Familial (autosomal dominant) Alzheimer's disease (FAD) is a genetically heterogeneous disorder. Mutations in exons 16 and 17 of the amyloid beta-protein precursor (beta PP) gene currently account for less than 2% of FAD kindreds. No known defect in beta PP quantity, structure, or processing accounts for disease-associated beta-amyloid deposition in the majority of early-onset FAD kindreds. Only two out of a sample of 48 pedigrees (particularly the early onset FAD 4 kindred) contributed noticeably to evidence of linkage at the D21S16/13 and S1/S11 loci in the chromosomal region 21q21 . Many early onset FAD pedigrees (including the FAD 1 and FAD 4 kindreds) show strong evidence of linkage to markers in the chromosome 14q24.3 region. Patients with trisomy 21 (Down's syndrome, DS) virtually always develop a histopathological phenotype indistinguishable from FAD, presumably on the basis of increased beta PP gene dosage and transcription. Whereas no beta PP gene duplication has been found in FAD, other mechanisms that augment beta PP production by effects at the transcriptional level could explain some FAD cases. Here, we report that cultured fibroblasts from affected members of the FAD 1 pedigree show a approximately 1.9 fold increase (P = 0.007) in beta PP mRNA levels compared to unaffected members when the cells are grown under stressed conditions in 0.5% serum. The elevated levels of beta PP mRNA in cells cultured in 0.5% serum also cosegregate with haplotypes in the 14q24.3 region when analyzed by linkage methods (LOD score = 3.26 at theta = 0.001). This is the chromosomal region to which FAD in this family has previously been mapped. As expected, fibroblasts from patients with DS used as a control show a similar beta PP mRNA increase. Fibroblasts from the FAD 4 pedigree did not show this defect under the conditions utilized here. beta PP and A beta protein levels were determined quantitatively after metabolic labeling and immunoprecipitation and found to increase 2.0 and 2.5 fold, respectively, in the fibroblasts from affected FAD 1 members. Finally, transient transfections of a beta PP promoter/chloramphenicol acetyl transferase reporter gene construct demonstrated a approximately 3-4 fold increase in beta PP promoter activity in affected fibroblasts from the FAD 1 but not the FAD 4 pedigree. Taken together, these data raise the possibility that an increase in beta PP transcription may underlie the AD phenotype in at least some of the chromosome 14-linked FAD families.