- Email: [email protected]
Implication of CLU gene polymorphisms in Chinese patients with Alzheimer's disease
Clinica Chimica Acta 411 (2010) 1516–1519
Contents lists available at ScienceDirect
Clinica Chimica Acta j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c l i n c h i m
Implication of CLU gene polymorphisms in Chinese patients with Alzheimer's disease Jin-Tai Yu a, Lu Li b, Qi-Xiu Zhu c, Qun Zhang a, Wei Zhang d, Zhong-Chen Wu a, Jun Guan e,⁎, Lan Tan a,⁎ a
Department of Neurology, Qingdao Municipal Hospital, School of Medicine, Qingdao University, Qingdao, Shandong Province 266071, PR China Shandong Medical College, Jinan, Shandong Province 250002, PR China c Department of Neurological Rehabilitation, The Affiliated Hospital of the Medical College of Qingdao University, Qingdao, Shandong Province 266071, PR China d Department of Emergency, Qingdao Municipal Hospital, School of Medicine, Qingdao University, Qingdao, Shandong Province 266071, PR China e Department of Cardiology, Qingdao Municipal Hospital, School of Medicine, Qingdao University, Qingdao, Shandong Province 266071, PR China b
a r t i c l e
i n f o
Article history: Received 23 May 2010 Received in revised form 15 June 2010 Accepted 15 June 2010 Available online 22 June 2010 Keywords: ClU Clusterin Apolipoprotein J Polymorphism Alzheimer's disease
a b s t r a c t Background: Clusterin (also called apolipoprotein J) has a potential central role in the pathogenesis of Alzheimer's disease (AD). Recently, two genome-wide association studies have identified three variants in CLU gene encoding clusterin associated with AD risk in Caucasians, while there are no studies on the association of CLU with AD risk in Asians. Methods: The study investigated 324 sporadic late-onset AD (LOAD) and 388 healthy controls matched for sex and age in a Han Chinese population. Three common genetic variants (rs2279590, rs11136000 and rs9331888) in CLU gene were genotyped using MALDI-TOF mass spectrometry. Results: The minor allele (G) of the rs9331888 polymorphism within CLU was significantly associated with an increased risk of LOAD (OR = 1.39, 95% CI = 1.13–1.72, P = 0.002). Logistic regression analysis revealed that the rs9331888 polymorphism presented strong associations with LOAD in the dominant, recessive and additive models. No significant differences in genotype and allele frequencies of the rs2279590 and rs11136000 polymorphisms were found between LOAD patients and controls. Haplotype analysis identified a risk haplotype (CCG) (OR = 1.66) and a protective haplotype (CCC)(OR = 0.70). Conclusions: Our findings implicate CLU as a susceptibility gene for LOAD in Han Chinese. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the leading cause of dementia in the elderly, with a prevalence of over 35 million worldwide [1]. About 1% of cases of AD is considered familial and shows autosomal dominant inheritance. Mutations in genes coding for amyloid precursor protein (APP), presenilin 1 (PSEN1), and presenilin 2 (PSEN2) proteins account for most cases of familial AD [2]. The more common late-onset sporadic AD (LOAD) is complex with numerous systems, cells and molecules affected. Several genes may interact to cause the disease or serve as risk factors [2]. The ε4 allele of apolipoprotein E (APOE) is the most prevalent identified risk factor and accounts for 40–50% of the risk as a single factor [2]. However, the ε4 allele is found in only around 38% (a lower frequency among Asian patients) of those with AD (relative to 14% of healthy subjects), and these genetic variants do not account for the full genetic risk for LOAD [3]. Aiming to identify new Alzheimer's disease loci, several genome-wide association studies (GWAS) have been previously conducted [2–4]. The most compelling novel non-
⁎ Corresponding authors. Tel.: + 86 532 8890 5659; fax: + 86 532 85968434. E-mail address: [email protected] (L. Tan). 0009-8981/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2010.06.013
APOE-related published GWAS signals have been observed in GAB2 (GRB2-associated binding protein 2), GAB2, followed by less consistently replicated signals in galanin-like peptide (GALP), piggyback transposable element derived 1 (PGBD1), tyrosine kinase, non-receptor 1 (TNK1) [2–4]. Recently, the two largest GWAS in AD have identified three single nucleotide polymorphisms (SNPs) (rs2279590, rs11136000 and rs9331888) in CLU gene as being associated with the risk of developing AD [5,6]. The attributive fractions of risk were to be 25.5% for APOE and 8.9% for CLU [5]. The original findings were replicated in a new large-scale GWAS [7]. Moreover, the association results for CLU are also supported by a recent meta-analysis of linkage studies of AD, in which a region of 15.1 Mb on chromosome 8p that encompasses CLU had the strongest evidence genome-wide for linkage [4], and several lines of evidence suggest that CLU is a strong candidate for involvement in the disease [8]. However, the other two case–control studies showed no evidence of association between CLU polymorphisms and AD risk [9,10]. Studies on the association of CLU polymorphisms with the risk of AD were performed primarily in Caucasians, while there are no similar studies in Asians by now. As variants and their frequencies of CLU are different in various genetically distinct racial/ethnic groups [10], replication is needed to confirm the potential effects of CLU in other groups, especially in Asians. The present study was conducted to assess the association of CLU polymorphisms (rs2279590, rs11136000 and rs9331888) with LOAD in a Han Chinese population.
