2 COMMONLY EXPANDED CAG CTG REPEAT LOCI - INVOLVEMENT IN AFFECTIVE-DISORDERS/

An association between bipolar affective disorder and CAG/CTG trinucle otide repeat expansions (TRE) has previously been detected using the r epeat expansion detection (RED) method. Here we report that 89% of RED products (CAG/CTG repeats) >120 nt (n = 202) detected in affective di sorder patients as well as unaffected family members and controls corr elate with expansions at two repeat loci, ERDA1 on chromosome 17q21.3 and CTG18.1 on 18q21.1. In a set of patients and controls in which we had previously found a significant difference in RED size distribution , the frequency of expansions at the CTG18.1 locus was 13% in bipolar patients (n = 60) and 5% in controls (n = 114) (P < 0.07) with a signi ficantly different size distribution (P < 0.03). A second set of patie nts were ascertained from 14 affective disorder families showing antic ipation. Twelve of the families had members with RED products >120 nt. The RED product distribution was significantly different (P < 0.0007) between affected (n = 53) and unaffected (n = 123) offspring. Using P CR, a higher frequency (P < 0.04) of CTG18.1 expansions as well as a d ifferent (P < 0.02) repeat size distribution was seen between affected and unaffected offspring. In addition, a negative correlation between RED product size and the age-of-onset could be seen in affected offsp ring (r(s) = -0.3, P = 0.05, n = 43). This effect was due to an earlie r onset in individuals with long CTG18.1 expansions. No difference in ERDA1 expansion frequency was seen either between bipolar patients (35 %, n = 60) and matched controls (29%, n = 114), or between affected an d unaffected offspring in the families. We conclude that expanded alle les at the CTG18.1 locus confers an odds ratio of 2.6-2.8 and may thus act as a vulnerability factor for affective disorder, while the ERDA1 locus seems unrelated to disease.