Replication in mammalian cells recapitulates the locus-specific differences in somatic instability of genomic GAA triplet-repeats.

Friedreich ataxia is caused by an expanded (GAA.TTC)n sequence in intron 1 of the FXN gene. Small pool PCR analysis showed that pure (GAA.TTC)44+ sequences at the FXN locus are unstable in somatic cells in vivo, displaying both expansions and contractions. On searching the entire human and mouse genomes we identified three other genomic loci with pure (GAA.TTC)44+ sequences. Alleles at these loci showed mutation loads of <1% compared with 6.3-30% for FXN alleles of similar length, indicating that somatic instability in vivo is regulated by locus-specific factors. Since distance between the origin of replication and the (CTG.CAG)n sequence modulates repeat instability in mammalian cells, we tested if this could also recapitulate the locus-specific differences for genomic (GAA.TTC)n sequences. Repeat instability was evaluated following replication of a (GAA.TTC)115 sequence in transfected COS1 cells under the control of the SV40 origin of replication located at one of five different distances from the repeat. Indeed, depending on the location of the SV40 origin relative to the (GAA.TTC)n sequence, we noted either no instability, predominant expansion or both expansion and contraction. These data suggest that mammalian DNA replication is a possible mechanism underlying locus-specific differences in instability of GAA triplet-repeat sequences.

The GAA triplet-repeat is unstable in the context of the human FXN locus and displays age-dependent expansions in cerebellum and DRG in a transgenic mouse model.

Friedreich ataxia (FRDA) is caused by homozygosity for FXN alleles containing an expanded GAA triplet-repeat (GAA-TR) sequence. Patients have progressive neurodegeneration of the dorsal root ganglia (DRG) and in later stages the cerebellum may be involved. The expanded GAA-TR sequence is unstable in somatic cells in vivo, and although the mechanism of instability remains unknown, we hypothesized that age-dependent and tissue-specific somatic instability may be a determinant of the progressive pathology involving DRG and cerebellum. We show that transgenic mice containing the expanded GAA-TR sequence (190 or 82 triplets) in the context of the human FXN locus show tissue-specific and age-dependent somatic instability that is compatible with this hypothesis. Small pool PCR analysis, which allows quantitative analysis of repeat instability by assaying individual transgenes in vivo, showed age-dependent expansions specifically in the cerebellum and DRG. The (GAA)(190) allele showed some instability by 2 months, progressed at about 0.3-0.4 triplets per week, resulting in a significant number of expansions by 12 months. Repeat length was found to determine the age of onset of somatic instability, and the rate and magnitude of mutation. Given the low level of cerebellar instability seen by others in multiple transgenic mice with expanded CAG/CTG repeats, our data indicate that somatic instability of the GAA-TR sequence is likely mediated by unique tissue-specific factors. This mouse model will serve as a useful tool to delineate the mechanism(s) of disease-specific somatic instability in FRDA.

Expansion of GAA trinucleotide repeats in mammals.

We have previously shown that GAA trinucleotide repeats have undergone significant expansion in the human genome. Here we present the analysis of the length distribution of all 10 nonredundant trinucleotide repeat motifs in 20 complete eukaryotic genomes (6 mammalian, 2 nonmammalian vertebrates, 4 arthropods, 4 fungi, and 1 each of nematode, amoebozoa, alveolate, and plant), which showed that the abundance of large expansions of GAA trinucleotide repeats is specific to mammals. Analysis of human-chimpanzee-gorilla orthologs revealed that loci with large expansions are species-specific and have occurred after divergence from the common ancestor. PCR analysis of human controls revealed large expansions at multiple human (GAA)(30+) loci; nine loci showed expanded alleles containing >65 triplets, analogous to disease-causing expansions in Friedreich ataxia, including two that are in introns of genes of unknown function. The abundance of long GAA trinucleotide repeat tracts in mammalian genomes represents a significant mutation potential and source of interindividual variability.