J.-T. Yu et al. / Clinica Chimica Acta 411 (2010) 1516–1519
2. Materials and methods 2.1. Subjects Our study sample comprised of 324 sporadic late-onset AD (LOAD) (age at onset ≥65 years) patients (women 181; mean age: 76.87± 5.58 years) and 388 healthy controls (women 211; mean age: 75.93± 4.69 years) matched for gender and age. All of the above subjects were Northern Han Chinese in origin. The patients were recruited from the Department of Neurology of the Qingdao Municipal Hospital, and several other hospitals in Shandong Province. A clinical diagnosis of probable AD was established according to the criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS–ADRDA) [11]. No AD patient had a family history of dementia. The control groups were selected from the Health Examination Center of the Qingdao Municipal Hospital, and they were confirmed healthy and neurologically normal by medical history, general examinations, laboratory examinations and Mini Mental State Examination (MMSE) (score N28). Subjects with myocardial infarction, congestive heart failure, stroke, type 2 diabetes mellitus (T2DM) and atherosclerosis were excluded from this study. The presence of atherosclerosis of the carotid arteries and the lower extremities arteries were examined with color Doppler ultrasound sonography. An informed consent to participate in this study was obtained from each subject or from a guardian, and the protocol of this study was approved by the Ethical Committee of Qingdao Municipal Hospital. 2.2. Genotyping DNA was extracted from the peripheral blood using the Wizard genomic DNA purification kit (Cat. #A1125, Promega, USA). The polymorphisms at positions rs2279590, rs11136000 and rs9331888 within CLU gene were genotyped by Beijing Genomics Institute (BGI, Shenzhen, China) using matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) assay (MassArrayTM, Sequenom Inc., San Diego, CA, USA) under standard conditions with little modification according to Buetow et al. [12]. PCR and extension primers were designed using the Sequenom MassARRAY assay-design software (primer sequences available in Table 1). Details are available from the authors upon request. 10% randomly selected DNA samples from both patients and controls were sequenced to validate the genotyping by MALDI-TOF MS. Results of the MALDI-TOF MS method corresponded with the results of sequencing. Apolipoprotein E (APOE) genotypes were determined as the method described by Donohoe et al. [13]. 2.3. Statistical analysis Genotype and allele frequencies were estimated by counting. Hardy– Weinberg equilibrium between expected and observed genotype distributions was assessed using the Chi-square test. Genotype and allele distributions were compared using the Chi-square test. Differences in allele and genotype distributions between cases and controls were analyzed using logistic regression adjusted for age, gender, and APOE ε4 status under various genetic models that were defined as 1 (aa +Aa)
1517
versus 0 (AA) for dominant, 1 (aa) versus 0 (AA+Aa) for recessive, and 0 (AA) versus 1 (Aa) versus 2 (aa) for additive (A: major allele; a: minor allele). The P value, odds ratios (OR) and 95% confidence intervals (CI) were calculated. Estimation of the statistical power was performed with the STPLAN 4.3 software. Linkage disequilibrium (LD) and Haplotypic frequencies were estimated using SHEsis software (http://www.analysis. bio-x.cn/myAnalysis.php). Data were analyzed using a commercially available statistical package (SPSS Version 11.5, SPSS Inc., Chicago, IL). The criterion for significant difference is P b 0.05. 3. Results Allelic and genotype frequencies of rs2279590, rs11136000 and rs9331888 in LOAD patients and controls are reported in Table 2. Distributions of the CLU and APOE polymorphisms in both patients and controls did not deviate from those predicted by the Hardy–Weinberg equilibrium. As expected, higher frequencies of ApoE ε4 allele carriers were observed in patients with LOAD compared with control subjects (46.9% vs 22.2%, P b 0.001). We then tested the possible association of the CLU polymorphisms with LOAD susceptibility. A power analysis was added to calculate the sufficiency of sample size in this study. Based on the observed prevalence of the minor alleles in controls, our sample size