Detection of classical and atypical/Nor98 scrapie by the paraffin-embedded tissue blot method

The paraffin-embedded tissue (PET) blot method was used to investigate sections of the central nervous system and lymphatic tissues from 24 cases of classical scrapie and 25 cases of atypical/Nor98 scrapie in sheep and four healthy control sheep. The PET blot detected deposits of PrPSc in the brain tissue of all 49 sheep with scrapie but no PrPSc labelling could be detected in the control sheep. By contrast, not all the atypical/Nor98 scrapie cases were detectable by immunohistochemistry. The high sensitivity of the PET blot method made it possible to observe that in some atypical/Nor98 cases, deposits of PrPSc may be restricted to supratentorial brain structures and that the diagnosis may be missed when only testing the obex area, where deposits are common in classical scrapie, and the cerebellar structures, where deposits are considered to be common in atypical/Nor98 cases.

, and not least the choice of brain sections are of great importance. This paper describes the highly sensitive and specific topographical detection of disease-associated PrP in both classical and atypical/Nor98 scrapie by the paraffin-embedded tissue (PET) blot method.

Sheep material
Sections were cut from the central nervous system (CNS) and lymphatic tissues (tonsils and/or lymph nodes) of 49 scrapie-positive sheep and four clinically healthy sheep from scrapie-free flocks. The scrapie-positive group consisted of 19 German and five Norwegian sheep diagnosed with classical scrapie either by Western blot or immunohistochemistry analysis of brain tissue and/or of a tonsil biopsy, and one German and 24 Norwegian Nor98 cases that all displayed the characteristic small molecular fragment of approximately 12 kDa in Western blot analysis; their genotypes and breeds are listed in Table 1. All the Norwegian and German atypical/Nor98 scrapie sheep were single cases from different flocks, but the German classical cases came from three flocks and the Norwegian classical cases came from four flocks, all of which had more than one case of classical scrapie.

Histopathology
Samples of CNS tissue and lymphoid tissue were fixed in 4 per cent buffered formaldehyde (usually for approximately one week) and embedded in paraffin. Most of the German tissue blocks were also decontaminated with 98 per cent formic acid, followed by another 12 to 24 hours of fixa-deposition of PrP Sc detected by immunohistochemistry. Reports of similar scrapie cases, often referred to as atypical scrapie, soon followed all over Europe (Buschmann and others 2004a, b, De Bosschere and others 2004, Gavier-Widen and others 2004, Onnasch and others 2004, Orge and others 2004, Everest and others 2006, Konold and others 2006, Nentwig and others 2007, Dagleish and others 2008, the Falkland Islands (Epstein and others 2005) and North America (Cook 2007). In contrast to classical scrapie, sheep with an AHQ (Lühken and others 2004) and/or ARF 141 Q allele (Moum and others 2005) proved to be more susceptible to atypical/Nor98 scrapie, and genotypes with the ARR allele were also affected (EFSA 2005, Buschmann and others 2004b, Orge and others 2004). The retrospective diagnosis of a sheep in the UK with Nor98 in 1989 indicates that the disease had probably been present for a longer time (Bruce and others 2007) but might have been overlooked, possibly due to its occurrence as single cases in flocks others 2007, Benestad andothers 2008), the low levels of PrP Sc in the obex region (the standard region used for scrapie diagnosis) and its comparatively low resistance towards proteinase K (Buschmann and others 2004a), which is involved in many diagnostic procedures for recognising PrP Sc .
In general, PrP Sc can be detected using suitable commercially available ELISA kits, but to differentiate between the classical and atypical/ Nor98 forms of scrapie other methods such as Western blot or immunohistochemistry are required. To ensure the detection of atypical/ Nor98 scrapie by these methods, the choice of antibody (Gretzschel and others 2006), the tissue treatment, for example, by proteases, detergents and denaturants, its fixation time in formalin (Privat and

