelated studies were used in the present study. These were:
(i) 10 ARQ/ARQ Suffolk sheep clinically affected with natural scrapie,
originating from a closed flock [Moredun Research Institute (MRI)].
Details of the clinico-epidemiological aspects of the infection and of the
pathological and PrPd IHC characteristics of the disease in this flock
have been extensively documented elsewhere (Jeffrey et al., 2001b;
González et al., 2002, 2010b, 2012; Sisó et al., 2010). Nine of the 10
sheep selected for this study were born in 2002 and one in 2000; all of
them (seven females and three males) had reached clinical end-point of
http://vir.sgmjournals.org
scrapie at 702–860 days of age and all showed indistinguishable PrPd
profiles by brain IHC. (ii) Seven ARQ/ARQ and three ARR/ARR
Romney sheep intracerebrally inoculated with a sheep-derived BSE
inoculum (González et al., 2007). Three, three and one of the ARQ/
ARQ sheep (five males and two females) had been dosed with 1023,
1024 and 1025 dilutions of the original brain homogenate, respectively,
and had reached clinical end-point of laboratory-confirmed BSE at
476–562, 499–604 and 828 days post-infection (p.i.), respectively. The
three ARR/ARR sheep (two females and one male) had been inoculated
with a 1023 dilution of the same brain homogenate and reached BSE
clinical end-point at 1414–1710 days p.i.
None of the 17 ARQ/ARQ sheep showed polymorphisms at any
codon of the PRNP gene.
From each of the 20 sheep, samples of frontal cerebral cortex, corpus
striatum, cerebellum and medulla oblongata at the obex from the left
hemi-brain were fixed in formaldehyde and processed for IHC examinations. Tissue samples from the same areas of the right hemi-brain
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
2525
L. González and others
(except for obex, which was not available for the seven BSE-affected ARQ/
ARQ sheep) were kept frozen at 280 uC for BCT analyses. Therefore, a
total of 73 brain samples, 40 from scrapie- and 33 from BSE-affected sheep,
were available for comparison between IHC features and BCT results.
Eleven of the 73 brain samples, three from scrapie and eight from
BSE-affected sheep, were selected for bioassay in TgshpXI transgenic
mice, which overexpress ovine ARQ/ARQ PrP (Kupfer et al., 2007).
The selection of the samples was made on the basis of differences in
magnitude of PrPd accumulation observed by IHC (scores from 1.0 to
16.7) and/or in the proportion of different morphological and cellassociated PrPd types.
IHC and BCT in sheep brain samples. The processing of brain
samples for IHC, which was performed with PrP antibody R145
(AHVLA), and the scoring system for quantification of PrPd accumulation have been described in detail previously (González et al., 2002,
2005a). Briefly, the magnitude of different types of disease-associated
(PrPd) aggregates was scored from 0 to 3 in the 73 samples included
in this study. To facilitate comparisons with BCT and infectivity
titration results, the PrPd types were grouped as follows: (i)
intracellular (ITCL): intra-neuronal, intra-microglial and intraastrocytic, for a maximum score of 9; (ii) stellate (STEL) for a
maximum score of 3; (iii) astrocyte-associated (ASTR): subpial, perivascular, subependymal and peri-vacuolar, for a maximum score of
12; (iv) grey matter neuropil-associated (NRPL): particulate/
coalescing, linear and peri-neuronal, for a maximum score of 9.
The only other type of PrPd observed in the samples examined were
non-vascular plaques, exclusively found in the three BSE-infected
ARR/ARR sheep; this type is not considered in the analyses of
individual PrPd types but was included in the score for total PrPd in
each sample, which was worked out as the sum of scores of the
different types considered (maximum theoretical score of 36). Since
the IHC protocol did not include a protease-treatment step, PrPd is
putatively composed of both protease-resistant and protease-sensitive
forms of abnormal, disease-associated PrP (González et al., 2005b).
Three BCTs were applied to the 73 selected brain samples. Two of the
tests, the TeSeE ‘sheep and goat’ (Bio-Rad) and the Bio-Rad TeSeE
‘sheep and goat’ Western blot analysis (WB), are based on
immunoaffinity and relative PK resistance at two different PK
concentrations (fivefold lower in the ELISA than in the WB), thus
detecting PrPres. The other, IDEXX HerdChek BSE/scrapie enzyme
immunoassay (EIA; Idexx Laboratories), is a polyanionic ligandaffinity based method and does not require PK treatment, thus
detecting PrPd. All three BCTs enable quantification of the abnormal
PrP in the samples by means of end-point dilution assays, the results of
which were used for comparison with the IHC scores and the titration
results in the bioassay. Dilution curves for each sample were achieved
by serially diluting samples 1 : 2 to give a dilution range from neat to
1 : 32 768. ELISA and WB samples were diluted post-extraction using
sample diluent (R6) and Laemmli solution, respectively and then
analysed according to manufacturer’s instructions. EIA samples were
diluted using de-ionized water prior to the addition of working plate
diluent; once diluted all were analysed following manufacturer’s
instructions with diluted samples treated exactly the same as neat
samples. We had previously shown that there were no noticeable
differences between EIA dilution curves prepared using negative
control brain homogenate [25 % (w/v) in kit homogenization buffer]
and water (data not shown). End-point dilutions were given as the
dilution at which the absorbance reading for the test sample was the
same as that for the test kit negative control (ELISA and EIA) or the last
dilution at which PrPres bands were observed (WB).