had an 80% power to detect the odds ratio of 1.51 (or 0.69), 1.50 (or 0.70) and 1.35 (or 0.74) for rs2279590, rs11136000 and rs9331888, respectively, at a significance level (alpha) of 0.05. For rs9331888, there were significant differences in genotype and allele frequencies between LOAD and control (genotype P = 0.009, allele P = 0.002). Higher frequencies of the minor allele (G) carriers were observed in patients with LOAD compared with control subjects (OR = 1.39, 95% CI = 1.13–1.72, P = 0.002). In subjects without APOE ε4 allele, the allele and genotype distributions in rs9331888 between LOAD patients and controls remain significantly different (genotype P = 0.009, allele P b 0.001). In the subjects with APOE ε4 allele, no significant differences were observed (genotype P = 0.197, allele P = 0.284). For rs2279590, there were no significant differences in genotype and allele frequencies between LOAD and control (genotype P = 0.055, allele P = 0.157). However, when these data were stratified by the APOE ε4 status, significant association was found in subjects with APOE ε4 allele (genotype P = 0.036, allele P = 0.024). No significant difference was determined between genotype and allele frequencies in rs11136000 polymorphism when LOAD patients and controls were compared, even after stratification by APOE ε4 status. Furthermore, logistic regression revealed that rs9331888 polymorphism was still strongly associated with LOAD (dominant model: OR=1.651, 95% CI=1.12–2.42, P=0.011; recessive model: OR=1.65, 95% CI=1.14–2.37, P=0.007; additive model: OR=1.44, 95% CI=1.15– 1.81, P=0.002) after adjusting for age, gender, and the APOE ε4 status (Table 3). The three SNPs were located in a linkage disequilibrium (LD) block. LD between rs2279590 and rs9331888 (D′ = 0.999, r2 = 0.243) or between rs11136000 and rs9331888 (D′ = 0.999, r2 = 0.260) were stronger than that between rs2279590 and rs11136000 (D′ = 0.965, r2 = 0.870). They define three common haplotypes (frequency N2%) that together account for 98% of the observations in controls. The CCG haplotype occurred at greater frequencies in LOAD vs control (39.5%
Table 1 PCR and MassEXTEND primer for CLU polymorphisms. SNPs
PCR primers
MassEXTEND primers
rs2279590
5′-ACGTTGGATGGGGTCAGCTCTCTAGGTTTC-3′ 5′-ACGTTGGATGATGCGCTCTGCAACAGAAGT-3′ 5′-ACGTTGGATGGAATGGCAGGCATTCAGCAC-3′ 5′-ACGTTGGATGTATTGGGTCAAGTGGCAAGG-3′ 5′-ACGTTGGATGGCAGAGCAAGAGGACTCATC-3′ 5′-ACGTTGGATGAGTAGAGAGGGTTCGCAGTG-3′
5′-TAAGGAAGTCCTCCTGCT-3′
rs11136000 rs9331888
5′-CAAAGCCACACCAGCTATCA AAA-3′ 5′-GGACTCATCCTTCCAAA-3′
1518
J.-T. Yu et al. / Clinica Chimica Acta 411 (2010) 1516–1519
Table 2 Distribution of the CLU polymorphisms in LOAD cases and controls. n
rs2279590 LOAD Controls APOE ε4(−) LOAD Control APOE ε4(+) LOAD Control rs11136000 LOAD Control APOE ε4(−) LOAD Control APOE ε4(+) LOAD Control
rs9331888 LOAD Control APOE ε4(−) LOAD Control APOE ε4(+) LOAD Control a
Genotypes n(%)
Alleles n(%)
CC
CT
324 388
226(69.8) 258(66.5)
96(29.6) 118(30.4)
172 302
108(62.8) 202(66.9)
152 86
TT
P
C
T
P
2(0.6) 12(3.1)
0.055
548(84.6) 634(81.7)
100(15.4) 142(18.3)
0.157
62(36.0) 90(29.8)
2(1.2) 10(3.3)
0.164
278(80.8) 494(81.8)
66(19.2) 110(18.2)
0.711
118(77.6) 56(65.1)
34(22.4) 28(32.6)
0(0) 2(2.3)
0.036a
270(88.8) 140(81.4)
34(11.2) 32(18.6)
0.024
324 388
218(67.3) 250(64.4)
104(32.1) 126(32.5)
2(0.6) 12(3.1)
0.057
540(83.3) 626(80.7)
108(16.7) 150(19.3)
0.194
172 302
104(60.5) 194(64.2)
66(38.4) 98(32.5)
2(1.1) 10(3.3)
0.187
274(79.7) 486(80.5)
70(20.3) 118(19.5)
0.763
152 86
114(75.0) 56(65.1)
38(25.0) 28(32.6)
0(0) 2(2.3)
0.105a
266(87.5) 140(81.4)
38(12.5) 32(18.6)
0.071
CC
GC
GG
P
C
G
P
324 388
63(19.4) 110(28.4)
158(48.8) 184(47.4)
103(31.8) 94(24.2)
0.009
284(43.8) 404(52.1)
364(56.2) 372(47.9)
0.002
172 302
32(18.6) 84(27.8)
80(46.5) 148(49.0)
60(34.9) 70(23.2)
0.009
144(41.9) 316(52.3)
200(58.1) 288(47.7)
0.001
152 86
31(20.4) 26(30.2)
78(51.3) 36(41.9)
43(28.3) 24(27.9)
0.197
140(46.1) 88(51.2)
164(53.9) 84(48.8)
0.284
As 2 cells have expected count less than 5, CC versus CT + TT analysis were conducted.
vs 28.1%, OR = 1.66, 95% CI = 1.33–2.07, P b 0.001), suggesting that it is possible risk factors for LOAD, while CCC haplotype was at lower frequencies in LOAD vs control (43.8% vs 52.1%, OR = 0.70, 95% CI = 0.57–0.81, P = 0.001) (Table 4).
