We examined a large family with autosomal dominant PD (average age of onset, 34 years), ranging clinically from dementia with Lewy bodies to typical PD (3). Neuropathological examination of affected members revealed profound pathology including extensive Lewy bodies and some glial cell cytoplasmic inclusions. Screening this family for mutations in, or linkage to, SNCA was negative. Linkage analysis revealed a chromosome 4p15 haplotype segregating with parkinsonism and essential tremor, with suggestive evidence for linkage to PARK4 [multipoint logarithm of odds (LOD) 2.64 at D4S1609](4). However, an unaffected individual who did not share the 4p15 haplotype became ill. This prompted a second genome-wide search at higher resolution, which revealed a haplotype co-segregating with disease over 26 cM (D4S2367–D4S1560), with a multipoint LOD of 3.50 at D4S2460. The SNCA genotypes were inconsistent with previous data, leading to initial exclusion; re-evaluation of the original linkage revealed a sample swap. Resequencing of SNCA failed to reveal pathogenic mutations.

The heterozygous single nucleotide polymorphisms in the SNCA promoter and in intron 5 suggested that deletion of this region was unlikely. Reverse transcriptase–polymerase chain reaction (RT-PCR) amplification of SNCA revealed only transcripts of normal length and sequence in affected family members. Analysis of intragenic markers MG4S2 and MG4S5 at the SNCA locus showed apparent examples of non-mendelian inheritance, which could be interpreted as multiple alleles. Quantitative real-time PCR amplification of SNCA exons yielded results consistent with whole gene triplication (Fig. 1). To confirm SNCA triplication, we performed fluorescent in situ hybridization (FISH) of chromosomes from Epstein-Barr virus (EBV)-immortalized lymphocytes from an affected family member (9-77). SNCA triplication in this family segregates with