Purpose: Down syndrome (DS) is a major cause of congenital heart disease (C
HD) and the most frequent known cause of atrioventricular septal defects (A
VSDs). Molecular studies of rare individuals with CHD and partial duplicati
ons of chromosome 21 established a candidate region that included D21S55 th
rough the telomere. We now report human molecular and cardiac data that nar
row the DS-CHD region, excluding two candidate regions, and propose DSCAM (
Down syndrome cell adhesion molecule) as a candidate gene. Methods: A panel
of 19 individuals with partial trisomy 21 was evaluated using quantitative
Southern blot dosage analysis and fluorescence in situ hybridization (FISH
) with subsets of 32 BACs spanning the region defined by D21S16 (21q11.2) t
hrough the telomere. These BACs span the molecular markers D21S55, ERG, ETS
2, MX1/2, collagen XVIII and collagen VI A1/A2. Fourteen individuals are du
plicated for the candidate region, of whom eight (57%) have the characteris
tic spectrum of DS-CHD. Results: Combining the results from these eight ind
ividuals suggests the candidate region for DS-CHD is demarcated by D21S3 (d
efined by ventricular septal defect), through PFKL (defined by tetralogy of
Fallot). Conclusions: These data suggest that the presence of three copies
of gene(s) from the region is sufficient for the production of subsets of
DS-CHD. This region does not include genes located near D21S55, previously
proposed as a "DS critical region," or the genes encoding collagens VI and
XVIII. Of the potential gene candidates in the narrowed DS-CHD region, DSCA
M is notable in that it encodes a cell adhesion molecule, spans more than 8
40 kb of the candidate region, and is expressed in the heart during cardiac
development. Given these properties, we propose DSCAM as a candidate for D
S-CHD.