4. Discussion CLU is a single copy gene, located on human chromosome 8p21-p12 [14], a chromosomal region of interest in LOAD defined by the genomewide linkage study [15]. Butler et al. have pooled linkage results from five independent genome scans and applied the genome search meta-analysis (GSMA) method to linkage studies for LOAD [4]. The GSMA finding suggests that the chromosome 8p region achieved the lowest summed rank P value of 0.001 as a locus harbouring susceptibility genes for LOAD. Recently, two large genome-wide association studies in Alzheimer's Disease (AD) have identified variants in three different genes (CLU,
Table 3 Adjusted OR estimated by multivariate logistic regression analysis. SNPs
Model
OR (95% CI)
Wald
P value
rs2279590
Dom Rec Add Dom Rec Add Dom Rec Add
0.87(0.61–1.24) 0.36(0.08–1.67) 0.84(0.61–1.16) 0.89(0.63–1.25) 0.36(0.08–1.67) 0.85(0.62–1.17) 1.65(1.12–2.42) 1.65(1.14–2.37) 1.44(1.15–1.81)
0.61 1.71 1.14 0.47 1.71 0.97 6.49 7.16 10.03
0.437 0.191 0.285 0.492 0.191 0.325 0.011 0.007 0.002
rs11136000
rs9331888
Adjusted for age, gender, and the carriage of at least one APOE ε4 allele. Dom, dominant model; Rec, recessive model; Add, additive model; OR, odds ratio; CI, confidence interval.
PICALM and CR1) as being associated with the risk of developing AD [5,6]. The strongest association was reported for SNPS in CLU. In Lambert's study, three SNPs (rs2279590, rs11136000, rs9331888) within CLU gene showed statistically significant association with AD (OR: 0.86, 0.86 and 1.12, respectively) [5]. In Harold's study, they identified that the intronic SNP (rs11136000) is associated with the development of AD (OR=0.84) [6]. The association between rs111360001 polymorphisms was replicated by a new GWAS conducted by Seshadri et al. (0R=0.90) [7]. However, the other two case–control studies failed to find evidence of association between CLU and AD risk [9,10]. As a smaller real effect size than that reported from GWAS, the disparities possibly arise from that these two case–control studies have no sufficient power to detect very small genetic effects. Here, we report for the first time the significant association of the rs9331888 SNP in CLU with an increased risk of LOAD in a Han Chinese population. This association was evident in non-carriers of the APOE ε4 allele, but not in its carriers. Logistic regression analysis revealed that the rs9331888 polymorphism presented strong associations with LOAD in the dominant, recessive and additive models. Haplotype analysis identified a risk haplotype (CCG) and a protective haplotype (CCC). In Caucasian historical controls [5–7], the minor allele frequencies of the rs2279590, rs11136000 and rs9331888 SNPs within CLU were 41%, 36%–40% and 28%, respectively, while in our study, the minor allele
Table 4 Estimated haplotype frequencies of CLU in LOAD cases and controls. Haplotypes
Cases
Controls
χ2
OR (95% CI)
P
TTG CCC CCG Global test
0.154 0.438 0.395
0.178 0.521 0.281
1.573 10.753 19.888
0.84(0.63–1.11) 0.70(0.57–0.81) 1.66(1.33–2.07)
0.210 0.001 b 0.001 b 0.001
The SNPs as ordered from left to right (5′ to 3′) are rs2279590, rs11136000 and rs9331888. Minor alleles are underlined.
