Natural scrapie in sheep is associated with polymorphisms of the PrP gene, particularly at amino acid codons 136, 154 and 171. This paper reports the results of nine scrapie case-control studies in Bleu du Maine, Herdwick, Merino x Shetland, Poll Dorset, Scottish Halfbred, Shetland, Soay, Suffolk and Swaledale sheep from British flocks affected by scrapie. In some outbreaks, scrapie was found to occur only in animals with at least one PrP allele encoding valine at codon 136 (V136), usually a relatively rare allele in healthy controls. In other outbreaks, the V136, PrP allele was either not found or was not an absolute prerequisite for scrapie to develop. Although scrapie had a strong tendency to affect sheep with PrP genotypes homozygous for glutamine at codon 171 (QQ171), these genotypes (QQ171 but varying at other codon positions) were relatively common in healthy controls. The reliable prediction of scrapie susceptibility in previously uninvestigated sheep flocks will therefore require information at least about PrP genotypes at codons 136 and 171.

Page 2, text line 9 should read ' … for the record, this M r for type B is 20 200 (p800), for type A 24 000 (p1300) and for type C 22 700 (p1000). The glycoform ratios for each type were as follows : type B, %H 66, %L 27 ; type A, %H 15-50, %L 22-46 ; … ' IFB

In scrapie-affected sheep flocks, host PrP genotype plays a vital role in determining which sheep will succumb to scrapie and the incubation period. Consequently, within-flock scrapie dynamics is best understood within the context of the genotype profile of the flock. Here we describe a 17 month epidemic of scrapie in a commercially farmed flock of 230 genotyped Texel sheep. At the start of the study, 70 % of the sheep were of three genotypes only : ARR/ARQ, ARH/ARQ and ARQ/ARQ. Only 15 % of sheep encoded the disease-associated VRQ allele and only a single sheep (0n4 %) was of the most susceptible VRQ/VRQ genotype. For susceptible genotypes there was a marked deficit (P 0n025) of older animals ( 3 years), implying that some cases of scrapie had occurred previously. In the ensuing 17 months, 18 sheep of known genotype were confirmed positive for the disease : seven VRQ/ARQ, six VRQ/ARH, two VRQ/ARR, three ARQ/ARQ. Median ages at death were 2n7, 2n8, 4n2 and 3n8 years respectively. Mortality rates were 55, 86, 13 and 3 % respectively. Survival analysis revealed a highly significant effect of genotype on survivorship, but no difference between VRQ/ARQ and VRQ/ARH, or between VRQ/ARR and ARQ/ARQ. There was no difference in the survivorship of middle-and older-age cohorts of susceptible sheep. Scrapie risk group (as defined by PrP genotype) was not associated with submission as a scrapie suspect but later found to be negative, or with dying of unknown causes on the farm.

No abstract