Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2009 Jan 28.
Published in final edited form as: Neuroreport. 2007 Nov 19;18(17):1761–1764. doi: 10.1097/WNR.0b013e3282f13e7a

Association between SORL1 and Alzheimer disease in a genome-wide study

Yan Meng 1,*, Joseph H Lee 3, Rong Cheng 3, Peter St George-Hyslop 4, Richard Mayeux 3, Lindsay A Farrer 1,2
PMCID: PMC2631643  NIHMSID: NIHMS88296  PMID: 18090307

Abstract

Several studies [1–3] have reported an association of Alzheimer disease (AD) with polymorphic markers in SORL1. Data from a recently published genome wide association study in AD [4] have been made publically available. We tested the association of AD with SORL1 in this dataset (TGEN), which included 31 SORL1 SNPs, 8 of which overlapped the original study [1]. Six SNPs, near the 3′ region of SORL1 containing SNPs which were strongly associated with AD in previous studies, showed significant association in the TGEN dataset. These results provide an independent replication of the association between AD and SORL1.

Keywords: Alzheimer disease, SORL1, association, genome-wide association study

Introduction

Rogaeva et al.[1] reported an association between Alzheimer disease (AD) and several SNPs in the gene encoding the sortilin-related receptor, low-density lipoprotein receptor class A repeat–containing protein - SORL1 (11q23–q24) in four different ethnic groups (Caucasians, African Americans, Israeli-Arabs and Hispanics) analyzed separately. Although the precise identity of the genetic effectors in SORL1 remains to be determined, this initial study pointed to two clusters of SNPs in distinct regions of the gene, implying the existence of multiple allelic variants associated with AD in different populations. Two positive replications have been reported [2,3]. Recently, results of a genome wide association study in AD was published [4] with a focus on a novel association of SNPs in the GAB2 gene with AD. This report however did not mention any results for SORL1. We were therefore interested to test whether the association of AD with SORL1 is replicated in this dataset (TGEN) which is publicly available.

Methods

Subjects

The TGEN dataset [4] was obtained from from the website http://www.tgen.org/neurogenomics/data. Although both the original study by Rogaeva et al [1] and the study by Reiman et al describing the TGEN dataset [4] included subjects ascertained at the Mayo Clinic in Rochester, Minnesota, the Mayo subjects in the TGEN data set and the Mayo subjects included in Rogaeva et al. study are independent. Consequently, we analyzed SORL1 data for 1,408 subjects in the TGEN database which included 1,044 autopsied individuals (641 cases, 403 controls) and 364 clinically examined subjects from the Mayo Clinic (218 cases and 146 controls).

Statistical analyses

SNP marker data were assessed for deviations from Hardy-Weinberg equilibrium using Haploview [5] software. Single point allelic and genotypic tests were performed using PLINK [6]. Marker genotype distributions in cases and controls were compared in several ways: (1) a genotypic test with two degrees of freedom, models assuming (2) dominant and (3) recessive inheritance, and (4) the Cochran-Armitage trend test.

Linkage disequilibrium

The LD structure among the SORL1 SNPs was examined using Haploview[5]. Haplotype blocks were defined using the confidence intervals algorithm. The default settings were used in these analyses, which create 95% confidence bounds on D′ to delineate SNP pairs in strong LD.

Results

There were 31 SORL1 SNPs in the TGEN database, and eight of those overlapped the 29 SNPs in the by Rogaeva et al study [1]. These 31 SNPs are referred to by their sequential order on the physical map in TGEN database, i.e. T.1, T.2, T.n, T.31. Therefore, a total of 52 unique SNPs were analyzed in these two studies (Table 1). All SNPs are in Hardy-Weinberg equlibrium in control samples. The LD structures of SNPs in the 5′ and 3′ regions are similar in the north European family data [1] and TGEN data [2] (Figure 1).

Table 1.

SORL1 SNP information in the original study and in the TGEN dataset.