J.-T. Yu et al. / Clinica Chimica Acta 411 (2010) 1516–1519
frequencies of these three SNPs within CLU in controls were 18.3%, 19.3% and 47.9%, respectively. The minor allele frequencies of the rs2279590, rs11136000 and rs9331888 within CLU in our healthy controls were similar to those in the Han Chinese population from the dbSNP database (21.1%, 23.3% and 41.1%, respectively) (http://www.ncbi.nlm.nih.gov/ SNP/) and significantly different from the Caucasian historical controls. This disparity between our findings and those of the historical studies might also be attributable to the different genetic backgrounds of the Chinese Han and Caucasian populations. Moreover, the study does not have sufficient power to detect very small genetic effects. A smaller real effect size than the originally reported from GWAS, (which in turn would have provoked an underestimation of sample size calculations of replication studies), could also partially explain the observed heterogeneity in CLU association between our study and the studies with Caucasian patients. CLU encodes ‘clusterin’ (also known as apolipoprotein J), which is a lipoprotein expressed in most mammalian tissues with higher levels in the brain than in many other tissues [16]. Clusterin possesses several characteristic properties that can connect it to the pathology of Alzheimer's Disease [8], i.e. (i) it binds avidly to amyloid-β oligomers and prevent their fibrillization and is present in neuritic plaques, (ii) it is involved in the clearance of amyloid-β peptides and fibrils by binding to megalin receptors and enhancing their endocytosis within glial cells in the brain, (iii) it is a complement inhibitor and can suppress complement activation observed in AD, (iv) it is included in lipid particles and thus affects cholesterol transport in conjunction with apolipoprotein E, (v) it is a stress-induced chaperone which can be transported to cytoplasm where it can bind to Bax protein and inhibit neuronal apoptosis, and (vi) it can also bind to Smad2/3 proteins and potentiate the neuroprotective TGFβ signaling. These biological properties indicate that clusterin could have beneficial, preventive effects on the AD pathology. Here, our data demonstrated that the CLU rs9331888 polymorphism is significantly associated with increased risk of AD in Han Chinese. The associated SNP (rs9331888) is located in the five prime untranslated regions (5′-UTR) of CLU gene. Thus, this polymorphism does not modify the encoded protein directly. However, 5′-UTR variants may potentially influence the protein expression levels and therefore the disease susceptibility [17]. This is potentially very important as it suggests that a therapy designed to counteract the functional susceptibility of CLU may be beneficial even when administered to mature individuals. So, further experimental research is necessary to define the direct functional association between CLU rs9331888 polymorphism and the occurrence of LOAD. In summary, our study suggests that genetic variations in the CLU gene might modify the risk for LOAD in Han Chinese. Future study
1519
should focus on detailed functional research of rs9331888, which will provide stronger evidence to support our findings. Acknowledgements We are grateful to all of the subjects who kindly agreed to participate in this study. This study was supported by grants from the National Natural Science Foundation of China (30870884) and the Shandong Provincial Outstanding Medical Academic Professional Program. References [1] Querfurth HW, LaFerla FM. Alzheimer's disease. N Engl J Med 2010;362:329–44. [2] Bertram L, Tanzi RE. Thirty years of Alzheimer's disease genetics: the implications of systematic meta-analyses. Nat Rev Neurosci 2008;9:768–78. [3] Bertram L, McQueen MB, Mullin K, et al. Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat Genet 2007;39:17–23. [4] Butler AW, Ng MY, Hamshere ML, et al. Meta-analysis of linkage studies for Alzheimer's disease — a web resource. Neurobiol Aging 2009;30:1037–47. [5] Lambert JC, Heath S, Even G, et al. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease. Nat Genet 2009;41:1094–9. [6] Harold D, Abraham R, Hollingworth P, et al. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nat Genet 2009;41:1088–93. [7] Seshadri S, Fitzpatrick AL, Ikram MA, et al. Genome-wide analysis of genetic loci associated with Alzheimer disease. JAMA 2010;303:1832–40. [8] Nuutinen T, Suuronen T, Kauppinen A, et al. Clusterin: a forgotten player in Alzheimer's disease. Brain Res Rev 2009;61:89–104. [9] Guerreiro RJ, Beck J, Gibbs JR, et al. Genetic variability in CLU and its association with Alzheimer's disease. PLoS ONE 2010;5:e9510. [10] Tycko B, Feng L, Nguyen L, et al. Polymorphisms in the human apolipoprotein-J/ clusterin gene: ethnic variation and distribution in Alzheimer's disease. Hum Genet 1996;98:430–6. [11] McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer's disease: report of the NINCDSADRDA work group under the auspices of Department of Health and Human Services task force on Alzheimer's disease. Neurology 1984;34:939–44. [12] Buetow KH, Edmonson M, MacDonald R, et al. High-throughput development and characterization of a genomewide collection of gene-based single nucleotide polymorphism markers by chip-based matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Proc Natl Acad Sci USA 2001;98:581–4. [13] Donohoe GG, Salomäki A, Lehtimäki T, et al. Rapid identification of apolipoprotein E genotypes by multiplex amplification refractory mutation system PCR and capillary gel electrophoresis. Clin Chem 1999;45:143–6. [14] Wong P, Taillefer D, Lakins J, et al. Molecular characterization of human TRPM-2/ clusterin, a gene associated with sperm maturation, apoptosis and neurodegeneration. Eur J Biochem 1994;221:917–25. [15] Lee JH, Cheng R, Graff-Radford N, et al. Analyses of the National Institute on Aging Late-Onset Alzheimer's Disease Family Study: implication of additional loci. Arch Neurol 2008;65:1518–26. [16] de Silva HV, Harmony JA, Stuart WD, et al. Apolipoprotein J: structure and tissue distribution. Biochemistry 1990;29:5380–9. [17] Yan H, Yuan W, Velculescu VE, et al. Allelic variation in human gene expression. Science 2002;297:1143.