selected because of Prnp genotypic homology with most of the samples
inoculated; conventional wild-type mice were disregarded in view of
previous evidence of inefficient transmissibility of scrapie isolates from
the MRI Suffolk flock to such mice (Bruce et al., 2002). Mice reaching
clinical end-point of neurological disease compatible with scrapie/BSE,
and those showing significant mobility deterioration or inability to eat
or drink, were euthanized with carbon dioxide. Brains were removed
and placed in buffered formalin and confirmation of scrapie/BSE was
carried out by IHC for PrPd with rabbit polyclonal R486, which
recognizes amino acid sequence 221–233 of bovine PrP, by procedures
described elsewhere (Beck et al., 2010). All mouse procedures were
performed in compliance with the Animal (Experimental Procedures)
Act 1986 under licence no. 70/6310 issued by the UK Home Office and
were approved by the local ethics committee.
Infectious titres (LD50) were determined by the Spearman–Karber
formula (Hamilton et al., 1977): log10LD505(X02[d/2])+d(Sri/ni),
where for this case:
X05log10 of the lowest dilution at which all mice died of scrapie or
BSE
d5dilution factor
ri5number of positive mice in the X0 and subsequent dilutions
ni5number of mice inoculated in the X0 and subsequent dilutions
(discounting those dying from intercurrent deaths)
Since 20 ml of inocula were used for the intracerebral inoculations and
the LD50 is expressed by gram of tissue, a conversion factor of 6500
was applied to the titres obtained.
Statistical analyses. The relationship between the results of the
three BCTs on individual samples (end-point dilution values) was
assessed by means of Spearman’s non-parametric correlation analyses
(Instat GraphPad Software). The same approach was used to compare
between individual sample results in IHC and in each of the three
BCTs. These analyses were done for the total number of samples
(n540 in scrapie, n521 in ARQ/ARQ BSE and n512 in ARR/ARR
BSE samples) regardless of brain area of provenance, and also for each
individual area separately. The latest were not performed on ARR/
ARR BSE samples due to the low number of sheep involved (n53).
The relationship between infectivity and laboratory test results was
assessed by Mann–Whitney non-parametric analysis applied to
samples grouped according to their titre. This strategy was necessary
in view of the bi-modal distribution of the infectious titres obtained
in the 11 samples selected for bioassay.
ACKNOWLEDGEMENTS
This study was funded by Defra project SE2007, with sheep samples
used for bioassay and laboratory tests coming from projects SE1842
(BSE) and SE1949 (scrapie), also funded by Defra. The authors are
indebted to L. Fairlie, A. Dunachie and M. Oliva (AHVLA-Lasswade)
for technical support on sheep and mouse immunohistochemistry and
to the staff in the Animal Service Unit (ASU-Weybridge) and in the
Moredun Research Institute for mouse and sheep up-keeping and postmortems, respectively. We also thank Martin Groschup (FriedrichLoeffler Institute) for provision of TgshpXI mice and to Jim Hope
(AHVLA) for critical and constructive reading of the manuscript.
REFERENCES
Bioassay in TgshpXI mice. Groups of 10 TgshpXI mice were each
Ayers, J. I., Schutt, C. R., Shikiya, R. A., Aguzzi, A., Kincaid, A. E. &
Bartz, J. C. (2011). The strain-encoded relationship between PrPSc
inoculated intracerebrally with 1021, 1023 and 1025 dilutions of each
of the selected 11 sheep brain samples. This transgenic mouse line was
replication, stability and processing in neurons is predictive of the
incubation period of disease. PLoS Pathog 7, e1001317.