PET blot
The PET blot was performed as described in cattle and human beings by Schulz-Schaeffer and others (2000a, b). In brief, 1 to 3 µm sections of paraffin-embedded tissues were cut, placed on 0·45 µm nitrocellulose membranes (Bio-Rad) and dried for two days at 55°C. The membranes were deparaffinised, rehydrated stepwise and then treated with 250 µg/ml proteinase K (Sigma-Aldrich) overnight at 55°C in proteinase K buffer (10mM Tris-HCl, pH 7·8, 100mM NaCl and 0·1 per cent Brij). After washing the membranes with Tris-buffered saline containing 0·1 per cent Tween 20 (TBST), the proteins were denatured in 4M guanidine thiocyanate for 30 minutes and the membranes were washed again. Immunodetection was performed after preincubation for 45 minutes in 0·2 per cent casein in PBS containing 0·1 per cent Tween 20 (PBST). The primary antibodies used were the monoclonal antibodies (mAb) L42, P4 (R-Biopharm) and F89/160.1.5 (VMRD), each at a dilution of 1:5000 in TBST for an incubation time of 90 minutes. After washing in TBST, an alkaline phosphatase-coupled goat anti-mouse antibody (D0486; Dako) diluted 1:1000 in TBST was applied for 60 minutes. After thorough washing with TBST, the pH was adjusted by rinsing the blots twice with NTM (100mM Tris-HCl, pH 9·5, 100mM NaCl and 50mM MgCl 2 ). The antibody reaction was visualised by the formazan reaction using NBT/BCIP. The blots were dried and evaluated under a dissection microscope.

Immunohistochemistry
The Norwegian scrapie cases were all stained according to the standard protocol (slightly modified from Benestad and others 2003) used at the National Veterinary Institute of Oslo. In brief, the pretreatment con-sisted of immersion in 98 per cent formic acid for 15 minutes followed by hydrated autoclaving at 121°C in 0·01M citric acid (pH 6·1) for 30 minutes. A mild digestion in 4 µg/ ml proteinase K at 37°C for five minutes was performed and the sections were incubated overnight at 4°C with mAb F89/160.1.5 (VMRD) at a dilution of 1:2000, alone and together with the mAb 2G11 at a dilution of 1:200. For weakly stained sections, the additional antibodies mAb F99/97.6.1 (VMRD), R145 (Jan Langeveld, IDLO), SAF84 and 8G8 (J. Grassi, CEA, France) were used. A commercially available kit (K4005, Envision AEC; Dako) was used to enhance the immunolabelling.
In addition, the German cases, the control sheep and selected tissue blocks of each Norwegian scrapie case were immunostained in Göttingen by the following protocol.
Tissue sections (1 to 3 µm) were incubated with 50 µg/ml proteinase K (Sigma-Aldrich) in pK-buffer (10mM Tris-HCl, pH 7·8, 100mM NaCl and 0·1 per cent Brij) for 15 minutes at room temperature. The slides were treated seven times for three minutes with 10mM citric acid (pH 6) in a microwave oven at 700 W and incubated with 4M guanidine thiocyanate for 15 minutes. Blocking steps with 0·1 per cent hydrogen peroxide and 0·2 per cent casein in PBST were performed before slides were incubated with the respective primary mAb (P4, L42 or F89/160.1.5) diluted 1:500 in 0·02 per cent casein/TBS for 90 minutes. The secondary antibody (D0486; Dako) was applied at a dilution of 1:500 in TBS for 60 minutes. Between each step, the slides were rinsed thoroughly with TBS. The antibody reaction became visible upon incubation with neofuchsin chromogen as substrate and the slides were lightly counterstained with haematoxylin.

Results
The three mAbs (P4, L42 and F89/160.1.5) revealed identical deposition patterns of PrP Sc in affected structures of the classical and atypical/Nor98 scrapie cases in PET blots and immunohistochemistry, independently of the protocol used. Nevertheless, variations in the intensity of the immunostaining were recorded, especially with the PET blot method. The P4 monoclonal antibody provided the most sensitive PrP Sc detection in the atypical/Nor98 scrapie cases and became the antibody of choice for this method, whereas with immunohistochemistry, mAb F89 gave the best results. No PrP Sc deposition was visible in the brain and lymphatic tissue sections of the four healthy control sheep (Fig 1a). With the PET blot, PrP Sc immuno staining was visible in the brain tissue of all the classical and atypical/Nor98 scrapie sheep, but one case of atypical/Nor98 scrapie was consistently negative by immunohistochemistry (Fig 2a).
The following results refer to the PET blot unless indicated otherwise: the deposition patterns and distribution of PrP Sc in the brain corresponded basically with those described for immunohistochemistry in classical scrapie (van Keulen and others 1995) and atypical/Nor98 scrapie (Benestad and others 2008). In one classical case, discrete accumulations of PrP Sc in the dorsal motor nucleus of the vagus nerve (DMNV) and the solitary tract nucleus were visible by immunohistochemistry and stronger staining was obtained by the PET blot method (Figs 1b, 1e). This case was therefore considered to be an early classical case (van Keulen and others 2008). The other classical cases generally had intense staining of the brainstem and especially of the DMNV (Fig 1c). In contrast, with the PET blot no deposits of PrP Sc were detected in the DMNV of the atypical/Nor98 scrapie cases (Fig 1d). PrP Sc accumulation in the obex of the atypical/Nor98 scrapie cases was often found in the spinal trigeminal nucleus (Figs 1d, 1f), and granules of PrP Sc were frequently present