SNP order dbSNP rs Number Physical Map Location (bp) Distance (bp) from Previous Marker SNP Type Orig order TGEN order NE FAD Israeli Arab TGEN
1 rs4935774 120,826,964 Upstream of 5′ UTR 1 T.1 0.20 0.18 0.39
2 rs578506 120,828,687 1,723 intron 2 0.07 ND
3 rs582446 120,833,069 4,382 intron 3 0.32 0.08
4 rs661057 120,834,164 1,095 intron 4 0.12 0.07
5 rs676160 120,845,443 11,279 intron T.2 0.49
6 rs11218304 120,854,321 8,878 intron 5 0.27 0.046
7 rs676759 120,864,475 10,154 intron T.3 0.57
8 rs560573 120,866,094 1,619 intron 6 T.4 0.64 0.29 0.55
9 rs985421 120,867,526 1,432 intron T.5 0.28
10 rs593769 120,869,213 1,687 intron T.6 0.57
11 rs12364988 120,872,836 3,623 H269H 7 T.7 0.15 0.041 0.33
12 rs668387 120,873,131 295 intron 8 T.8 0.52 0.021 0.32
13 rs689021 120,876,330 3,199 intron 9 0.71 0.040
14 rs641120 120,886,175 9,845 intron 10 0.57 0.07
15 rs11218313 120,888,081 1,906 intron T.9 0.83
16 rs4935775 120,894,712 6,631 intron 11 0.22 0.030
17 rs12285364 120,898,436 3,724 intron 12 0.014 0.56
18 rs2298813 120,898,894 458 T528A 13 0.012 0.43
19 rs11600231 120,911,918 13,024 intron 14 0.69 0.12
20 rs2276346 120,919,686 7,768 intron 15 T.10 0.61 0.36 0.25
21 rs10502262 120,920,522 836 intron T.11 0.17
22 SORL1-T833T 120,931,165 10,643 T833T 16 0.41 0.0026
23 rs556349 120,931,417 252 intron 17 0.0057 0.20
24 rs7131432 120,932,080 663 intron T.12 0.14
25 rs11218340 120,936,564 4,484 intron 18 0.08 0.14
26 rs10892756 120,939,766 3,202 intron T.13 0.49
27 rs11218346 120,944,391 4,625 intron T.14 0.48
28 rs1790213 120,947,399 3,008 intron T.15 0.33
29 rs11218347 120,951,758 4,359 intron T.16 0.78
30 rs2070045 120,953,300 1,542 S1187S 19 0.031 0.00082
31 rs3824966 120,953,393 93 intron 20 0.15 0.15
32 rs1699103 120,957,136 3,743 intron T.17 0.045
33 rs7116734 120,957,150 14 intron T.18 0.30
34 rs11218350 120,957,861 711 intron T.19 0.040
35 SORL1-18ex26 120,959,359 1,498 (−18) 5′ of exon 26 21 0.41 0.0091
36 rs1699102 120,962,172 2,813 N1246N 22 0.06 0.91
37 rs10892759 120,969,298 7,126 intron T.20 0.021
38 rs1792113 120,970,156 858 intron T.21 0.20
39 rs11218360 120,978,601 8,445 intron T.22 0.42
40 rs7128608 120,978,808 207 intron T.23 0.44
41 rs3824968 120,981,132 2,324 A1584A 23 0.0031 0.00073
42 rs1629493 120,982,306 1,174 intron T.24 0.09
43 rs2282649 120,984,168 1,862 intron 24 T.25 0.020 0.64 0.26
44 rs726601 120,986,617 2,449 intron T.26 0.07
45 rs1784931 120,988,148 1,531 intron T.27 0.049
46 rs1010159 120,988,611 463 intron 25 T.28 0.09 0.91 0.14
47 rs1784933 120,994,626 6,015 intron 26 0.07 0.80
48 rs1614735 120,998,211 3,585 intron 27 T.29 0.010 0.96 0.19
49 rs17125558 120,999,237 1,026 intron T.30 0.70
50 rs10892761 121,003,094 3,857 intron T.31 0.56
51 rs1133174 121,006,965 3,871 downstream of 3′UTR 28 0.07 0.09
52 rs1131497 121,007,955 990 downstream of 3′UTR 29 0.20 0.06
Open in a new tab

Notes: The nominal P values for allelic association with AD are shown for each dataset. NE FAD -- north European familial AD dataset; Israeli-Arab -- Israeli-Arab case-control dataset, and TGEN – TGEN dataset. Boldface rs number indicates that the SNP was genotyped in both Rogaeva et al. [1] and Reiman et al. [4] studies. Bold face p-value indicates that the SNP is nominally associated with AD in the original study. Bold face p-value with white font, black background refer to significant results from analyses of the TGEN data.

Figure 1.

