Genetic variation in genes involved in Abeta-degradation in Alzheimer disease

Abstract

Alzheimer disease (AD) is a neurodegenerative disorder that causes dementia among mainly elderly people and is increasing in prevalence in industrialized societies from 1 to 3 % in 60-65 year-olds to as much as 50 % by age 95. This disease is complex. Multiple genes are involved, and they potentially affect each other. The strongest known genetic risk factor for developing AD is the carriage of the å4-allele of the gene coding for apolipoprotein E (APOE) (Strittmatter et al., 1993).

This thesis aims to improve the understanding of the genetic basis for AD by comparative re-sequencing of candidate genes for AD in order to discover sequence variants between individuals with or without the disease in the studied population. Comparing such potential differences with known variants is one approach to begin to understand the roles of genes involved in complex diseases. There are today several genes that compete for being the new gene of interest, in addition to APOE, as risk factor for AD.

One such candidate gene is alpha-3 catenin (CTNNA3), which was tested for association with AD in paper I. I genotyped selected markers in connection with CTNNA3, and the results indicated an association with AD. Using these markers in a Swedish and a Scottish AD material, we found no association. More tests would however be needed in order to entirely eliminate this gene.

Paper II provides a special focus on a part of chromosome 10q where the gene encoding the protein insulin degrading enzyme (IDE) is positioned. The purpose of this study was to investigate if polymorphisms within an LD (linkage disequilibrium) block encompassing IDE (Prince et al., 2003), might also influence Parkinson s disease (PD), since the quantitative trait age at onset (AAO) has previously shown linkage to IDE in both AD and PD (Li et al., 2002), suggesting pathogenic alleles influencing both disorders. Our findings pointed towards a linkage disequilibrium (LD) block on chromosome 10q that harbors alleles influencing AD and PD in a similar way, containing the three genes IDE, KIF11 (kinesin family member 11), and HHEX (hematopoietically expressed homeobox).

In paper III, the aim was to further investigate an earlier (Prince et al., 2003) association between DNA variation in IDE and AD pathology and severity. More specifically, the association of IDE variants with the density of senile plaques and neurofibrillary tangles was observed, and I therefore attempted to replicate these findings in an autopsy population in order to measure cerebral amyloid beta (Aâ) levels. This resulted in an observation of significant effects on plasma Aâ levels, which together with no association with direct disease risk in AD case-control samples of one specific single nucleotide polymorphism (SNP), suggested that IDE may have a role as modifier of severity of disease rather than risk.

The aim of paper IV was to examine whether abundant data could exist that might help resolve the role of gene(s) in complex phenotypes, but that many reasonably powered studies producing negative results are not being published. A literature study of association studies based on candidate genes in AD confirmed a publication bias. We also tested 62 genetic markers for association with AD risk in addition to possible effects upon quantitative indices of AD severity, resulting in only modest signals, i.e. the overall results were mostly negative. Identification of the underlying genes for complex human diseases can strengthen our knowledge about disease causes and possibly suggest potential treatment methods. This strategy could also increase our knowledge about gene function at a molecular level by providing new information about changes in DNA-sequences that can influence function of genes. Ultimately, the identification of disease contributing genes could clear the way for more personally formulated methods of treatment.