Founder effect and recurrent mutational events in fatal familial insomnia

Abstract

The origin and distribution of the D178N-129M mutationcausing fatal familial insomnia (FFI) are controversial.Some authors have found evidence of founder effects whileworking with genealogical data [1, 2], whereas otherspropose recurrent mutations based on STR data [3]. Arecent study on SNP haplotypes close to D178N [4]conflicts with findings derived from STRs and seems tosupport the notion of founder effect. However, a commentappearing in the present issue of Neurogenetics questionssuch results. The main points in Lee and Goldfarb’s critique[5] concern the use of G-test and the SNPs selected for theanalysis, concluding that we failed to provide conclusiveevidence that only two founder haplotypes are associatedwith all known European FFI cases.A thorough reading of the discussion section permits toperceive clearly that results are not presented as conclusiveevidence but rather pointing to a certain direction. Specif-ically, in the last paragraph, we consider several uncertain-ties that leave the debate as for the origin of FFI open [4]:(1) “the results presented herein do not exclude thepossibility that other less successful mutational eventsmight have occurred” and (2) “probably, the true numberof founding events will be greater than two (minimalnumber estimated from SNPs)”.As Lee and Goldfarb indicate, G-test was employed toevaluate whether there were significant differences infrequencies of SNP haplotypes between patients andcontrol group from Spain. Nevertheless, conclusion aboutminimal number of founder mutations does not stem fromG-test results but essentially from the meticulous examina-tion of haplotypic and genealogical data. The notion offounding events for the D178N mutation is not new. Basedon demographic records, Harder et al. [1] reconstructed alarge German FFI kindred and assigned nine patients tothree related pedigrees comprising more than 800 individ-uals, whose common ancestors were traced to the 17thCentury. Likewise, a recent survey demonstrated that 12FFI cases from the Basque Country belonged to threeprobably related pedigrees traced back to the 16th and 17thCenturies [2].The SNPs used to construct haplotypes were selectedaccording to proximity to the D178N mutation to avoidoverestimation of haplotypic diversity by recombination.Lee and Goldfarb consider “especially troubling” that thehaplotypes identified as cosegregating with D178N-129Mare common in the background population. This finding isneither surprising nor unexpected. Owing to elementaryprobabilistic reasons, the opportunity that a deleteriousmutation occurs in a chromosome harboring a relativelycommon haplotype in the population is conspicuously high.Last but not least, an evolutionary genetics consideration—bearing in mind that onset of FFI usually occurs after themutated individuals have had offspring, intensity of naturalselection against the autosomal dominant mutation shouldnot be too high, as deduced from published genealogies [1,2]. It seems, thereby, little probable that this neurodegen-erative disorder appears by recurrent mutations in different