Figure 1

Open in a new tab

Linkage disequilibrium (LD) structure of SORL1. Relative locations of SNPs included in each dataset are shown on two parallel stick diagrams, with LD maps for the TGEN dataset located above and the north European familial AD dataset studied by Rogaeva et al. [1] below the gene structure. The measure of LD (D′) among all possible pairs of SNPs is shown graphically according to the shade of red where white represents very low D′ and dark red represents very high D′. High D′ estimates associated with a large confidence interval (most likely due to one of the alleles being rare) are denoted by blue squares.

Six SNPs (T.17, T.19, T.20, T.21, T.26, T.27) showed nominally significant association (0.01 ≤ p <0.05) with AD under at least one model (Table 2). These six SNPs span a region of approximately 35 kb including SNPs 21–25 near the 3′ end of SORL1 which were strongly associated with AD in the Rogaeva et al. study. TGEN SNPs T.17 and T.19 are located between SNPs 20 and 21, TGEN SNPs T.20 and T.21 are between SNPs 22 and 23, and TGEN SNPs T.26 and T.27 are between SNPs 24 and 25 (Table 2). Haplotype analysis strengthened the association signal in the region including the two SNPs between SNPs 22 and 23 (global p-value = 0.005, data not shown).

Table 2.

Single-SNP association results in the TGEN dataset.

order SNP interval of SNPs in the original paper Minor Allele F_A F_U OR Allele.P Geno.P DOM.P REC.P TREND.P
T.1 rs4935774 1 1 0.215 0.201 1.09 0.39 0.54 0.54 0.29 0.38
T.2 rs676160 4~5 2 0.100 0.092 1.10 0.49 0.78 0.50 0.73 0.48
T.3 rs676759 5~6 2 0.389 0.378 1.05 0.57 0.35 0.88 0.19 0.57
T.4 rs560573 6 2 0.386 0.374 1.05 0.55 0.40 0.95 0.21 0.55
T.5 rs985421 6~7 2 0.016 0.022 0.74 0.28 NA NA NA 0.27
T.6 rs593769 6~7 1 0.385 0.374 1.05 0.57 0.48 0.98 0.25 0.57
T.7 rs12364988 7 2 0.496 0.477 1.08 0.33 0.56 0.61 0.28 0.34
T.8 rs668387 8 2 0.430 0.411 1.08 0.32 0.57 0.52 0.31 0.32
T.9 rs11218313 10~11 2 0.088 0.086 1.03 0.83 0.92 0.78 0.82 0.83
T.10 rs2276346 15 1 0.351 0.372 0.91 0.25 0.22 0.10 0.95 0.24
T.11 rs10502262 15~16 1 0.288 0.312 0.89 0.17 0.33 0.14 0.60 0.17
T.12 rs7131432 17~18 2 0.016 0.024 0.67 0.14 NA NA NA 0.14
T.13 rs10892756 18~19 1 0.075 0.068 1.11 0.49 0.40 0.65 0.18 0.49
T.14 rs11218346 18~19 1 0.085 0.078 1.11 0.48 0.60 0.55 0.37 0.47
T.15 rs1790213 18~19 2 0.375 0.393 0.93 0.33 0.34 0.17 0.99 0.33
T.16 rs11218347 18~19 1 0.076 0.073 1.04 0.78 0.94 0.81 0.78 0.78
T.17 rs1699103 20~21 1 0.404 0.442 0.85 0.045 0.13 0.07 0.14 0.043
T.18 rs7116734 20~21 1 0.399 0.419 0.92 0.30 0.42 0.61 0.19 0.29
T.19 rs11218350 20~21 2 0.215 0.248 0.83 0.040 0.11 0.11 0.07 0.045
T.20 rs10892759 22~23 1 0.325 0.368 0.83 0.021 0.040 0.011 0.35 0.020
T.21 rs1792113 22~23 1 0.351 0.375 0.90 0.20 0.06 0.041 0.68 0.21
T.22 rs11218360 22~23 1 0.027 0.032 0.83 0.42 NA NA NA 0.41
T.23 rs7128608 22~23 1 0.044 0.038 1.17 0.44 0.75 0.45 0.84 0.45
T.24 rs1629493 23~24 2 0.356 0.388 0.87 0.09 0.20 0.15 0.14 0.08
T.25 rs2282649 24 2 0.289 0.309 0.91 0.26 0.52 0.31 0.43 0.26
T.26 rs726601 24~25 2 0.296 0.329 0.86 0.07 0.12 0.038 0.56 0.06
T.27 rs1784931 24~25 1 0.377 0.415 0.86 0.049 0.12 0.049 0.24 0.046
T.28 rs1010159 25 1 0.336 0.363 0.89 0.14 0.32 0.21 0.24 0.14
T.29 rs1614735 27 2 0.479 0.505 0.90 0.19 0.27 0.11 0.58 0.18
T.30 rs17125558 27~28 2 0.020 0.022 0.90 0.70 NA NA NA 0.70
T.31 rs10892761 27~28 2 0.411 0.422 0.96 0.56 0.07 0.10 0.32 0.56
Open in a new tab