Contents lists available at ScienceDirect
Clinica Chimica Acta j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c l i n c h i m
Implication of CLU gene polymorphisms in Chinese patients with Alzheimer's disease Jin-Tai Yu a, Lu Li b, Qi-Xiu Zhu c, Qun Zhang a, Wei Zhang d, Zhong-Chen Wu a, Jun Guan e,⁎, Lan Tan a,⁎ a
Department of Neurology, Qingdao Municipal Hospital, School of Medicine, Qingdao University, Qingdao, Shandong Province 266071, PR China Shandong Medical College, Jinan, Shandong Province 250002, PR China c Department of Neurological Rehabilitation, The Affiliated Hospital of the Medical College of Qingdao University, Qingdao, Shandong Province 266071, PR China d Department of Emergency, Qingdao Municipal Hospital, School of Medicine, Qingdao University, Qingdao, Shandong Province 266071, PR China e Department of Cardiology, Qingdao Municipal Hospital, School of Medicine, Qingdao University, Qingdao, Shandong Province 266071, PR China b
a r t i c l e
i n f o
Article history: Received 23 May 2010 Received in revised form 15 June 2010 Accepted 15 June 2010 Available online 22 June 2010 Keywords: ClU Clusterin Apolipoprotein J Polymorphism Alzheimer's disease
a b s t r a c t Background: Clusterin (also called apolipoprotein J) has a potential central role in the pathogenesis of Alzheimer's disease (AD). Recently, two genome-wide association studies have identified three variants in CLU gene encoding clusterin associated with AD risk in Caucasians, while there are no studies on the association of CLU with AD risk in Asians. Methods: The study investigated 324 sporadic late-onset AD (LOAD) and 388 healthy controls matched for sex and age in a Han Chinese population. Three common genetic variants (rs2279590, rs11136000 and rs9331888) in CLU gene were genotyped using MALDI-TOF mass spectrometry. Results: The minor allele (G) of the rs9331888 polymorphism within CLU was significantly associated with an increased risk of LOAD (OR = 1.39, 95% CI = 1.13–1.72, P = 0.002). Logistic regression analysis revealed that the rs9331888 polymorphism presented strong associations with LOAD in the dominant, recessive and additive models. No significant differences in genotype and allele frequencies of the rs2279590 and rs11136000 polymorphisms were found between LOAD patients and controls. Haplotype analysis identified a risk haplotype (CCG) (OR = 1.66) and a protective haplotype (CCC)(OR = 0.70). Conclusions: Our findings implicate CLU as a susceptibility gene for LOAD in Han Chinese. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Alzheimer's disease (AD) is a progressive neurodegenerative disorder and the leading cause of dementia in the elderly, with a prevalence of over 35 million worldwide [1]. About 1% of cases of AD is considered familial and shows autosomal dominant inheritance. Mutations in genes coding for amyloid precursor protein (APP), presenilin 1 (PSEN1), and presenilin 2 (PSEN2) proteins account for most cases of familial AD [2]. The more common late-onset sporadic AD (LOAD) is complex with numerous systems, cells and molecules affected. Several genes may interact to cause the disease or serve as risk factors [2]. The ε4 allele of apolipoprotein E (APOE) is the most prevalent identified risk factor and accounts for 40–50% of the risk as a single factor [2]. However, the ε4 allele is found in only around 38% (a lower frequency among Asian patients) of those with AD (relative to 14% of healthy subjects), and these genetic variants do not account for the full genetic risk for LOAD [3]. Aiming to identify new Alzheimer's disease loci, several genome-wide association studies (GWAS) have been previously conducted [2–4]. The most compelling novel non-
⁎ Corresponding authors. Tel.: + 86 532 8890 5659; fax: + 86 532 85968434. E-mail address: [email protected] (L. Tan). 0009-8981/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2010.06.013
APOE-related published GWAS signals have been observed in GAB2 (GRB2-associated binding protein 2), GAB2, followed by less consistently replicated signals in galanin-like peptide (GALP), piggyback transposable element derived 1 (PGBD1), tyrosine kinase, non-receptor 1 (TNK1) [2–4]. Recently, the two largest GWAS in AD have identified three single nucleotide polymorphisms (SNPs) (rs2279590, rs11136000 and rs9331888) in CLU gene as being associated with the risk of developing AD [5,6]. The attributive fractions of risk were to be 25.5% for APOE and 8.9% for CLU [5]. The original findings were replicated in a new large-scale GWAS [7]. Moreover, the association results for CLU are also supported by a recent meta-analysis of linkage studies of AD, in which a region of 15.1 Mb on chromosome 8p that encompasses CLU had the strongest evidence genome-wide for linkage [4], and several lines of evidence suggest that CLU is a strong candidate for involvement in the disease [8]. However, the other two case–control studies showed no evidence of association between CLU polymorphisms and AD risk [9,10]. Studies on the association of CLU polymorphisms with the risk of AD were performed primarily in Caucasians, while there are no similar studies in Asians by now. As variants and their frequencies of CLU are different in various genetically distinct racial/ethnic groups [10], replication is needed to confirm the potential effects of CLU in other groups, especially in Asians. The present study was conducted to assess the association of CLU polymorphisms (rs2279590, rs11136000 and rs9331888) with LOAD in a Han Chinese population.
