Alzheimer Disease 18

A number sign (#) is used with this entry because of evidence that susceptibility to the development of late-onset Alzheimer disease-18 (AD18) is conferred by heterozygous mutation in the ADAM10 gene (602192) on chromosome 15q21.

For a general phenotypic description and a discussion of genetic heterogeneity of Alzheimer disease, see 104300.

Molecular Genetics

Kim et al. (2009) presented evidence from SNP analysis for genetic association of ADAM10 with late-onset Alzheimer disease (LOAD) (rs2305421; p = 0.008). Direct sequencing of the gene identified 2 rare, potentially disease-associated nonsynonymous mutations, Q170H (602192.0006) and R181G (602192.0007), in the ADAM10 prodomain. Heterozygosity for these mutations was found in 11 of 16 affected individuals from 7 of 436 AD families (NIMH cohort), as well as in 2 of 5 affected individuals from 2 of 351 additional affected families (NIA cohort). However, there was incomplete segregation with AD; several affected individuals did not carry the mutation, and several unaffected individuals did carry the mutation. These findings suggested either incomplete penetrance or the involvement of additional factors. Combining both mutations to 1 aggregate genotype showed an association with AD across both samples (p = 0.0186), but a nonsignificant p value of 0.212 after exclusion of the probands initially selected for sequencing. In vitro functional studies showed that both mutations significantly attenuated alpha-secretase activity of ADAM10 (greater than 70% decrease) and elevated A-beta levels (1.5- to 3.5-fold). In contrast, Cai et al. (2012) did not find an association between variation in the ADAM10 gene among 305 unrelated AD probands and 271 controls of European ancestry. One identified variant (L9V) was determined to be associated with Ashkenazi ancestry and not with AD. The Q170H and R181G variants found by Kim et al. (2009) were not identified.

Suh et al. (2013) confirmed the pathogenicity of the Q170H and R181G variants in animal studies; see ANIMAL MODEL.

Animal Model

Suh et al. (2013) generated transgenic mice carrying the late-onset Alzheimer disease (LOAD)-associated ADAM10 prodomain variants Q170H or R181G, as well as an artificial dominant-negative mutant E384A. All 3 types of mutant mice showed decreased levels of Adam10 C-terminal fragments compared to wildtype, indicating that these mutations interfere with the normal ectodomain shedding of Adam10 itself. Q170H and R181G mutant mice showed significant attenuation of APP processing compared to wildtype, with a decrease in APP-CTF-alpha levels and an increase in sAPP-beta levels, indicating that the mutations attenuated Adam10 alpha-secretase activity on APP. Crossing these Adam10 mutant mice with the Tg2576 AD mouse model showed that the Adam10 mutations increased amyloidogenic APP processing, as manifest by a shift from the alpha-secretase to the amyloidogenic beta-secretase pathway. This was associated with increased beta-amyloid plaque load and reactive gliosis in the brains of mutant transgenic mice. The changes were not as significant as those observed with the dominant-negative Adam10 mutation, suggesting that the Q170H and R181G mutants diminish, but do not abolish alpha-secretase activity. The LOAD-associated mutations were shown to decrease hippocampal neurogenesis compared to wildtype Adam10. The LOAD mutations impaired the intramolecular chaperone function of the Adam10 prodomain. Collectively, these findings suggested that diminished alpha-secretase activity of ADAM10 on APP resulting from mutations in the ADAM10 prodomain can cause AD-related pathology.