Many polymorphic CAG/CTG repeats have been identified in the human genome. Although these repeats are highly polymorphic, their number usually does not exceed 40 repeats in normal individuals. These short repeats are stably transmitted from generation to generation. In contrast, abnormally expanded CAG/CTG repeats, typically more than 40, have been found to cause disease. Longer expansions result in earlier onset and more severe clinical manifestations. The expanded repeats are unstably transmitted to succeeding generations, with a tendency toward elongation of some of the repeats. This genetic instability causes earlier onset and increased severity in successive generations (anticipation). The phenomenon of anticipation in pedigrees has led to the suspicion that instability in the expanded repeat underlies a given disorder. The method of repeat expansion detection (RED) allows detection of expanded CAG/CTG repeats (Schalling et al., 1993). Nakamoto et al. (1998) used a method for cloning gene fragments with expanded CAG/CTG repeats from RED-positive DNAs applied to RED-positive DNAs from a patient with familial spastic paraplegia and another with the Holmes type of cerebellar ataxia. They succeeded in cloning the gene fragments with expanded CAGs. Both gene fragments included pure CAG/CTG repeats with no interruption. Unexpectedly, their surrounding sequences were identical, and FISH experiments revealed that both fragments were derived from the same genetic locus at 17q21.3. Thus, 'this expansion seemed not to relate to a specific disorder.' Subsequently, it was shown to be a polymorphism; 32 of 75 Japanese students (42.7%) had alleles containing more than 51 CAG/CTG repeats; of 150 alleles, 39 (26%) showed the expansion. The length of the polymorphic repeat ranged from 10 to 92 in these 75 normal Japanese individuals. The authors found that 4 of 30 randomly chosen Caucasians (13.3%) had alleles with more than 51 CAG/CTG repeats (5 of 60 alleles, or 8.3%). Most repeats fell into 2 clusters: one from 10 to 30 and the other from 55 to 92. This feature was reminiscent of previously identified disease-related alleles. All of the sequenced alleles with the long and the short CAG/CTG repeats contained polymorphic C and A residues, respectively, next to the CAG repeats. This observation suggested the existence of a founder chromosome with the expanded repeat which was common in both Japanese and Caucasians. The flanking sequences of the CAG/CTG repeat showed no homology with any reported sequences or ESTs. Northern blot and RT-PCR analyses failed to detect any transcriptions of the ERDA1 CAG/CTG repeats.

Deka et al. (1999) studied allelic distribution at the ERDA1 locus in 5 human populations and characterized the mutational patterns by direct observation of 731 meioses. The data showed that large alleles with 40 CAG repeats or more are generally most common in Asian populations, less common in populations of European ancestry, and least common among Africans. They observed high intergenerational instability (46.3%) of the large alleles. Although the mutation rate was not dependent on parental sex, paternal transmission resulted predominantly in contractions, whereas maternal transmission yielded expansions. Within this class of large alleles, the mutation rate increased concomitantly with increasing allele size, but the magnitude of repeat size change did not depend on the size of the progenitor allele. Sequencing of specific alleles showed that the intermediate-sized alleles (30 to 40 repeats) have CAT/CAC interruptions within the CAG-repeat array. These results indicated that expansion and instability of trinucleotide repeats are not exclusively disease-associated phenomena.

Del-Favero et al. (2002) studied the CAG repeat at the ERDA1 locus in a large combined European case-control sample of bipolar affective disorder (125480). The sample consisted of 403 patients and 486 controls matched for age, gender, and ethnicity. The patients were consecutively recruited from 5 participating centers in Belgium, Croatia, Denmark, Scotland, and Sweden. Dichotomous analysis of the combined sample did not show a significant difference in expansion frequency between cases and controls. The findings did not support the involvement of this repeat locus in bipolar affective disorder.