27 June 2010. To the Alzheimer disease geneticist, they’re becoming household names. CLU, CR1, and PICALM—top hits from recent AD genomewide association studies—emerge again as late-onset AD risk genes in a large case-control association study published online June 14 in the Archives of Neurology. “It was gratifying to obtain near-perfect replication of the results previously reported,” principal investigator Steven Younkin of Mayo Clinic, Jacksonville, Florida, wrote in an e-mail to ARF. Furthermore, a study in this month’s print issue of the Archives of Neurology finds that CR1, PICALM, and two other loci (BIN1 and CNTN5) track with brain imaging measures that predict AD risk and progression, suggesting these genes influence biological processes relevant to disease. Jonathan Rosand of Massachusetts General Hospital, Boston, led that research, which analyzed freely available data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI).

Since researchers discovered ApoE 17 years ago as a genetic risk factor for sporadic AD (Corder et al., 1993), hundreds of other genes have emerged as contenders but linger in obscurity because researchers failed to replicate their association with late-onset AD (LOAD) in other cohorts. That’s why CLU (clusterin/apolipoprotein J), CR1 (complement receptor 1), and PICALM) (phosphatidylinositol-binding clathrin assembly protein) made headlines last year when scientists reported that variants of these genes had achieved genomewide significance in two large, independent GWASs (Harold et al., 2009; Lambert et al., 2009).

The Younkin study provides further evidence that these risk genes are here to stay. Scouring the genomes of 1,829 sporadic AD cases and 2,576 controls—none overlapping with samples from last year’s GWAS publications—first author Minerva Carrasquillo and colleagues confirmed the LOAD associations of the tried-and-true ApoE and, more importantly, the three 2009 loci. “It looks like the variants in PICALM, CLU, and CR1 are going to hold up well,” Younkin told ARF. “Let us hope these ‘new’ LOAD genes lead to new therapeutic targets and help to identify the at-risk population.” He added that preliminary analysis “looks promising” for variants of BIN1 and PCDH11X. The latter showed strong association with LOAD in a genomewide study by Younkin and colleagues (Carrasquillo et al., 2009), and Younkin will present follow-up data next month in a talk at the International Conference on Alzheimer’s Disease in Honolulu. BIN1 was one of two novel candidate loci reported earlier this year in AD’s largest GWAS to date (Seshadri et al., 2010 and ARF related news story).

Reassuringly, BIN1 also came to the fore in the current report from Rosand and colleagues. Led by first author Alessandro Biffi, the researchers wanted to see if brain characteristics assessed by neuroimaging might help fish out potential risk genes that otherwise escape detection using strictly genomic approaches. To gauge the gene-trawling utility of brain imaging, the team determined whether the four most established LOAD loci of the time—ApoE, CLU, CR1, and PICALM—showed association with magnetic resonance imaging (MRI) measures that were known to correlate with disease risk and progression. They also looked at variants of two other loci, BIN1 and CNTN5, which looked promising but just missed the cutoff for genomewide significance in the 2009 GWASs.

Combing through ADNI data from 740 patients (168 with probable AD, 357 with mild cognitive impairment, and 215 cognitively normal), the scientists focused on six MRI measures: hippocampal volume, amygdala volume, white matter lesion volume, entorhinal cortex thickness, parahippocampal gyrus thickness, and temporal pole cortex thickness. When assessed cumulatively, the non-ApoE loci showed association with each imaging measure. All but CLU associated individually with certain brain measures (for example, BIN1 with entorhinal and temporal pole cortical thicknesses, PICALM with hippocampal volume and entorhinal cortex thickness), and ApoE associated with all measures except white matter lesion volume.

The study drives home the importance of a multi-pronged approach to finding disease genes. Though clinical studies will uncover some genes, pathology and neuroimaging may help with others. “You really want to bring all of these ways of characterizing patients into finding genes,” Rosand said. Incidentally, this month marked the 10-year anniversary of the first completed draft of the human genome—a feat seen by some as overblown (see The New York Times story), by others as the dawn of an exciting, but possibly daunting, era (see Economist story) of research on complex genetic disorders.

Along those lines, the current studies add to a growing body of research suggesting that common diseases may not be caused by single common genetic variants but rather by a combination of rare loci. To complicate matters, variants at a single locus can associate with multiple disease, as is the case with the Gaucher disease gene glucocerebrosidase (GBA), where a slew of variations confer genetic risk to Parkinson disease and other Lewy body diseases, too (see ARF related news story). A recent report makes the same point by linking common autoimmune diseases to more than a dozen rare variants of sialic acid acetylesterase, an enzyme involved in B cell signaling (Surolia et al., 2010).

Whole-genome association studies will not uncover rare loci “because they are individually responsible for too few cases of disease,” UK scientists John Hardy, University College London, and Julie Williams, Cardiff University, suggest in an editorial accompanying Rosand’s paper (Hardy and Williams, 2010). The development of cost-effective methods for whole-exome (i.e., all exons in the genome) sequencing could help reveal rare variants, they note. More importantly, as these approaches lead us to additional AD genes, “we should be able to map these risk loci onto biochemical pathways,” they write.—Esther Landhuis.

Author's update: In addition to the research covered above, a study published online June 11 in the journal Human Molecular Genetics similarly reported replication of APOE, CLU, PICALM, and CR1 variants in a unique cohort of 1,019 neuropathologically confirmed AD cases and 591 controls (Corneveaux et al., 2010).

References:
Carrasquillo MM, Belbin O, Hunter TA, Ma L, Bisceglio GD, Zou F, Crook JE, Pankratz S, Dickson DW, Graff-Radford NR, Petersen RC, Morgan K, Younkin SG. Replication of CLU, CR1, and PICALM Associations with Alzheimer Disease. Arch Neurol. 2010 Jun 14. Abstract

Biffi A, Anderson CD, Desikan RS, Sabuncu M, Cortellini L, Schmansky N, Salat D, Rosand J for ADNI. Genetic variation and neuroimaging measures in Alzheimer disease. Arch Neurol. 2010 Jun;67(6):677-85. Abstract