Notes: F_A: Minor allele frequency in cases; F_U: Minor allele frequency in controls; OR: estimated odds ratio for AD of minor allele; Allele.P: p-value for allelic association; Geno.P: p-value for genotypic (2 df) test; DOM.P: p-value for genotypic (1 df) test with dominant model; REC.P: p-value for genotypic (1 df) test with recessive model; TREND.P: Cochran-Armitage trend test; NA: Not available (allelic, genotypic and dominant/recessive tests require that each cell has a frequency of 1 or greater). SNPs studied by both Rogaeva et al. [1] and Reiman et al. [4] are indicated in boldface.

Discussion

The results from analysis of the TGEN data provide an independent replication of the association between AD and SORL1. We recognize that the magnitude of the significance of these results would not survive correction for multiple testing in a hypothesis-generating study (e.g., genome-wide association study); however a p-value of less than 0.05 is sufficient to confirm a previously reported association between variants within a specific gene and disease. Furthermore, although the particular SNP associations in this study are novel, a Bonferroni or similar correction is not appropriate or justified because all of the associated SNPs are within the region of and in LD with SNPs implicated in previous investigations[1–3], and therefore these are not truly independent tests.

Conclusion

While the precise identity of genetic effectors in SORL1 remains unknown, the TGEN data support the hypothesis that at least one disease risk enhancing allele is located near the 3′ end of SORL1.

Acknowledgments

This work was supported by grants from the US National Institute on Aging: R01-AG09029, R01-AG25259, R01-AG17173, and P30-AG13846 (LAF); R37-AG15473 and P01-AG07232 (RM); the Canadian Institutes of Health Research, Howard Hughes Medical Institute, Alzheimer Society of Ontario, Canada Foundation for Innovation, Ontario Research and Development Challenge Fund, Ontario Mental Health Foundation, Genome Canada (PSGH); the Blanchett Hooker Rockefeller Foundation and Charles S. Robertson Gift from the Banbury Fund (RM); and Fonds de la Recherche en Santé (YM). This research was conducted using the Boston University Linux Cluster for Genetic Analysis funded by an NCRR Shared Instrumentation grant (RR163736).

Footnotes

The authors declare that there are no conflicts of interest.

References

  • 1.Rogaeva E, Meng Y, Lee JH, Gu Y, Kawarai T, Zou F, et al. The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease. Nat Genet. 2007;39(2):168–177. doi: 10.1038/ng1943. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Lee JH, Cheng R, Schupf N, Manly J, Lantigua R, Stern Y, et al. The association between genetic variants in SORL1 and Alzheimer disease in an urban, multiethnic, community-based cohort. Arch Neurol. 2007;64(4):501–506. doi: 10.1001/archneur.64.4.501. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Lee JH, Cheng R, Schupf N, Honig LS, Vonsattel GJ, Clark L, et al. The Association Between Genetic Variants in SORL1 and Autopsy-Confirmed Alzheimer’s Disease. Neurology. 2007 doi: 10.1212/01.wnl.0000280581.39755.89. In press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Reiman EM, Webster JA, Myers AJ, Hardy J, Dunckley T, Zismann VL, et al. GAB2 alleles modify Alzheimer’s risk in APOE epsilon4 carriers. Neuron. 2007;54(5):713–720. doi: 10.1016/j.neuron.2007.05.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and visualization of LD and haplotype maps. Bioinformatics. 2005;21(2):263–265. doi: 10.1093/bioinformatics/bth457. [DOI] [PubMed] [Google Scholar]
  • 6.Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, et al. PLINK: A tools set for whole-genome association and population-based linkage analyses. Am J Hum Genet. 2007:81. doi: 10.1086/519795. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES