MAPPING THE FRIEDREICH ATAXIA LOCUS (FRDA) BY LINKAGE DISEQUILIBRIUM ANALYSIS WITH HIGHLY POLYMORPHIC MICROSATELLITES

The Friedreich's ataxia locus (FRDA) is tightly linked to markers D9S5 and D9S15 located in 9q13-q21. Cumulated maximum lod scores between F RDA and D9S5 and between FRDA and D9S15 are above 36 and 61, respectiv ely, at a recombination fraction of 0, indicating that recombination e vents needed to orient the search of the gene are very difficult to id entify and ascertain. We have established a 1 Megabase PFGE map around D9S5 and D9S15 and isolated a corresponding 530 kb YAC contig. We fou nd that the two markers are 260 kb apart. This result was surprising, since D9S5 and D9S15 were independently isolated, but in agreement wit h the strong linkage between the two loci (lod score >35 at a recombin ation fraction of 0). Seven clusters of rare cutter enzyme sites (CpG islands), which are potential indicators of genes, were identified in the 1 Megabase region by PFGE analysis and YAC mapping. The search for genes around the CpG islands is in progress. To map the Friedreich at axia locus in the absence of clearly identified recombination events, we chose an alternative approach based on haplotype analysis of patien ts from small populations with precise geographic and historical origi ns, such as the Louisiana-Acadians, deported from Nova-Scotia about 15 0 years ago and who remained isolated for historical and cultural reas ons. In this population, a single mutation, associated with a specific haplotype may account for the majority of Friedreich ataxia cases. Ha plotypes different from the major haplotype at one or the other extrem ity can indicate ancient recombinations. Analysis of extended haplotyp es allows therefore to scan rapidly a large number of meiosis for reco mbination events. We isolated, mainly from YAC clones, six multi-allel ic markers (one RFLP and five CA(11) microsatellites) around D9S5 and D9S15. The first five polymorphisms were used to define 405 kb long ha plotypes from Louisiana-Acadian patients [13]. A major haplotype was f ound in 50% of 22 -->independant affected chromosomes and this haploty pe was never found on 16 normal chromosomes. Four additional ''affecte d'' haplotypes were identical to the major haplotype for alleles at FD 1, 26P, GS2 and D9S15 but different for the GS4 allele. Again the four haplotypes were not found on normal chromosomes and they are very lik ely derived from the major haplotype by ancient recombination events. The recombinations would exclude the region beyond GS4 as a possible l ocation for the FRDA locus. Other explanations, polymorphism instabili ty and mutation heterogeneity, cannot be fully excluded. Another indic ation of an ancient recombination event between GS4 and FRDA is given by a patient from a family with known remote consanguinity, who is hom ozygous for a rare haplotype that shows divergence at GS4. Extension o f the haplotypes with the new microsatellite markers should allow to r einforce the present conclusions, as well as identify new recombinant haplotypes in order to narrow down the localization of the Friedreich ataxia gene.