J.-T. Yu et al. / Clinica Chimica Acta 411 (2010) 1516–1519
2. Materials and methods 2.1. Subjects Our study sample comprised of 324 sporadic late-onset AD (LOAD) (age at onset ≥65 years) patients (women 181; mean age: 76.87± 5.58 years) and 388 healthy controls (women 211; mean age: 75.93± 4.69 years) matched for gender and age. All of the above subjects were Northern Han Chinese in origin. The patients were recruited from the Department of Neurology of the Qingdao Municipal Hospital, and several other hospitals in Shandong Province. A clinical diagnosis of probable AD was established according to the criteria of the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer's Disease and Related Disorders Association (NINCDS–ADRDA) [11]. No AD patient had a family history of dementia. The control groups were selected from the Health Examination Center of the Qingdao Municipal Hospital, and they were confirmed healthy and neurologically normal by medical history, general examinations, laboratory examinations and Mini Mental State Examination (MMSE) (score N28). Subjects with myocardial infarction, congestive heart failure, stroke, type 2 diabetes mellitus (T2DM) and atherosclerosis were excluded from this study. The presence of atherosclerosis of the carotid arteries and the lower extremities arteries were examined with color Doppler ultrasound sonography. An informed consent to participate in this study was obtained from each subject or from a guardian, and the protocol of this study was approved by the Ethical Committee of Qingdao Municipal Hospital. 2.2. Genotyping DNA was extracted from the peripheral blood using the Wizard genomic DNA purification kit (Cat. #A1125, Promega, USA). The polymorphisms at positions rs2279590, rs11136000 and rs9331888 within CLU gene were genotyped by Beijing Genomics Institute (BGI, Shenzhen, China) using matrix-assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) assay (MassArrayTM, Sequenom Inc., San Diego, CA, USA) under standard conditions with little modification according to Buetow et al. [12]. PCR and extension primers were designed using the Sequenom MassARRAY assay-design software (primer sequences available in Table 1). Details are available from the authors upon request. 10% randomly selected DNA samples from both patients and controls were sequenced to validate the genotyping by MALDI-TOF MS. Results of the MALDI-TOF MS method corresponded with the results of sequencing. Apolipoprotein E (APOE) genotypes were determined as the method described by Donohoe et al. [13]. 2.3. Statistical analysis Genotype and allele frequencies were estimated by counting. Hardy– Weinberg equilibrium between expected and observed genotype distributions was assessed using the Chi-square test. Genotype and allele distributions were compared using the Chi-square test. Differences in allele and genotype distributions between cases and controls were analyzed using logistic regression adjusted for age, gender, and APOE ε4 status under various genetic models that were defined as 1 (aa +Aa)
1517
versus 0 (AA) for dominant, 1 (aa) versus 0 (AA+Aa) for recessive, and 0 (AA) versus 1 (Aa) versus 2 (aa) for additive (A: major allele; a: minor allele). The P value, odds ratios (OR) and 95% confidence intervals (CI) were calculated. Estimation of the statistical power was performed with the STPLAN 4.3 software. Linkage disequilibrium (LD) and Haplotypic frequencies were estimated using SHEsis software (http://www.analysis. bio-x.cn/myAnalysis.php). Data were analyzed using a commercially available statistical package (SPSS Version 11.5, SPSS Inc., Chicago, IL). The criterion for significant difference is P b 0.05. 