Customize your JAMA Network experience by selecting one or more topics from the list below.
Author Affiliations: Department of Molecular Neuroscience, Reta Lilla Weston Laboratories, UCL Institute of Neurology, Queen Square, London, England.
Twenty years have passed since the first time APOE genotypes were associated with Alzheimer disease (AD): the frequency of the APOE-E4 allele was found to differ significantly (P = .01) between 30 patients with AD selected from 30 families and 91 age-matched unrelated controls.1 This is one of the most replicated observations in biology and, especially since the advent of genome-wide association studies, this particular genetic association has been shown to be true in many cohorts from different populations. Furthermore, since the first genome-wide association studies in AD it was clear that no other locus in the genome would present such a strong association with the disease as that on chromosome 19 around the APOE locus.2 In fact, not only did genetic variants within APOE show association with AD but single-nucleotide polymorphisms (SNPs) around the locus also presented strong associations with the disease. Thus, the whole chromosome region showed association with disease.
There are 3 possible explanations for this extended region of association. The first is that there is actually another gene of pathogenic importance within the region, the second is that genetic variability in the regulatory regions around APOE contributes to disease risk, and the third is that the association simply reflects that there is strong linkage disequilibrium in the region, that is, that sequence variants across the region travel together so that association with APOE-E4 also shows up as association with polymorphisms close to the E4 variant. Although this debate is a new one, the basic observation, that there was an extended haplotype associated with the E4 allele and with disease, was made immediately after the original report of association,3 and in fact, the association between the chromosome 19 haplotype and AD predates the identification of APOE as a locus for disease.4
The debate about the nature of other risk at the chromosome 19 locus beyond the APOE protein polymorphisms was restarted when Roses and colleagues5 used a phylogenetic analysis approach to assess risk at the locus. After deep sequencing of the APOE region, they identified different phylogenetic clades that could be differentiated by the size of a poly-T repeat located in an intron of TOMM40. In their analysis, a longer poly-T was associated with an increased risk of developing AD and with an earlier age at onset of the disease in the small number of samples studied. TOMM40 is a channel-forming subunit of the translocase of the mitochondrial outer membrane (TOM complex),6 which forms the protein-conducting channel in the outer mitochondrial membranes, facilitating the translocation of unfolded proteins from the cytosol into the mitochondrial intermembrane space.7 Given that mitochondrial dysfunction has been widely implicated in the etiology of AD,8TOMM40 was proposed to be a good candidate gene for AD. However, close to 700 good candidate genes are now represented in the AlzGene database, a field synopsis of genetic associations studies in AD, from which less than 15 can be interpreted as true susceptibility AD genes.9 The golden rule to consider a gene as a true risk factor for a determined disease has been the ability to replicate the original association. Several studies have tried to untangle the roles of genetic variability in APOE and the nearby genes in AD and associated phenotypes with contrasting results,10-15 but with the exception of the report by Cruchaga et al,12 where approximately 2500 cases and controls were assessed, all the other studies rely on relatively small numbers of samples. Although no association with age at onset was identified by Cruchaga and colleagues, a significant association between the TOMM40 poly-T repeat and the risk of developing AD was found to have the opposite direction as the original report.5
In this issue of the Archives of Neurology, Jun and colleagues16 evaluate the association of SNPs in the APOE region and attempt to replicate the original association of TOMM40 poly-T rs10524523 with the risk and age at onset of AD. To do this, they used a series of conditional logistic regression models and survival analysis in a very large cohort of more than 10 000 cases and 10 000 controls. After adjusting the models for APOE genotypes, no significant independent association was found between AD and any of the SNPs studied in the APOE region. The conclusions from this study arise from a well-designed series of analyses conducted in a very robust set of data, as Jun and colleagues went to great efforts to have good-quality genotyping calls. In addition to this, the large number of samples studied allowed the preservation of statistical power in analyses of stratified sets of data. In the light of these results, it is very difficult to attribute a genetic and APOE -independent role of TOMM40 in the risk and age at onset of AD development.
Although Cruchaga et al have failed to detect any association between the studied SNPs and TOMM40 or APOE complementary DNA levels, larger samples may be needed to detect expression quantitative trait loci signals, which means one cannot yet rule out the possibility of expression effects in addition to APOE genotypes as having an effect in disease.17 In fact, a recent functional analysis of the APOE locus has revealed that genetic variation in this locus cis-regulatory element enhancer region significantly influences, at least in SHSY5Y cells, the in vitro expression driven by APOE promoters. The data from this report indicate that multiple APOE locus cis-elements may influence APOE promoter activity according to haplotype and cell type, suggesting that complex transcriptional regulatory structures may modulate regional gene expression.18
The role of the E2, E3, and E4APOE alleles was established by studying 30 cases and 90 controls. Now the strong linkage disequilibrium in this genomic region is demanding more than 20 000 samples for the elucidation of the role of the nearby gene, TOMM40.
The identification of true genes associated with diseases is of clear importance as it can lead to a better understanding of pathways and biological processes underlying and contributing to the disease and point the way to the identification of additional factors that may be important in the development of therapies. The literature regarding the genetic role of TOMM40 in AD has been growing at a steady state and the genetic data have been repeatedly suggested to be of importance for diagnostic, prognostic, and therapeutic strategies.5,19,20 These suggestions should all be reassessed in light of the present definitive negative results.
Correspondence: Dr Hardy, Department of Molecular Neuroscience, Reta Lilla Weston Laboratories, UCL Institute of Neurology, Queen Square House, 9th Floor, Queen Square, London WC1N 3BG, England (firstname.lastname@example.org).
Published Online: August 6, 2012. doi:10.1001/archneurol.2012.1935
Author Contributions:Study concept and design: Guerreiro and Hardy. Acquisition of data: Hardy. Drafting of the manuscript: Guerreiro and Hardy. Critical revision of the manuscript for important intellectual content: Guerreiro and Hardy. Obtained funding: Hardy.
Financial Disclosure: None reported.
Guerreiro RJ, Hardy J. TOMM40 Association With Alzheimer Disease: Tales of APOE and Linkage Disequilibrium. Arch Neurol. 2012;69(10):1243–1244. doi:10.1001/archneurol.2012.1935
Browse and subscribe to JAMA Network podcasts!
Create a personal account or sign in to: