Search

The identification of mutations in the amyloid precursor protein (APP) gene associated with the presence of early-onset familial Alzheimer disease (AD) raises the possibility of their practical clinical application, at least in some circumstances, in the diagnostic assessment for AD. As a stimulus for discussion, a hypothetical, illustrative case vignette is presented. A 48-year-old man, concerned about recent memory loss and with a family history of early-onset AD, requested testing for the APP717 Val-->Ile mutation, previously identified in his relatives affected with AD. Whether the testing should be undertaken is considered in the context of the current interpretation of potential test results as well as the competency of the individual who requested the test to provide informed consent. Informed consent includes an understanding of the foreseeable risks and benefits associated with disclosure of test results. Although molecular genetic testing in particular individuals, such as the man described herein, could be appropriate, it should not be interpreted to apply in general at this stage to individuals suspected of having AD. In view of a number of caveats, including the genetic heterogeneity of AD, which significantly limits the sensitivity and specificity of the currently available genetic tests, further research and discussion is strongly recommended before widespread introduction of molecular genetic testing for individuals with suspected AD.

A genetic locus associated with familial Alzheimer disease (FAD) and a candidate gene, APP, encoding the amyloid protein precursor have both been assigned previously to chromosome 21, and, in a few FAD families, mutations of APP have been detected. However, obligate crossovers between APP and FAD have also been reported in several FAD pedigrees, including FAD4, a large kindred showing highly suggestive evidence for linkage of the disorder to chromosome 21. In case the apparent APP crossover in FAD4 actually represented an intragenic recombination event or segregation of different mutations in different family branches, we have performed a more detailed assessment of APP as a candidate gene in this family. The entire coding region of the APP gene was sequenced for FAD4 and for FAD1, a second large kindred. No mutations were found, indicating that, in at least one chromosome 21-linked FAD pedigree, the gene defect is not accounted for by a mutation in the known coding region of the APP gene. A total of 25 well-characterized early- and late-onset FAD pedigrees were typed for genetic linkage to APP, to assess the percentage of FAD families predicted to carry mutations in the APP gene. None of the FAD families yielded positive lod scores at a recombination fraction of 0.0. To estimate the overall prevalence of FAD-associated mutations in the beta A4 domain of APP, we sequenced exons 16 and 17 in 30 (20 early- and 10 late-onset) FAD kindreds and in 11 sporadic AD cases, and we screened 56 FAD kindreds and 81 cases of sporadic AD for the presence of the originally reported FAD-associated mutation, APP717 Val----Ile (by BclI digestion). No APP gene mutations were found in any of the FAD families or sporadic-AD samples examined in this study, suggesting that the mutations in exons 16 and 17 are a rare cause of FAD. Overall, these data suggest that APP gene mutations account for a very small portion of FAD.