2526
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
Journal of General Virology 93
TSE infectivity versus laboratory tests
Balkema-Buschmann, A., Eiden, M., Hoffmann, C., Kaatz, M.,
Ziegler, U., Keller, M. & Groschup, M. H. (2011). BSE infectivity in
Head, M. W., Knight, R., Zeidler, M., Yull, H., Barlow, A. & Ironside,
J. W. (2009). A case of protease sensitive prionopathy in a patient in
the absence of detectable PrPSc accumulation in the tongue and nasal
mucosa of terminally diseased cattle. J Gen Virol 92, 467–476.
the UK. Neuropathol Appl Neurobiol 35, 628–632.
Barron, R. M. & Manson, J. C. (2003). A gene-targeted mouse model of P102L
Gerstmann-Sträussler-Scheinker syndrome. Clin Lab Med 23, 161–173.
accumulation of PrPres in murine ME7 scrapie in relation to pathological
and PrP immunohistochemical changes. J Comp Pathol 124, 20–28.
Barron, R. M., Campbell, S. L., King, D., Bellon, A., Chapman, K. E.,
Williamson, R. A. & Manson, J. C. (2007). High titers of transmissible
Jeffrey, M., Martin, S., Thomson, J. R., Dingwall, W. S., BegaraMcGorum, I. & González, L. (2001b). Onset and distribution of tissue
spongiform encephalopathy infectivity associated with extremely low
levels of PrPSc in vivo. J Biol Chem 282, 35878–35886.
PrP accumulation in scrapie-affected suffolk sheep as demonstrated by
sequential necropsies and tonsillar biopsies. J Comp Pathol 125, 48–57.
Beck, K. E., Sallis, R. E., Lockey, R., Simmons, M. M. & Spiropoulos, J.
(2010). Ovine PrP genotype is linked with lesion profile and
Jeffrey, M., González, L., Chong, A., Foster, J., Goldmann, W., Hunter,
N. & Martin, S. (2006). Ovine infection with the agents of scrapie
immunohistochemistry patterns after primary transmission of classical
scrapie to wild-type mice. J Neuropathol Exp Neurol 69, 483–497.
Bruce, M. E., Boyle, A., Cousens, S., McConnell, I., Foster, J.,
Goldmann, W. & Fraser, H. (2002). Strain characterization of natural
sheep scrapie and comparison with BSE. J Gen Virol 83, 695–704.
Caughey, B., Raymond, G. J., Kocisko, D. A. & Lansbury, P. T., Jr
(1997). Scrapie infectivity correlates with converting activity, protease
resistance, and aggregation of scrapie-associated prion protein in
guanidine denaturation studies. J Virol 71, 4107–4110.
EFSA Panel on Biological Hazards (2009). Scientific opinion on
analytical sensitivity of approved TSE rapid tests. EFSA Journal 7, 1436.
Gambetti, P., Dong, Z., Yuan, J., Xiao, X., Zheng, M., Alshekhlee, A.,
Castellani, R., Cohen, M., Barria, M. A. & other authors (2008). A
novel human disease with abnormal prion protein sensitive to
protease. Ann Neurol 63, 697–708.
González, L., Martin, S., Begara-McGorum, I., Hunter, N., Houston,
F., Simmons, M. & Jeffrey, M. (2002). Effects of agent strain and host
genotype on PrP accumulation in the brain of sheep naturally and
experimentally affected with scrapie. J Comp Pathol 126, 17–29.
González, L., Martin, S. & Jeffrey, M. (2003). Distinct profiles of PrPd
immunoreactivity in the brain of scrapie- and BSE-infected sheep:
implications for differential cell targeting and PrP processing. J Gen
Virol 84, 1339–1350.
Jeffrey, M., Martin, S., Barr, J., Chong, A. & Fraser, J. R. (2001a). Onset of
(CH1641 isolate) and bovine spongiform encephalopathy: immunochemical similarities can be resolved by immunohistochemistry.
J Comp Pathol 134, 17–29.
Jeffrey, M., McGovern, G., Sisó, S. & González, L. (2011). Cellular
and sub-cellular pathology of animal prion diseases: relationship
between morphological changes, accumulation of abnormal prion
protein and clinical disease. Acta Neuropathol 121, 113–134.
Jeffrey, M., McGovern, G., Chambers, E. V., King, D., González, L.,
Manson, J. C., Ghetti, B., Piccardo, P. & Barron, R. M. (2012).
Mechanism of PrP-amyloid formation in mice without transmissible
spongiform encephalopathy. Brain Pathol 22, 58–66.
Konold, T., Bone, G. E., Phelan, L. J., Simmons, M. M., González, L.,
Sisó, S., Goldmann, W., Cawthraw, S. & Hawkins, S. A. C. (2010).
Monitoring of clinical signs in goats with transmissible spongiform
encephalopathies. BMC Vet Res 6, 13.
Kupfer, L., Eiden, M., Buschmann, A. & Groschup, M. H. (2007).
