Atypical scrapie was first identified in Norwegian sheep in 1998 [1], and has subsequently been identified in many European countries [2] and elsewhere around the world, including North America [3] and the Falkland Islands [4]. Retrospective studies have identified cases predating this initial identification [5, 6] and epidemiological studies have indicated that this form of scrapie does not conform to the behaviour of an infectious disease [7]. In many cases, atypical scrapie does not co-exist with classical scrapie (e.g. on individual farms, but also at a national level, such as in Portugal and the Falkland Islands [8]), although some field cases originate from flocks which have also reported classical scrapie, and it has occurred in a closed research flock with endemic classical disease [9]. It has also been reported in a research flock in which detailed biosecurity measures have been maintained, founded with sheep from a country free of classical scrapie [10]. This has led to proposals [2, 10, 11] that this form of scrapie is in fact a spontaneous disease which may have existed undetected for a long time.

Despite lack of evidence for infectivity in the epidemiological data, atypical scrapie isolates have been shown to be infectious experimentally, through intracerebral inoculation into ovinised transgenic mice [12] and sheep [13].

The interactions of host and infecting strain on ovine transmissible spongiform encephalopathies (TSEs) are known to be complex, and have a profound effect on the resulting phenotype of disease. Following experimental challenges distinct 'strain' characteristics can be identified [14], but when naturally-occurring disease is studied it is clear that host prion protein gene (PRNP) genotype can have a significant effect on the resulting phenotype [15], although a single genotype may support more than one phenotype [15, 16]. The preferred genotype 'target' range of atypical scrapie is different to that of classical scrapie [17, 18] but there is some overlap in susceptible genotypes.

In contrast to the widespread variation in pathology and immunopathology characteristics seen in classical scrapie [15], the pathology observed in atypical cases appears to be much more consistent, regardless of genotype [1, 19–21].

The first successful challenge of an AHQ/AHQ sheep with 'field' atypical scrapie from an homologous donor sheep was reported in 2007 [13]. This paper provides data on further homologous AHQ/AHQ transmissions, including sub-passage of the initial case, and compares this with similar data from field cases of classical scrapie, and experimental bovine spongiform encephalopathy (BSE) in the same genotype.

The scale of such studies, and the limited availability of both inoculum and recipient sheep precludes a study design which includes every genotype/strain/route of challenge combination. To enable us to assess the ability of a single PRNP genotype to support various TSEs with distinct pathological characteristics, data is presented for comparison from AHQ/AHQ animals orally challenged with BSE (cases 10-14) [22–24] and field cases of classical scrapie which have arisen naturally in AHQ/AHQ sheep (cases 15-17) [25, 26].

Animals challenged with field atypical scrapie succumbed to disease from 378 - 1057 days post challenge (mean 751 days), which compares with 126, 537 and 605 days for the one VRQ/VRQ and two ARQ/VRQ classical scrapie intracerebrally inoculated controls respectively.

Pathology and immunohistochemistry

Both the vacuolation profiles and immunopathological characteristics were very consistent within the animals challenged with atypical scrapie, and distinct from the other AHQ/AHQ animals where disease was attributable to other strains.

Full vacuolation profiles could only be obtained for four of the five primary transmission cases, and one sub-passage animal. Vacuolation (Figures 1a, b) in the atypical cases was absent in the brainstem, and most pronounced in the cerebellar and cerebral cortices, as compared to the BSE and classical scrapie cases in which brainstem lesions were predominant. In the thalamic and frontal areas the main difference between the classical scrapie and BSE profiles related to the relative intensity of vacuolation, which was greater in the classical scrapie cases.

Immunohistochemistry (IHC; Figure 2) shows that immunolabelling patterns are also distinct, with both the type of labelling and the neuroanatomical distribution being different for the atypical scrapie compared to the other strains. However, the atypical scrapie pattern remains the same on primary versus secondary passage (Figure 3), and is consistent with the IHC patterns described in natural disease [20].

Western immunoblotting (WB; Figure 4) shows that the banding pattern associated with atypical scrapie [1, 27] is present in all the atypical challenged cases, and is consistent regardless of whether it is primary or secondary passage (Figure 4). The blot of the field case which was examined in the context of a retrospective review of natural disease [28] was consistent with classical scrapie. The BSE cases were blotted within the remit of their source project (Chaplin and Bellworthy, unpublished data), and conform to the blot pattern (low molecular mass of the unglycosylated band with antibody 6H4 and a weak P4 antibody reaction) associated with experimental BSE in sheep.

No immunolabelling was seen in any lymphoid tissue from the atypical challenged sheep, despite widespread involvement of the equivalent tissues in the positive classical scrapie controls and the BSE-challenged sheep (Table 2).

Case number		1	2	3	4	5	18	19	20
Lateral retropharnyngeal lymph node		Neg	Neg	Neg	NA	Neg	Pos	Pos	Neg
Palatine tonsil		Neg	NA	NA	Neg	Neg	NA	NA	NA
Nictitating membrane		IT	IT	Neg	Neg	IT	Neg	IT	Neg
Spleen		Neg	Neg	Neg	Neg	Neg	Pos	Pos	Pos
Mesenteric lymph node		Neg	Neg	Neg	Neg	Neg	Pos	Pos	Pos
Distal ileum - Peyer's patches		Neg	Neg	Neg	Neg	Neg	Pos	Pos	Neg
Recto-anal mucosa associated lymphoid tissue (RAMALT)		Neg	Neg	Neg	Neg	Neg	Pos	Pos	Pos

Cases 1-5: atypical scrapie.

