23 Jul 2002

Stockholm. In the early days of the California gold rush, it looked like all you had to do to join the ranks of the rich and famous was to pack up a mule, stake a claim, and start picking up the nuggets. But apart from a limited number of rich veins discovered by a few lucky individuals, finding and harvesting those riches turned out to be hard work and, for most folks, unsuccessful. Might the same thing be in store for genetic miners working on Alzheimer’s disease?

Many investigators would trace the advent of “modern” AD research to the discovery of genetic mutations associated with familial forms of the disease. These well-established loci are associated with three genes, APP, PS1 and PS2. The fact that virtually all of the mutations result in elevation of Aß42 levels has bolstered the “amyloid cascade” hypothesis (reviewed by John Hardy in one of the plenary talks here), which, in spite of ongoing modification, still places the Aß peptide as central to disease etiology. The nagging doubts that continue to plague some investigators regarding this virtual dogma arise from the fact these mutations occur in a remarkably small number of families and, according to most estimates, probably account for, at most, five percent of AD cases. Yet the success of the gene-screening approach in identifying potential causes of familial forms of the disease has spurred continued effort to locate comparable genetic risk factors for the so-called “non-familial” or sporadic forms of the disease, most of which have a later age of onset (LOAD).

Virtually every Alzheimer’s conference has an obligatory session devoted to the latest progress in the search for these genes, and the international conference in Stockholm is no exception. Several such reports were presented from this apparently limitless frontier by leaders of the search, several of whom have already earned their place in the miners' Hall of Fame. The details of the reports presented at this session might make fascinating yarns to those who truly grasp concepts such as Markov Chain Monte Carlo simulation methods and SNP haplotype analysis. However, to to the rest of us, the bottom line is whether or not specific genes are implicated in AD risk. It probably is premature to say that the search should be called off, but the take-home message may not be too different from the quip that Rudy Tanzi offered when his Powerpoint slides initially failed to appear on the screen. “There aren’t any more Alzheimer’s genes,” he announced. “Thank you. Any questions?” The resulting laughter from the audience was genuine, but seemed to carry a note of anxiety. This is not to say that the list of candidates has gotten any shorter. The territory is still expansive and a number of loci across several chromosomes (1, 5, 6, 9, 10, 12, 19 and 21) have been inconsistently reported to show linkage to AD. But an association comparable to what appears to now be the gold standard of ApoE, remains to be detected.

The panel of speakers included Christine Van Broeckhoven, Gerry Schellenberg, Rudy Tanzi, Lannfelt, Alison Goate and Peggy Pericak-Vance, a stellar lineup of seasoned miners to be sure. And they all seem to agree that there is still gold in “them thar hills” and that “candidate” genes are likely to exist on chromosomes 10 (Goate, Pericak-Vance, Tanzi), 12 (Schellenberg) and 19 (Schellenberg, apparently at a different site than ApoE). On chromosome 10, insulin degrading enzyme (IDE), is a contender and is consistent with other recent lines of evidence indicating that this enzyme may be an important regulator of Aß levels (Tanzi). But Goate noted that none of the candidates they have looked at on chromosome 10, including IDE, have survived scrutiny. (In fact, data to be presented later at this meeting by Steve Younkin suggest that the locus on chromosome 10 may be a gene for an α-catenin that might have a role in presenilin function.) A large scale study of SNPs on chromosome 9 (Pericak-Vance, Schellenberg) have also failed to identify clear leads but the Pericak-Vance group did find some evidence for disease risk associated with specific mitochrondrial haplotypes. Schellenberg provided evidence for a hot spot on chromosome 19 that is distinct from ApoE. So at this time, chromosomes 10 and 19 appear to harbor relevant genetic loci for the risk of late-onset AD but there is no consensus on the identity of the genes.

Each speaker made cases for particular gene candidates, but it seemed clear that no new El Dorado has been identified. One question that none of the speakers addressed is whether it would ever be possible to know when to call off the search (by the way, there are still places in California where you can pan for gold). Certainly, the difficulty now facing this talented group of genetic sleuths is the establishment of the boundaries of the search. Is there a minimal LOD score that can be reliably used to identify relevant genes? Are negative results ever definitive? Tanzi made the point, for example, that the linkage to chromosome 12 is not especially strong, but that their haplotype analysis still implicates α-2-macroglobulin (a still controversial candidate for AD risk). This is just one indication of the difficulty facing further mining in this harsh genetic landscape. But if new genes are to be found, these seasoned prospectors are likely to find them.—Keith Crutcher