Editorials

The genetics of Alzheimer's disease

The number of genetic risk factors associated with this disorder is increasing steadily 

The genetics of Alzheimer's disease is proving to be complex and controversial. Nature Genetics this month contains a paper from Tanzi's group in Boston suggesting that a common polymorphism of ₂ macroglobulin is associated with a major increase in the risk of developing late onset Alzheimer's disease.¹ The data generated heated debate at the sixth international conference on Alzheimer's disease held in Amsterdam at the end of July. At least 70 other reports on the genetics of Alzheimer's disease were presented, implicating over a dozen other genes or genetic loci.² So what is our current state of knowledge?

Missense mutations in three genes are known to cause autosomal dominant forms of early onset Alzheimer's disease: these are the amyloid precursor protein gene located on chromosome 21³ and genes for presenilin 1 and presenilin 2 located on chromosomes 14 and 1, respectively. ^{4 5} Studies on these missense mutations have given strong support to the "amyloid cascade hypothesis" of Alzheimer's disease.⁶ The amyloid precursor protein mutations code for amino acids at or near points where the precursor is cleaved enzymically and result in slightly longer forms of  amyloid being secreted. These aggregate readily into highly insoluble amyloid fibrils which form the major component of senile plaques. Similar changes in  amyloid production are observed with the mutations linked to Alzheimer's disease in presenilin 1 and 2. The presenilin proteins show marked homology, with multiple membrane-spanning domains, and may act as chaperone molecules in the processing of amyloid precursor protein, exposing sites in the molecule to enzymatic cleavage.⁷ While mutations associated with amyloid precursor protein are extremely rare, the 50 or so mutations associated with presenilin 1 may explain up to half of all cases of early onset Alzheimer's disease.

In contrast to early onset Alzheimer's disease, there is to date only one genetic factor indisputably linked with late onset forms of this disorder, and that is the e4 allele of apolipoprotein E.⁸ Three common allelic variants of apolipoprotein E existe2, e3, and e4encoded at a single gene locus on chromosome 19; several large, neuropathologically verified cohort studies have shown that apolipoprotein E e4 predicts risk of Alzheimer's disease. However, apolipoprotein E e4 is neither necessary or sufficient to cause Alzheimer's disease, and a population based study of almost 5000 elderly people by Meyer et al,⁹ also reported in this month's Nature Genetics, indicates that the apolipoprotein E genotype predicts whennot whetherindividuals are predisposed to develop Alzheimer's disease. A "plateau" appeared in the survival curve for all groups surviving to old age, so that even with the homozygous e4/e4 condition the last onset of dementia occurred at 84 years, with a significant number of individuals surviving disease free for longer periods. At most only half of individuals with late onset Alzheimer's disease carry an apolipoprotein E e4 allele, and the study by Meyer et al indicates that even in the presence of this susceptibility factor other genes are likely to be involved.

A whole string of genetic associations with late onset Alzheimer's disease have been reported by various groups, including polymorphisms in angiotensin converting enzyme, ₁ antichymotrypsin, bleomycin hydrolase, butyrylcholinesterase, HLA, low density lipoprotein receptor related protein, various mitochondrial enzymes, and a presenilin 1 intronic mutation.² So far none of these findings has been consistently replicated, and it remains to be seen whether the recent report on ₂ macroglobulin¹ will stand this test. The mutation reported by the Boston group is a common variant of the ₂ macroglobulin gene that causes a deletion in the nucleotide sequence for the "bait" region of the molecule, which binds proteases, and which in Tanzi's view may be implicated in the clearance of  amyloid from the synaptic cleft. It is present in 20% of the population, and the level of risk conferred for Alzheimer's disease appears to be similar to that associated with apolipoprotein E e4. However, the statistical analysis of this work used an as yet unpublished family based association method which, though it measures relative risk for the actual families studied, does not indicate the general population risk. In discussion at Amsterdam several groups claimed that they had been unable to confirm the association with ₂ macroglobulin in population samples, so it may be relevant to only a small proportion of familial cases of Alzheimer's disease.

The mutations in amyloid precursor protein, presenilin 1, and presenilin 2 allow for genetic screening in suspected cases of familial Alzheimer's disease with early onset and for appropriate genetic counselling and support. Studies on the underlying pathophysiological mechanisms support the rationale for therapeutic strategies aimed at preventing the formation of amyloid fibrils in vivo or promoting their dissolution. Tanzi's work, if confirmed, will extend the hunt for rational therapies based on the biological functions of ₂ macroglobulin and its role in Alzheimer's disease. Until disease slowing treatments become available there is little justification for predictive testing based on apolipoprotein E, ₂ macroglobulin, or any of the other genes so far linked with late onset Alzheimer's disease.