Cases 18-20: classical scrapie.

Neg = Negative; Pos = Positive; IT = insufficient lymphoid tissue to test; NA = sample not available.

This data confirms the experimental transmissibility of atypical scrapie, and the stability of disease phenotype - clinical, biochemical and pathological - on PRNP homologous experimental transmission in AHQ/AHQ animals. The mean incubation time of these cases (751 days) is approximately 50% of the average age at death in field cases with atypical scrapie [20]. This observation that intracerebral challenge shortens the incubation period is consistent with the substantial reductions in incubation period which are seen in the classical scrapie controls which are 90% (for VRQ/VRQ) and 60% (for VRQ/ARQ) less than the average age at onset of field cases (Ortiz-Pelaez, personal communication) in these genotypes.

There was a slight shortening of mean incubation period between the initial experimentally challenged sheep (790 days) and those which succumbed following sub-passage (695 days), but the group sizes are too small to draw any robust conclusions from this observation.

The wide range of incubation periods seen cannot be readily interpreted due to the very small numbers of animals in this study (a necessary restriction due to the very limited amount of suitable donor material, and limited number of recipients of a suitable genotype). It is interesting to note that the outliers (cases 1 and 2, Table 1) received the same inoculum. All donor inocula give similar incubation periods in ovinised transgenic (Tg338) mice when inoculated as part of a separate study (Spiropoulos, personal communication), although some difference in titre is likely to be masked by the sensitivity of these mice. In one titration of an atypical scrapie isolate the mean incubation period did not change more than 10% over the first three log dilutions (Spiropoulos and Simmons, unpublished data).

The absence of any visible lymphoreticular system (LRS) involvement in the experimentally challenged animals is consistent with what has been observed in natural cases of atypical scrapie [13] and cannot be attributed to the experimental route of exposure, since the classical scrapie controls challenged by the same route had widespread labelling of disease-associated PrP (PrP^Sc) in the lymphoid tissues.

The clinical presentation of atypical scrapie cases was different to classical scrapie or BSE in sheep: atypical scrapie did not appear to cause evident pruritus, whereas the scrapie and BSE cases described here usually presented with pruritus. Compulsive behaviour, such as circling, was only observed in atypical scrapie, and an impaired menace response was considerably more frequent in atypical scrapie cases. The difference in the clinical picture does not appear to be confined to the genotype or the route of inoculation since similar signs in the absence of pruritus have been observed in naturally affected atypical scrapie cases of various genotypes [1, 10, 29], whereas pruritus is frequent in naturally affected classical scrapie cases [26] or sheep orally challenged or naturally infected with the BSE agent under experimental settings [24]. Ataxia with hypermetria, head tremor and an absent menace response, which were predominantly observed in sheep with atypical scrapie, are signs indicative of a dysfunction of the cerebellum [30]. This concurs with the lesion and PrP^Sc immunolabelling pattern distribution in atypical scrapie, which is particularly prominent in the cerebellum. However, the lateralisation of clinical signs (circling to one side or a unilaterally deficient menace response) seen in atypical scrapie is unusual and to the authors' knowledge has not been described elsewhere in ovine BSE or classical scrapie. It is unlikely that the intracerebral route was responsible for the apparent lateralisation of clinical signs because it was not observed in ovine cases of BSE [24] and classical scrapie inoculated by the same route (cases 18-20, Table 1). Compulsive circling in the absence of vestibular signs like head tilt and nystagmus is suggestive of an asymmetric lesion of the forebrain, usually on the same side as the circling direction [31]. A unilateral absent menace response has been associated with a unilateral cerebellar lesion on the same side as the deficit [32]. This would suggest that the signs in case 2 (Table 1), which presented with a left absent menace response and circling to the left, was predominantly caused by a lesion of the left side of the brain. Unfortunately, the whole brain of this case was not available to evaluate symmetry of neuropathological changes, but there was no evidence of any pronounced asymmetrical lesion or PrP^Sc distribution in the bilateral sections examined in case 6, which also circled to the left.

This study shows that a single PRNP genotype, AHQ/AHQ can display a range of disease phenotypes which are consistent and readily distinguishable from each other, supporting the suggestion that such characteristics can be attributable, predominantly, to the agent strain. However, the relative over-representation of the AHQ genotype in the atypical naturally-affected populations supports the hypothesis that field strains have strong genotype 'tropisms' or that there is genotype selection of strains, as proposed by Spiropoulos [15].

The consistent disease phenotype seen in natural disease, primary experimental transmission and subsequent passage, supports the experimental inoculation model as a valid tool for the study of atypical scrapie, regardless of whether the disease in the field is acquired or spontaneous. It also enables the creation of atypical scrapie samples in a consistent manner for test evaluation and/or quality assurance purposes, helping to improve surveillance approaches for disease detection and confirmation in the field.