No evidence of the susceptibility of chickens to BSE emerged from these primary inoculation or sub-passage experiments. Kaplan-Meier survival data showed no significant differences between control and treatment groups. No pathology, which in mammals would have significance in relation to TSE, was identified in any of the chickens. Although there have been no substantiated reports of naturally occurring avian prion disease, Schoon et al.  described three cases of interest in red-necked ostriches (Struthio camelus) in two zoos in northern Germany, in 1986, 1988 and 1989. These birds, and two subsequent cases in 1992-1993, displayed progressive clinical signs of a nervous disorder with ataxia, changes to balance, uncoordinated movements during feeding and had vacuolar changes in the brainstem, but transmission studies failed to establish the nature of the disorder .
The two clinically defined syndromes observed in the present study appear to be more of interest in terms of the husbandry and experimental circumstances than of significance to the study aims.
The narcolepsy-like behaviour seemed to be associated with the approach of food satiety, but the behaviour was not observed in the sub-passage study when the housing was a free range floor system, suggesting that prolonged cage confinement in the primary study may have played a role.
Although in the primary study the MDS was observed only in exposed chickens, insufficient male control chickens survived to the terminal kill to determine whether there was a statistically significant correlation between exposure to BSE and the development of MDS, or whether MDS was a disorder peculiar to the males of the strain of chicken used in the experiment.
Similar signs to those of the MDS observed in the primary exposure study also occurred in each group in the sub-passage study, but the more frequent occurrence of such cases in the chickens that received tissues from the saline control chickens of the primary study than recipients of tissues from BSE exposed chickens, argues strongly against any relationship to exposure to the BSE agent.
Neither the feeding associated narcolepsy nor MDS have previously been reported in commercially farmed domestic chickens. However, these syndromes were observed in circumstances that were themselves unusual. The lifespan of commercially farmed chickens is usually approximately 45 days (broiler chickens) to 18 months (laying chickens) and their behaviour is not scrutinized in detail. Since neither of the syndromes or any other intercurrent diseases were observed in chickens less than 2 years of age it is possible that the observed behavioural abnormalities could be inherent to this strain of chickens without having been reported previously.
No lesions were observed in the central or peripheral nervous system that were significant either in terms of transmission of BSE or the clinical neurological signs observed. The vacuolation observed in the central nervous system of both the chickens and mice is most probably age and or host strain related.
Using biochemical extraction and PK digestion techniques, identical or similar to the one used in this study, SAF have been detected in a wide range of prion diseases, both natural and experimental [2, 5, 16, 32–41]. SAF are considered to be aggregates of the abnormal prion protein and, prior to the routine introduction of immunochemical methods for the detection of the protein, were the most studied pathological marker for the TSE . The detection of SAF in the diagnosis of mammalian TSE has been shown to be a relatively insensitive approach but unlike immunochemical methods, provides morphologic detection of the extracted altered protein, independent of the problems of specificity of immunochemical detection in a species for which reagents and appropriate control materials are lacking. Therefore, not withstanding sensitivity issues, negative results by this method in the present study might be considered to give more definitive information on the absence of significant accumulations of an altered form of the prion protein in central nervous tissue.
No disease-specific immunolabelling was seen with any of the antibodies used, including 6C2 and R568. The widespread punctate immunolabelling observed appear related to technical factors which are largely irresolvable in the absence of species and disease specific antibodies, positive control material and appropriately developed epitope demasking procedures.
In the present study mammalian derived PrP antibodies were not able to detect abnormal forms of PrP in the neural tissues of the chicken. Even if it were assumed that there was successful uptake of infectivity into nervous tissues there are a number of possible reasons for the failure of the mammalian abnormal prion protein to initiate a disease process.
The three-dimensional structures of mammalian and chicken PrPc are quite similar although there is only ~30% sequence identity [43–45]. The C-terminal domain of mammalian PrPc forms a globular domain with a unique fold which consists of three α-helices and a short, anti-parallel β-sheet. Chicken PrPc has a number of additional structural elements not found in mammalian PrPc: a 310 helix between helices 2 and 3, an insertion between helices 2 and 3 which forms a flexibly disordered loop containing a glycosylation site and elongation of the N-terminal end of helix 3 . The mechanism of conversion of PrPc to PrPSc is not known but it is likely that it involves close range interactions between the two molecules. The additional structural elements present in chicken PrPSc may inhibit or prevent this interaction. For example, changing the charge distribution at the protein's surface as occurs in the Gly200Lys mutation associated with familial Creutzfeldt-Jakob disease , which slows or prevents the propagation of PrPd.
The key difference between the structures of the PrPc and PrPSc isoforms is the relative proportions of α-helices and β-sheets (reviewed in ). Dima and Thirumali  showed that the amino acid sequence of helix 2 of chicken PrPc has a higher propensity to form α-helices than the same region of mammalian (mouse) PrPc. This suggests that for chicken PrPc the transition from α-helix to β-sheet may not occur as readily as it does in mammals, or may not occur at all. Therefore, even if bovine PrPSc was able to engage in close range interactions with chicken PrPc, it may not be able to induce the structural transformation which is necessary for formation of chicken PrPSc.
Even assuming successful conversion of chicken PrPc into PrPSc there is evidence that it lacks key metal-binding sites which may render it weakly- or non-pathogenic . In mammalian PrPc there is a high-affinity copper-binding site located around His96 [50–53]. This site was found to be highly conserved in mammals but absent in the non-mammalian species examined, including chicken . When copper or other divalent cations are absent or present at very low levels mammalian PrPc becomes more susceptible to proteolytic degradation and conversion efficiency is reduced (reviewed in ). If the affinity of chicken PrPc for copper is lower than in mammals it follows that chicken PrPSc would be more susceptible to proteolytic degradation.
Although in the present study the experimental models used involved same species sub-passage and wild type mouse tissue assay after primary exposure to the BSE agent, it is possible that the chicken, or other avian species, are susceptible to other TSE agents after parenteral exposure. In a study where chickens were intravenously inoculated with a TME agent passaged in mink  putatively small amounts of infectivity were recovered by mink bioassay from chicken lymphoreticular tissues sampled at 30 and 148 days p.i. but the chickens did not show any neurological signs or pathological changes in the brain. No other diagnostic studies of the chickens' tissues were performed. It is unclear whether this observation was related to uptake and persistence of inoculum, or replication of agent in the tissues.