During 2009–2010, we analyzed conformational stability of PrP^Sc from sheep TSE isolates by using a conformational stability and solubility assay (CSSA) that we developed (14). We showed that CSSA could reveal strain-specified PrP^Sc conformational stability in sheep isolates because it enabled discrimination of Nor98 from classical scrapie isolates (14). Scrapie isolates had GdnHCl_1/2 values (the concentration of guanidine hydrochloride able to dissolve half the insoluble PrP^Sc aggregates in a brain homogenate) of 2.0 mol/L–2.3 mol/L; Nor98 isolates were less stable (1.3–1.4 mol/L GdnHCl). We thus sought to determine the conformational stability of PrP^Sc aggregates (Technical Appendix) derived from CH1641 and BSE strains (Table 1), including one (TR316211) of the few CH1641-like field isolates described so far (10,12,13). Two other CH1641-like isolates (99–454 and 99–321) were found in a retrospective analysis of sheep scrapie cases in France.

Classical scrapie included as control displayed a GdnHCl_1/2 value (2.2 mol/L) in the range of previously analyzed isolates. CH1641 (provided by N. Hunter, Institute for Animal Health, Edinburgh, Scotland) and CH1641-like isolates showed conformational stabilities close to classical scrapie, with GdnHCl_1/2 values of 2.0–2.8 mol/L. In contrast, PrP^Sc from experimental sheep BSE (15) clearly showed higher conformational stability, with GdnHCl_1/2 values >3.8 mol/L (Table 1). These results suggest experimental sheep BSE might have a stronger resistance to denaturation than do most natural sheep scrapie isolates.

Because the discriminatory methods based on differential PrP^Sc N-terminal proteinase K (PK) cleavage (11) do not enable a clear-cut discrimination of CH1641-like from BSE (12), we investigated the potential of denaturation with GdnHCl as a further discriminatory strategy within the framework of the Istituto Superiore di Sanità discriminatory WB (11). To this aim, samples were untreated or treated with 3.5 mol/L GdnHCl before PK digestion and WB analysis with SAF84 and P4 monoclonal antibodies (Figure 1). This method was set up by analyzing representative scrapie, BSE, and CH1641 samples (Figure 1). As expected, BSE and CH1641 were poorly detected by P4, in contrast to classical scrapie. Treatment with 3.5 M GdnHCl, however, nearly abolished PK resistance of PrP^Sc from classical scrapie and CH1641, but not from sheep BSE, thus also enabling discrimination of CH1641 from BSE.

We then analyzed a larger set of samples (Table 2), including natural BSE in a goat (Figure 1). These experiments confirmed the higher resistance to denaturation of BSE samples, irrespective of the species, PrP genotype, and route of inoculation, compared with all other samples (Figure 1). When the antibody ratio and the denaturation ratio were measured and plotted as a scattergraph, classical scrapie, CH1641, and BSE isolates clustered into 3 distinct groups (Figure 2, panel A): 1) scrapie isolates displayed antibody ratios <2 and denaturation ratios were 0.02–0.13; 2) CH1641 samples had antibody ratios >2 and denaturation ratios were 0.06–0.29; and 3) BSE samples had antibody ratios >2, but denaturation ratios were >0.51.

Glycoform profiles, i.e., the relative proportion of diglycosylated, monoglycosylated, and unglycosylated PrP^res fragments, have also been reported as a discriminatory criterion for the identification of BSE in sheep (8–10), as well as when compared with CH1641 (8,10). With the Istituto Superiore di Sanità WB method (Figure 2, panel B), field scrapie isolates, including CH1641-like isolates, were characterized by a lower diglycosylated-to-monoglycosylated glycoform ratio (0.48:0.35–0.58:0.25) than sheep BSE (0.65:0.25–0.75:0.19) and the natural goat BSE (0.70:0.22).

Because the analysis of PrP^Sc from sheep prion isolates by CSSA showed an extremely high conformational stability of BSE samples, we improved the Istituto Superiore di Sanità discriminatory WB by including a pretreatment of brain homogenates with GdnHCl. Our results show that the combined use of 2 independent molecular features, N-terminal cleavage by PK and resistance to denaturation, could indeed differentiate classical scrapie and CH1641-like isolates from small ruminant BSE. Nonetheless, we observed some variability among the CH1641-like samples, either when analyzed by CSSA (Table 1) or by the discriminatory WB. As previously reported (12), the antibody ratios of some CH1641-like samples were close to the cutoff (Figure 2, panel A). Furthermore, the variable conformational stability observed by CSSA was also reflected in the denaturation ratios measured by discriminatory WB, with 2 CH1641-like samples showing a relatively higher resistance to GdnHCl than to all other scrapie samples. Because of the limited number of CH1641-like isolates, further studies are needed to evaluate their effective range of variability.

This variability may be disappointing for discriminatory purposes, but it may also hinder the possible presence of subtle PrP^Sc conformational (and possibly strain) variants in CH1641-like isolates. The biologic similarities of CH1641-like samples after transmission to ovine transgenic mice (13) and voles (U. Agrimi, unpub. data) were worth noting. Nevertheless, CH1641-like isolates induced a certain degree of PrP^Sc molecular variability in both rodent models (13; U. Agrimi, unpub. data), which might be related to the molecular variability in PrP^Sc extracted from sheep brain.

Although based on a limited set of samples, our study supports the notion that CH1641-like isolates can be convincingly discriminated from small ruminant BSE on molecular grounds. Furthermore, the high conformational stability of BSE, when compared with that in classical scrapie, Nor98, and CH1641-like isolates, suggests the potential of the new discriminatory WB here proposed for discriminating BSE from other known small ruminant TSEs.