Expansion of GAA triplet repeats in the human genome: unique origin of the FRDA mutation at the center of an Alu.

Friedreich ataxia is caused by expansion of a GAA triplet repeat (GAA-TR) in the FRDA gene. Normal alleles contain <30 triplets, and disease-causing expansions (66-1700 triplets) arise via hyperexpansion of premutations (30-65 triplets). To gain insight into GAA-TR instability we analyzed all triplet repeats in the human genome. We identified 988 (GAA)(8+) repeats, 291 with >or=20 triplets, including 29 potential premutations (30-62 triplets). Most other triplet repeats were restricted to <20 triplets. We estimated the expected frequency of (GAA)(6+) repeats to be negligible, further indicating that GAA-TRs have undergone significant expansion. Eighty-nine percent of (GAA)(8+) sequences map within G/A islands, and 58% map within the poly(A) tails of Alu elements. Only two other (GAA)(8+) sequences shared the central Alu location seen at the FRDA locus. One showed allelic variation, including expansions analogous to short Friedreich ataxia mutations. Our data demonstrate that GAA-TRs have expanded throughout primate evolution with the generation of potential premutation alleles at multiple loci.

The GAA triplet-repeat sequence in Friedreich ataxia shows a high level of somatic instability in vivo, with a significant predilection for large contractions.

Friedreich ataxia is commonly caused by large expansions of a GAA triplet-repeat (GAA-TR) sequence in the first intron of the FRDA gene. We used small-pool PCR to analyze somatic variability among 7190 individual FRDA molecules from peripheral blood DNA of subjects carrying 12 different expanded alleles, ranging in size from 241 to 1105 triplets. Expanded alleles showed a length-dependent increase in somatic variability, with mutation loads ranging from 47% to 78%. We noted a strong contraction bias among long alleles (>500 triplets), which showed a 4-fold higher frequency of large contractions versus expansions. Some contractions were very large; of all somatic mutations scored, approximately 5% involved contractions of >50% of the original allele length, and 0.29% involved complete reversion to the normal/premutation length (< or =60 triplets). These observations contrast sharply with the strong expansion bias seen in expanded CTG triplet repeats in myotonic dystrophy. No somatic variability was detected in >6000 individual FRDA molecules analyzed from 15 normal alleles (8-25 triplets). A premutation allele with 44 uninterrupted GAA repeats was found to be unstable, ranging in size from 6 to 113 triplets, thus establishing the threshold for somatic instability between 26 and 44 GAA triplets. Analysis of an additional 7850 FRDA molecules from serially passaged lymphoblastoid cell lines carrying nine expanded alleles (132-933 triplets) showed very low mutation loads, ranging from 0% to 6.2%. Our data indicate that expanded GAA-TR alleles in Friedreich ataxia are highly mutable and have a natural tendency to contract in vivo, and that these properties depend on multiple factors, including DNA sequence, triplet-repeat length and unknown cell-type-specific factors.

Genetic admixture of European FRDA genes is the cause of Friedreich ataxia in the Mexican population.

Friedreich ataxia accounts for approximately 75% of European recessive ataxia patients. Approximately 98% of pathogenic chromosomes have large expansions of a GAA triplet repeat in the FRDA gene (E alleles), and strong linkage disequilibrium among polymorphisms spanning the FRDA locus indicates a common origin for all European E alleles. In contrast, we found that only 14 of 151 (9.3%) Mexican Mestizo patients with recessive ataxia were homozygous for E alleles. Analysis of polymorphisms spanning the FRDA locus revealed that all Mestizo E alleles had the common European haplotype, indicating that they share a single origin. Genetic admixture levels were determined, which revealed that the relative contributions to the Mestizo FRDA gene pool by Native American and European genes were 76-87% and 13-24%, respectively, commensurate with the observed low prevalence of Friedreich ataxia in Mestizos. This indicates that Friedreich ataxia in Mexican Mestizos is due to genetic admixture of European mutant FRDA genes in the Native American gene pool that existed prior to contact with Europeans.