Evidence for linkage disequilibrium between D16S94 and the adult onset polycystic kidney disease (PKD1) gene.

Genetic mapping of the PKD1 gene in chro-mosomal region 16pl3.3 has depended on the identification of closely linked markers and rare recombinants that suggest the order of loci in relation to meiotic crossover points.'2 This has led to the proposal that the PKD1 gene lies in an approximately 750 kb region close to marker loci D16S84, D16S125, and D16S94, and proximal to the hypervariable probe D16S85 that first detected linkage to PKD1.34 However, genetic (meiotic) mapping can sometimes lead to inconsistent localisa-tions, as with the Huntington's disease gene,5 owing to factors such as rare double cross-overs, gene conversion events, gonadal mosaic-ism, genetic heterogeneity, misdiagnosis, mistyping, and non-paternity. While many of these factors can be minimised or excluded, some are very difficult to exclude and, although rare, become significant compared with the frequency of rare recombination events separating markers from disease genes. In the absence of patients with chromosomal rearrangements that can help to localise the PKD 1 gene physically, genetic methods of mapping are critical. A means of genetically mapping a disease gene, other than by linkage analysis, is by the identification of non-random allelic association (linkage disequilibrium) between a marker and disease locus. This can occur if a mutation causing disease arises in close association with a relatively uncommon marker allele or haplo-type, which can then become over-represented in the patient group relative to the general population. The finding of linkage disequilib-rium has served to refine the mapping both of the cystic fibrosis gene6 and of the Hunt-ington's disease gene.7 However, linkage dis-equilibrium is expected to occur only if the mutation rate is relatively low or if the frequency of a particular chromosomal haplotype is inflated in the population of affected subjects as a result of previous selective advantage, migration, or population stratification. It requires that many or most affected subjects carry the same original mutation. It has been suggested that the mutation rate in ADPKD is high,8 based on the apparently high frequency of unaffected parents of ADPKD subjects and on the high population prevalence. This study was carried out at a time when methods of presymptomatic diagnosis were less sensitive than with ultrason-ography and when paternity testing was less readily available. We therefore tested a genetically homogeneous set of 33 ADPKD families for evidence of linkage disequilibrium between the PKD1 locus and closely linked markers D16S84, D16S125, and D16S94. A set of 34 ADPKD families was ascertained …