He GPI anchor (GPI2) [15]. The GPI2 PrPSc produced by these mice is fully infectious, lacks the GPI anchor, and is largely unglycosylated, which reduces the heterogeneity in the C-terminal portion of the molecule [15,16]. These properties make it ideal to carry out structural studies, and have allowed us to obtain, for the first time, a complete survey of the whole PrPSc sequence, regarding its susceptibility to proteolysis.all PK cleavage sites. This allowed us to analyze samples by Western blot (WB) and by MS. We analyzed our samples with high mass accuracy using nanoLC-ESI-Qq-TOF MS (Figure S4) and identified three peaks of 17148, 16728, and 16371 Da (peptides DprE1-IN-2 site G81-S232, G85-S232, and G89-S232). The smaller peptides were analyzed by MALDI-TOF. MS-based analysis revealed that the seven bands present in the WB (vide infra) contained thirteen peptides with MWs of 17148, 16726, 16371, 13606, 13463, 12173, 12041, 11171, 9687, 9573, 8358, 7436 and 6274 Da. By comparing the observed masses with those calculated from the mouse GPI- PrP sequence, we determined that they correspond to peptides G81-S232, G85-S232, G89-S232, A116-S232, 79831-76-8 site G118-S232, M133-S232, S134-S232, G141-S232, N152-S232, M153-S232, Y162-S232, S169-S232 and V179-S232 (Figure 2 and Table 1). No C-terminally truncated peptides were observed in our MS or WB-based analysis (vide infra).Identification of PK Cleavage Sites in GPI-anchorless PrPSc by Western BlotIn parallel we used Tricine-SDS-PAGE [17] followed by WB to analyze the PK-digested GPI- PrPSc (Figure 3). When the WB was probed with the antibody #51 (epitope G92-K100), just one wide band (,17 kDa) was observed, suggesting a set of cleavage products near G89 with no C-terminally truncated fragments. A blot probed with the W226 antibody (epitope W144-N152), revealed three additional faint 
23977191 bands (,14.6, 13 and 12 kDa), suggesting three PK cleavage sites between the epitopes of these antibodies. Probing with the C-terminal R1 antibody (epitope Y225-S230) revealed three more bands (,10.2, 8 and 6.7 kDa), suggesting three additional cleavage sites near residues Y149, P164 and V175. These bands agree quite well with our MS-based analysis (vide supra). In order to exclude the possibility that the observed PKresistant fragments are the result of the known preference of PK of certain amino acid residues, rather than structural constraints, we subjected a similar amount of freshly refolded, recombinant MoPrP to cleavage by PK. A concentration of PK much lower than that used with mouse GPI- PrPSc, 1 mg/ml, completely destroyed all PrP, leaving no PK-resistant fragments larger than 3.5 kDa (Figure S5). Only PK concentrations below 1 mg/ml yielded some partially resistant fragments, whose sizes do not match those of PK-treated GPI- PrPSc.Results Accumulation of PrPSc in GPI-anchorless MiceHomozygous GPI-anchorless PrP mice were 
inoculated at 6 weeks of age with the RML strain of murine-adapted scrapie. Three-hundred sixty-five days post-inoculation, the mice were humanely euthanized. Their brains were surgically removed for further biochemical processing. The presence of PrPSc was confirmed by digesting a portion of some of these brains, after suitable homogenization, with proteinase K (PK) and analyzing the result by Western blot (Figure 1A and S1). The PK treatment yielded the characteristic PK resistant core protein, referred to as PrP27-30 in PK-treated wild-type PrPSc, although in this case its apparent MW is lower, given the lack o.He GPI anchor (GPI2) [15]. The GPI2 PrPSc produced by these mice is fully infectious, lacks the GPI anchor, and is largely unglycosylated, which reduces the heterogeneity in the C-terminal portion of the molecule [15,16]. These properties make it ideal to carry out structural studies, and have allowed us to obtain, for the first time, a complete survey of the whole PrPSc sequence, regarding its susceptibility to proteolysis.all PK cleavage sites. This allowed us to analyze samples by Western blot (WB) and by MS. We analyzed our samples with high mass accuracy using nanoLC-ESI-Qq-TOF MS (Figure S4) and identified three peaks of 17148, 16728, and 16371 Da (peptides G81-S232, G85-S232, and G89-S232). The smaller peptides were analyzed by MALDI-TOF. MS-based analysis revealed that the seven bands present in the WB (vide infra) contained thirteen peptides with MWs of 17148, 16726, 16371, 13606, 13463, 12173, 12041, 11171, 9687, 9573, 8358, 7436 and 6274 Da. By comparing the observed masses with those calculated from the mouse GPI- PrP sequence, we determined that they correspond to peptides G81-S232, G85-S232, G89-S232, A116-S232, G118-S232, M133-S232, S134-S232, G141-S232, N152-S232, M153-S232, Y162-S232, S169-S232 and V179-S232 (Figure 2 and Table 1). No C-terminally truncated peptides were observed in our MS or WB-based analysis (vide infra).Identification of PK Cleavage Sites in GPI-anchorless PrPSc by Western BlotIn parallel we used Tricine-SDS-PAGE [17] followed by WB to analyze the PK-digested GPI- PrPSc (Figure 3). When the WB was probed with the antibody #51 (epitope G92-K100), just one wide band (,17 kDa) was observed, suggesting a set of cleavage products near G89 with no C-terminally truncated fragments. A blot probed with the W226 antibody (epitope W144-N152), revealed three additional faint 23977191 bands (,14.6, 13 and 12 kDa), suggesting three PK cleavage sites between the epitopes of these antibodies. Probing with the C-terminal R1 antibody (epitope Y225-S230) revealed three more bands (,10.2, 8 and 6.7 kDa), suggesting three additional cleavage sites near residues Y149, P164 and V175. These bands agree quite well with our MS-based analysis (vide supra). In order to exclude the possibility that the observed PKresistant fragments are the result of the known preference of PK of certain amino acid residues, rather than structural constraints, we subjected a similar amount of freshly refolded, recombinant MoPrP to cleavage by PK. A concentration of PK much lower than that used with mouse GPI- PrPSc, 1 mg/ml, completely destroyed all PrP, leaving no PK-resistant fragments larger than 3.5 kDa (Figure S5). Only PK concentrations below 1 mg/ml yielded some partially resistant fragments, whose sizes do not match those of PK-treated GPI- PrPSc.Results Accumulation of PrPSc in GPI-anchorless MiceHomozygous GPI-anchorless PrP mice were inoculated at 6 weeks of age with the RML strain of murine-adapted scrapie. Three-hundred sixty-five days post-inoculation, the mice were humanely euthanized. Their brains were surgically removed for further biochemical processing. The presence of PrPSc was confirmed by digesting a portion of some of these brains, after suitable homogenization, with proteinase K (PK) and analyzing the result by Western blot (Figure 1A and S1). The PK treatment yielded the characteristic PK resistant core protein, referred to as PrP27-30 in PK-treated wild-type PrPSc, although in this case its apparent MW is lower, given the lack o.