The coding region of PRNP
has been analysed in (i) paediatric and (ii) adult kuru sampled around the peak of the epidemic (Cervenakova et al. 1999
), (iii) recent elderly kuru patients with a long incubation time (Collinge et al. 2006
), (iv) healthy modern-day young Fore, (v) neighbouring linguistic groups of EHP with low or no documented kuru, (vi) elderly men and (vii) women who attended mortuary feasts but have not developed kuru, sampled in the past 15 years (Mead et al. 2003
). These groups may be stratified a priori
in terms of their likely susceptibility to kuru. The interpretation of kuru patients' susceptibility is complicated by uncertainty about the time of exposure, as mortuary feasts of deceased kuru patients took place over many years prior to the cessation of cannibalism in 1960. The mean incubation period of kuru has been estimated as approximately 12 years (Alpers 2008
). Childhood or adolescent kuru would therefore reflect the most susceptible group as the incubation time is limited (to around the mean or shorter) by the age that children were old enough to participate in mortuary feasts. Healthy elderly women with multiple exposures at mortuary feasts reflect the most resistant group as they have clearly documented exposure but have proven disease free over many decades. Healthy elderly men may also be expected to be resistant to kuru but to a lesser degree than elderly women given their lower exposure. Elderly recent kuru patients with very long incubation times, over 50 years in some cases, are also likely to show genetic resistance given that their incubation time is much longer than the mean. Adult kuru patients sampled at the peak of the epidemic are difficult to classify in this way as their incubation time is uncertain. Also interesting is the possibility that the Fore population adapted to the kuru epidemic by inflation in the frequency of genetic resistance factors. The healthy young modern populations in EHP can be ranked by their overall exposure to kuru to consider this possibility.
codon 129 genotypes in groups of varying exposure and disease status are shown in . A generality, based on cumulative evidence from human studies, is that individuals homozygous for PRNP
codon 129 are susceptible to prion diseases whereas those heterozygous are resistant. Additionally, however, the homozygous genotypes 129MM and 129VV may show differential susceptibility; for example, all patients with vCJD have been 129MM, whereas early patients with iatrogenic CJD in the UK are particularly associated with codon 129VV (Collinge et al. 1991
). The molecular basis for this susceptibility is complex, invoking the efficiency of heterologous protein–protein interactions, and conformational selection (Collinge 1999
; Hill & Collinge 2003
; Collinge & Clarke 2007
): whether the infecting prion strain is a preferred or permissible conformation for PrP with 129M or 129V (Collinge et al. 1996
; Wadsworth et al. 2004
; Asante et al. 2006
PRNP codon 129 genotypes in susceptibility-stratified groups from EHP. (The p values are given for 3×2 (Χ2-tests, 2 d.f.) cross-tabulations. ‘n.s.’ denotes not significant.)
The stratified kuru groups show highly significant differences in codon 129 genotype frequencies. The most susceptible group, kuru children, is strongly associated with 129MM and to a lesser degree 129VV genotypes. The low susceptibility groups all show an excess of heterozygosity at codon 129 and particularly a deficit of 129MM. This finding is strongly significant in elderly women who have attended multiple mortuary feasts, but also very prominent in the small number of recent kuru patients with extremely long incubation times. The elevated heterozygosity in healthy old women is presumably caused by homozygous individuals having died of kuru prior to the group being sampled. Overall, these data support the inference of strong balancing selection acting on the Fore at PRNP
concurrent with the kuru epidemic (Mead et al. 2003
Two issues are worthy of more detailed consideration. The exposure of the elderly male group occurred when they were young boys. Infant males were exposed to high-risk tissues (brain and spinal cord) at mortuary feasts in a similar way to females of all ages. Boys around the ages of 6–8 years would alter their participation in mortuary feasts, joining the adolescent and adult males, with low exposure to high-risk tissues. Assuming that these anthropological observations are correct, which is reasonable given their extensive corroboration (Whitfield et al. 2008
), and that the PRNP
codon 129 heterozygosity of elderly men gives an indication of their exposure as boys, we can explore the genetic data to estimate the likely exposure at mortuary feasts in the early twentieth century. This analysis is illustrated in . Here, we show the stratification of the heterozygosity of elderly men by ordering on date of birth and displaying a sliding window of heterozygosity from 100 male codon 129 genotypes. A peak of heterozygosity in the mid-1950s is consistent with epidemiological data regarding the peak incidence of kuru. These genetic data confirm oral history that kuru was a new disease in the twentieth century as extremely elderly men did not appear to be exposed to kuru as children.
Figure 1 Heterozygosity at codon 129 in elderly men and women stratified by date of birth. This graph illustrates a significant excess of heterozygosity for both genders consistent with their exposure at mortuary feasts and the epidemiology of kuru. Males were (more ...)
A further issue of interest is whether the modern young Fore have a higher frequency of resistance genotypes than their neighbours in EHP, potentially as a consequence of the kuru epidemic in its recent history. It is important to consider the population consequences of heterozygote advantage over both homozygous genotypes. In this circumstance, a population may adapt to an equilibrium allele frequency determined by the relative fitness of the two homozygous genotypes. In kuru, 129MM is more susceptible than 129VV, which would be expected to result in equilibrium when 129V is the more frequent allele. 129V is more frequent in the Fore than its neighbours in EHP with no exposure to kuru; however, the difference is small (). With rapid increases in allele frequency, one would expect to observe extensive linkage disequilibrium around the selected allele. In the Fore, the diversity of microsatellite alleles linked to 129V is not consistent with a large and rapid increase in 129V frequency from a low level (Mead et al. 2003
). The population data are therefore consistent with some adaptation to kuru by an increase in 129V frequency in the Fore; however, it is also likely that 129V was a high-frequency allele in the EHP at the outset of the kuru epidemic.