Figure 1 A, Schematic representation of the CAG+ CCG polymorphic region in the HD gene. The repeat regions are boxed. Primer M 1 2 3 4 orientation is indicatedby thearrows. Primers are indicated as in the study by Andrew et al.(1994). Both the normal and the mutated HD alleles of patient HD16 are shown. The spacer region between the CAG and the CCG repeats is boxed. B, PCR amplification of genomic DNA from patient HD16, analyzed on a 2% agarose gel. Lane M, Molecular weight markers. Lane 1, PCR amplification of the CAG polymorphic region (reaction 1, primers hd344 and hd477). Lane 2, PCR amplification of the CAG+ CCG polymorphic region (reaction 2, primers hd344 and hd482). Lane 3, PCR amplification of the CCG polymorphic region (reaction 3, primers hd419 and hd482). Lane 4, PCR amplification of the CAG polymorphic region with the newly synthesized primer hd450 (reaction 4, primers hd344 and hd450). Asterisks (*) indicate oligonucleotide primers (right) and primer dimers (left). mutation changing the triplet CAA into thetriplet CAG. Whatever the case, since the mutation lies within the triplet at the 3'end of the conventional antisense primer hd447, it is likely to hamper annealing of the primer and/or polymerase-dependent elongation of the se-quence, thus preventing PCR amplification of the ex-panded allele. When PCR reaction was carried out with an ad hoc synthesized antisense primer (primer hd450, fig. 1A) whose 3'end was complementary to the nucleotide (nt 436) preceding the deleted CAA triplet, the ex-panded allele was correctly amplified (fig. 1B, lane 4). The identification of this deletion raises the question of whether this mutation was unique to this patient and inherited. Additional screening of HD families led us to identify two additional symptomatic HD siblings from a different unrelated family (family N), both harboring the CAA triplet deletion on the pathological expanded allele. It is interesting to note that a mutation in this triplet changing CAA to CAG has been recently identified in one HD family, resulting in a pure CAG tract that was associated with marked intergenerational instability (Goldberg et al. 1995). Although intergenerational in-stability could not be evaluated in our families, altogether, the data suggest that, in the HD gene, the region between the CAG and the CCG tracts is a hot spot for mutations. Thus, in spite of the low frequency of the mutation reported here-in our series, it hasnot been observed in other 140 HD patients from 113 unrelated families-we suggest that the occurrence of deletions or point mutations in this region may be a cause of diagnos-tic errors and misinterpreted genetic data in HD families, which can be avoided by performing routinely the whole set of three PCR reactions.