Amino acid sequence and prion strain specific effects on the in vitro
and in vivo convertibility of ovine/murine and bovine/murine prion
protein chimeras. Biochim Biophys Acta 1772, 704–713.
Lasmézas, C. I., Deslys, J. P., Robain, O., Jaegly, A., Beringue, V.,
Peyrin, J. M., Fournier, J. G., Hauw, J. J., Rossier, J. & Dormont, D.
(1997). Transmission of the BSE agent to mice in the absence of
detectable abnormal prion protein. Science 275, 402–404.
McKinley, M. P., Bolton, D. C. & Prusiner, S. B. (1983). A protease-resistant
protein is a structural component of the scrapie prion. Cell 35, 57–62.
González, L., Martin, S., Houston, F. E., Hunter, N., Reid, H. W.,
Bellworthy, S. J. & Jeffrey, M. (2005a). Phenotype of disease-associated
Miyazawa, K., Emmerling, K. & Manuelidis, L. (2011). High CJD
PrP accumulation in the brain of bovine spongiform encephalopathy
experimentally infected sheep. J Gen Virol 86, 827–838.
infectivity remains after prion protein is destroyed. J Cell Biochem
112, 3630–3637.
González, L., Terry, L. & Jeffrey, M. (2005b). Expression of prion protein
Mould, D. L., Dawson, A. M. & Smith, W. (1967). Determination of the
in the gut of mice infected orally with the 301V murine strain of the
bovine spongiform encephalopathy agent. J Comp Pathol 132, 273–282.
dosage-response curve of mice inoculated with scrapie. J Comp Pathol
77, 387–391.
González, L., Chianini, F., Martin, S., Sisó, S., Gibbard, L., Reid, H. W.
& Jeffrey, M. (2007). Comparative titration of experimental ovine BSE
Piccardo, P., Manson, J. C., King, D., Ghetti, B. & Barron, R. M. (2007).
infectivity in sheep and mice. J Gen Virol 88, 714–717.
Accumulation of prion protein in the brain that is not associated with
transmissible disease. Proc Natl Acad Sci U S A 104, 4712–4717.
González, L., Martin, S., Hawkins, S. A. C., Goldmann, W., Jeffrey, M. &
Sisó, S. (2010a). Pathogenesis of natural goat scrapie: modulation by
Rohwer, R. G. (1991). The scrapie agent: ‘‘a virus by any other name’’.
host PRNP genotype and effect of co-existent conditions. Vet Res 41, 48.
Safar, J. G., Geschwind, M. D., Deering, C., Didorenko, S., Sattavat, M.,
Sanchez, H., Serban, A., Vey, M., Baron, H. & other authors (2005).
González, L., Sisó, S., Monleón, E., Casalone, C., van Keulen, L. J. M.,
Balkema-Buschmann, A., Ortiz-Peláez, A., Iulini, B., Langeveld, J. P. M.
& other authors (2010b). Variability in disease phenotypes within a
single PRNP genotype suggests the existence of multiple natural sheep
scrapie strains within Europe. J Gen Virol 91, 2630–2641.
González, L., Jeffrey, M., Dagleish, M. P., Goldmann, W., Sisó, S., Eaton,
S. L., Martin, S., Finlayson, J., Stewart, P. & other authors (2012).
Susceptibility to scrapie and disease phenotype in sheep: cross-Prnp
genotype experimental transmissions with natural sources. Vet Res 43, 55.
Curr Top Microbiol Immunol 172, 195–232.
Diagnosis of human prion disease. Proc Natl Acad Sci U S A 102, 3501–3506.
Silveira, J. R., Raymond, G. J., Hughson, A. G., Race, R. E., Sim, V. L.,
Hayes, S. F. & Caughey, B. (2005). The most infectious prion protein
particles. Nature 437, 257–261.
Sisó, S., Jeffrey, M. & González, L. (2009). Neuroinvasion in sheep
transmissible spongiform encephalopathies: the role of the haematogenous route. Neuropathol Appl Neurobiol 35, 232–246.
Hamilton, M. A., Russo, R. C. & Thurston, R. V. (1977). Trimmed
Sisó, S., Jeffrey, M., Martin, S., Chianini, F., Dagleish, M. P. &
González, L. (2010). Characterization of strains of ovine transmissible
Spearman-Kärber method for estimating median lethal concentrations in toxicity bioassays. Environ Sci Technol 11, 714–719.
spongiform encephalopathy with a short PrPd profiling method.
J Comp Pathol 142, 300–310.
http://vir.sgmjournals.org
Downloaded from www.microbiologyresearch.org by
IP: 88.99.165.207
On: Wed, 12 Jul 2017 16:26:34
2527