Spatial and Temporal Patterns of Prion Gene Variation Are Consistent With a Response to Chronic Wasting Disease-Induced Selection in Wild White-Tailed Deer

Abstract

Chronic wasting disease (CWD) poses a threat to cervids and is increasingly prevalent throughout North America. Prion protein gene (PRNP) variation may confer some degree of genetic resilience, creating an impetus to examine changes in allelic variation and to assess signatures of selection. We investigated the association between CWD and PRNP variation in white-tailed deer (WTD) (Odocoileus virginianus) and mule deer (MD) (Odocoileus hemionus) sampled in Alberta, Canada between 2014 and 2017. We sequenced the PRNP gene of 575 WTD (67 CWD-positives) and 660 MD (202 CWD-positives) and detected 14 single nucleotide polymorphic loci in WTD and 8 in MD. No association was identified between the MD genetic variation and disease status. Notably, a variant at 286 was detected in WTD, resulting in an amino acid change at codon 96 (G96S). Genotype counts at this locus were significantly associated with CWD status, with the 96S allele under-represented among CWD-positive and over-represented among negative individuals. For a CWD-positive individual, the odds of being homozygous for the major allele (G96/G96) were more than sevenfold greater than being homozygous for the minor allele (96S/96S). Following additional sequencing of 1612 WTD, we examined spatial and temporal variability of this locus in association with the disease history on the landscape. Among females, the frequency of 96S varied negatively with the distance to where CWD was first detected. Additionally, the 96S allele frequency has increased over time, in line with expectations based on estimated disease selection coefficients. Our results are consistent with CWD selection pressure resulting in increasing frequency of the 96S allele in space and time, indicating it may confer some resilience and extended infection with CWD.

Keywords: PRNP; White‐tailed deer; cervid; chronic wasting disease; prion; transmissible spongiform encephalopathy.

FIGURE 1

White‐tailed deer (WTD) sampling within wildlife management units (WMUs; 2014–2017), as well as of chronic wasting disease (CWD) occurrence in WTD, in Alberta (Canada) from 2007 to 2016. The seven wildlife management units (WMUs) used for the subsampling are highlighted in yellow.

FIGURE 2

Genotype proportions for the G96S PRNP locus separated into CWD‐negative (black) and ‐positive (pink) individuals for white‐tailed deer in Alberta (Canada) sampled from 2014 to 2017.

FIGURE 3

Frequency of the 96S allele estimated for 35 wildlife management units (WMUs). Panels A, C, and E show trends for the 96S allele frequency among females, males, and all deer (both sexes), respectively, versus the smallest distance of the WMU from the two regions where CWD was first detected in kilometers (km). Panels B, D, and F show trends for the 96S allele frequency among females, males, and all deer, respectively, versus the number of years since CWD was first detected (zero years are WMUs where CWD had not been detected as of 2017).

FIGURE 4

Interpolated surface map of the 96S allele frequency in female white‐tailed deer across wildlife management units (WMUs) in Alberta, Canada. Colors represent the allele frequency gradient from low (yellow) to high values (dark blue). Interpolation was performed to estimate allele frequencies at unsampled locations.

FIGURE 5

Predicted change over time in the allele frequency for 96S in white‐tailed deer based on two models of gene action (“Additive” and “Dominant”), and four estimates of selection coefficients calculated in Haworth et al. (2021), and Robinson et al. (2012). The negative “−” and plus “+” signs denote the upper and lower estimates, respectively. Observed values are denoted by black squares – the starting allele frequency for this allele was obtained from Wilson et al. (2009), and the value for 2017 was estimated from 854 individuals we sampled in 2017.