FIG 2: Cortex of a three-year-old Dala sheep (AFRQ/AFRQ) that was consistently negative by immunohistochemistry using different protocols and a panel of monoclonal antibodies (mAbs), as in (a) with mAb P4, but gave positive signalling with the PET blot and the same mAb (b).
Bar=500 µm PAPERS minimal PrP Sc deposits in the brainstem and cerebellum by immunohistochemistry, but strong PrP Sc labelling in the cerebrum (Nentwig and others 2007), supports the importance of the present findings with respect to the location sampled for rapid testing. The animal in this study to which this observation refers was only three years old and was the youngest of the Nor98/atypical scrapie cases, suggesting that it might have been an early case of atypical/Nor98 scrapie. It was diagnosed by Western blot, which showed the characteristic Nor98 scrapie profile, and the isolate was also successfully transmitted to tg388 mice (Le Dur and others 2005), but immunohistochemistry did not confirm these results. The PET blot method can be considered to be a valuable tool for investigating sheep scrapie, because disease-associated PrP deposits can be detected in classical and atypical/Nor98 scrapie cases, even at low levels, so providing the high specificity necessary for diagnosis and research.
in the white matter, especially in the pyramidal and cerebellar tracts. This distribution corresponds with previous reports (Benestad and others 2008). Among the 15 atypical/Nor98 scrapie cases from which brainstem material was available, only one case had no PrP Sc in the obex region. Sections of spinal cord were available from some of the atypical/Nor98 scrapie sheep, and several of them showed discrete signals in the substantia gelatinosa and in the dorsolateral tract. The substantia gelatinosa and marginal cells were also the most intensely stained structures in spinal cord samples from advanced cases of classical scrapie, but in contrast with the atypical/Nor98 scrapie cases, all the grey and white matter structures were affected, including sympathetic chain ganglia. The cerebellar cortex in the atypical/Nor98 sheep usually had intense deposits of PrP Sc . In the sheep with no PrP Sc in the obex, only a faint local signal in the granular layer of the cerebellum was observed, which could easily have been missed. This case was analysed by immunohistochemistry using different protocols and a panel of mAbs (P4, L42, F89, F99, 2G11, 8G8 and R145), but was repeatedly found to be negative in all the brain sections. In contrast, marked PrP Sc labelling was visible in the cerebral cortex (Fig  2b), and there was less intense staining in the basal ganglia and thalamus with the PET blot using mAb P4.
In all the classical cases of scrapie, positive staining was visible in lymphatic tissues (in two cases, no lymphatic tissue was available), as shown for a tonsil in Fig 3. However, no PrP Sc deposits could be found in the lymphatic tissues of any of the 16 atypical/Nor98 sheep for which lymphatic tissues were available.

Discussion
In this study, the PET blot method was applied to sections of CNS and lymphatic tissues of sheep with classical and atypical/Nor98 scrapie. The results support those previously reported for typical and atypical scrapie cases (EFSA 2005), but in addition show that the method is highly sensitive and specific for both types of scrapie. It has already proved to be a useful tool for studies of different TSEs, for example, cattle BSE (Schulz-Schaeffer and others 2000a), BSE in C57B1/6 mice (Lezmi and others 2006), experimental scrapie in hamsters (McBride and others 2001, Thomzig and others 2004), human TSEs (Schulz-Schaeffer and others 2000b, Peden and others 2006), and for the detection of PrP Sc in the peripheral tissues of sheep diagnosed with classical scrapie (Andréoletti and others 2004, Thomzig and others 2007, Lacroux and others 2007. To the authors' knowledge, this is its first application to tissues of atypical/Nor98 scrapie cases. The high sensitivity of the PET blot method made it possible to observe that in some of the atypical/Nor98 cases, the deposits of PrP Sc may have been restricted to supra tentorial (cerebral) brain structures; PrP Sc is not necessarily seen in the obex region, as in classical scrapie, and more importantly, may not be visible in cerebellar structures, as is considered to be common for atypical/Nor98 scrapie. The case of a Swiss goat diagnosed with atypical scrapie that had only