13 April 2006. Steve Younkin, Mayo Clinic at Jacksonville, Florida, presented the recent progress in identification of multi-locus genotypes in novel late-onset AD (LOAD) genes, particularly presenilin 1 (PSEN1), insulin-degrading enzyme (IDE) and ubiquilin 1 (UBQNL1). First, Younkin emphasized the need to identify the remaining AD genes, and stated that beyond the major LOAD risk factor, ApoE4, the majority of the remaining AD genes will likely exhibit relatively modest effects on risk and protection. Accordingly, he argued that for case-control studies, up to 10,000 matched cases and controls will be necessary to identify the remaining AD genes. He then described seven DNA variants in the conserved non-coding regions of the gene encoding the insulin-degrading enzyme on chromosome 10. These variants form five common haplotypes, and four risk-conferring and nine protective multi-locus genotypes. Younkin also provided RT-PCR data showing that the protective IDE genotypes were functionally associated with increased IDE mRNA expression in brain. Other genetic data were also presented to support the existence of risk-conferring variants for LOAD in the ubiquilin 1 and presenilin 1 genes.

Mario Bengtson, Genomics Institute of Novartis Foundation, San Diego, California, presented a tauopathy phenotype in the lister mutant mouse. The responsible gene, called “listerin,” was found to encode a novel E3 ubiquitin ligase. The pathogenic mutation in this gene leads to a short, in-frame internal deletion. The mutant mouse shows progressive loss of hind limb reflex extension, motor neuron degeneration, and accumulation of hyperphosphorylated soluble forms of the tau protein. He emphasized that the lister mutant mouse is significant as a new mouse model for motor neuron/tauopathy disease and as the first mutation in a ubiquitin ligase identified in the disease. However, Bengston regretted that he could not reveal the identity of the gene or the human ortholog.

Lars Bertram, Massachusetts General Hospital, Boston, provided an overview of the recently developed “AlzGene” database. The database provides systematic and objective summaries of all peer-reviewed publications in the area of AD genetics and covers over 300 candidate genes and over 800 polymorphisms. Bertram emphasized the dire need for such a database, given the increasing number of AD gene papers and confusion created by mixed positive and negative results for the hundreds of AD candidate genes tested to date, especially since most exhibit modest to moderate effects on risk and protection. He also pointed out the usefulness and necessity of such a database for diseases that are characterized by complex genetics. The Alzgene.org site was described with regard to the data collection process and certification of the methods employed in the meta-analyses—ApoE4 results were used as an example of a positive control gene. Besides ApoE4, nine candidate gene polymorphisms were found to be positive so far by AlzGene meta-analyses: ApoE (promoter region), apolipoprotein C1 (ApoC1), angiotensin-converting enzyme (ACE), cystatin 3 (CST3), estrogen receptor 1 (ESR1), insulin-degrading enzyme (IDE), prion precursor protein (PRP), presenilin 1 (PSEN1), and transferrin (TF). He went on to describe results of family-based association analyses in NIMH and CAG samples, which revealed genetic association of AD with the ubiquilin 1 gene on chromosome 9 and the IDE gene on chromosome 10. Several negative candidate genes on chromosome 19 were also reviewed.

Ellen Wijsman, University of Washington, Seattle, outlined her strategy for identification of late-onset AD genes following up a genome scan in the Seattle LOFAD pedigrees using Markov Chain Monte Carlo oligogenic linkage analysis. Age-at-onset was used as an AD quantitative trait and ApoE as a major gene covariate for adjustment. She presented evidence of a new genetic association of LOAD with five single nucleotide polymorphisms in a locus on chromosome 19p at a position near 33 cM, but could not reveal the identity of the candidate gene.

Lisa Paige, Metabolon Inc., Research Triangle Park, North Carolina, presented a comparative metabolomic analysis of plasma from patients with AD or mild cognitive impairment (MCI) as a means to search for metabolic biomarkers for AD. After an overview of the metabolon process, she provided proof-of-concept data based on a strong correlation between metabolomic and clinical lab measures of creatinine, regarding both sensitivity and specificity. No metabolic markers were yet found in AD versus control plasma, emphasizing the challenges in developing biomarkers for the disease.