3. Results Allelic and genotype frequencies of rs2279590, rs11136000 and rs9331888 in LOAD patients and controls are reported in Table 2. Distributions of the CLU and APOE polymorphisms in both patients and controls did not deviate from those predicted by the Hardy–Weinberg equilibrium. As expected, higher frequencies of ApoE ε4 allele carriers were observed in patients with LOAD compared with control subjects (46.9% vs 22.2%, P b 0.001). We then tested the possible association of the CLU polymorphisms with LOAD susceptibility. A power analysis was added to calculate the sufficiency of sample size in this study. Based on the observed prevalence of the minor alleles in controls, our sample size had an 80% power to detect the odds ratio of 1.51 (or 0.69), 1.50 (or 0.70) and 1.35 (or 0.74) for rs2279590, rs11136000 and rs9331888, respectively, at a significance level (alpha) of 0.05. For rs9331888, there were significant differences in genotype and allele frequencies between LOAD and control (genotype P = 0.009, allele P = 0.002). Higher frequencies of the minor allele (G) carriers were observed in patients with LOAD compared with control subjects (OR = 1.39, 95% CI = 1.13–1.72, P = 0.002). In subjects without APOE ε4 allele, the allele and genotype distributions in rs9331888 between LOAD patients and controls remain significantly different (genotype P = 0.009, allele P b 0.001). In the subjects with APOE ε4 allele, no significant differences were observed (genotype P = 0.197, allele P = 0.284). For rs2279590, there were no significant differences in genotype and allele frequencies between LOAD and control (genotype P = 0.055, allele P = 0.157). However, when these data were stratified by the APOE ε4 status, significant association was found in subjects with APOE ε4 allele (genotype P = 0.036, allele P = 0.024). No significant difference was determined between genotype and allele frequencies in rs11136000 polymorphism when LOAD patients and controls were compared, even after stratification by APOE ε4 status. Furthermore, logistic regression revealed that rs9331888 polymorphism was still strongly associated with LOAD (dominant model: OR=1.651, 95% CI=1.12–2.42, P=0.011; recessive model: OR=1.65, 95% CI=1.14–2.37, P=0.007; additive model: OR=1.44, 95% CI=1.15– 1.81, P=0.002) after adjusting for age, gender, and the APOE ε4 status (Table 3). The three SNPs were located in a linkage disequilibrium (LD) block. LD between rs2279590 and rs9331888 (D′ = 0.999, r2 = 0.243) or between rs11136000 and rs9331888 (D′ = 0.999, r2 = 0.260) were stronger than that between rs2279590 and rs11136000 (D′ = 0.965, r2 = 0.870). They define three common haplotypes (frequency N2%) that together account for 98% of the observations in controls. The CCG haplotype occurred at greater frequencies in LOAD vs control (39.5%
Table 1 PCR and MassEXTEND primer for CLU polymorphisms. SNPs
PCR primers
MassEXTEND primers
rs2279590
5′-ACGTTGGATGGGGTCAGCTCTCTAGGTTTC-3′ 5′-ACGTTGGATGATGCGCTCTGCAACAGAAGT-3′ 5′-ACGTTGGATGGAATGGCAGGCATTCAGCAC-3′ 5′-ACGTTGGATGTATTGGGTCAAGTGGCAAGG-3′ 5′-ACGTTGGATGGCAGAGCAAGAGGACTCATC-3′ 5′-ACGTTGGATGAGTAGAGAGGGTTCGCAGTG-3′
5′-TAAGGAAGTCCTCCTGCT-3′
rs11136000 rs9331888
5′-CAAAGCCACACCAGCTATCA AAA-3′ 5′-GGACTCATCCTTCCAAA-3′
1518
J.-T. Yu et al. / Clinica Chimica Acta 411 (2010) 1516–1519
Table 2 Distribution of the CLU polymorphisms in LOAD cases and controls. n
rs2279590 LOAD Controls APOE ε4(−) LOAD Control APOE ε4(+) LOAD Control rs11136000 LOAD Control APOE ε4(−) LOAD Control APOE ε4(+) LOAD Control
rs9331888 LOAD Control APOE ε4(−) LOAD Control APOE ε4(+) LOAD Control a
Genotypes n(%)
Alleles n(%)
CC
CT
324 388
226(69.8) 258(66.5)
96(29.6) 118(30.4)
172 302
108(62.8) 202(66.9)
152 86
TT
P
C
T
P
2(0.6) 12(3.1)
0.055
548(84.6) 634(81.7)
100(15.4) 142(18.3)
0.157
62(36.0) 90(29.8)
2(1.2) 10(3.3)
0.164
278(80.8) 494(81.8)
66(19.2) 110(18.2)
0.711
118(77.6) 56(65.1)
34(22.4) 28(32.6)
0(0) 2(2.3)
0.036a
270(88.8) 140(81.4)
34(11.2) 32(18.6)
0.024
324 388
218(67.3) 250(64.4)
104(32.1) 126(32.5)
2(0.6) 12(3.1)
0.057
540(83.3) 626(80.7)
108(16.7) 150(19.3)
0.194
172 302
104(60.5) 194(64.2)
66(38.4) 98(32.5)
2(1.1) 10(3.3)
0.187
274(79.7) 486(80.5)
70(20.3) 118(19.5)
0.763
152 86
114(75.0) 56(65.1)
38(25.0) 28(32.6)
0(0) 2(2.3)
0.105a
266(87.5) 140(81.4)
38(12.5) 32(18.6)
0.071
CC
GC
GG
P
C
G
P
324 388
63(19.4) 110(28.4)
158(48.8) 184(47.4)
103(31.8) 94(24.2)
0.009
284(43.8) 404(52.1)
364(56.2) 372(47.9)
0.002
172 302
32(18.6) 84(27.8)
80(46.5) 148(49.0)
60(34.9) 70(23.2)
0.009
144(41.9) 316(52.3)
200(58.1) 288(47.7)
0.001
152 86
31(20.4) 26(30.2)
78(51.3) 36(41.9)
43(28.3) 24(27.9)
0.197
140(46.1) 88(51.2)
164(53.9) 84(48.8)
0.284
As 2 cells have expected count less than 5, CC versus CT + TT analysis were conducted.
vs 28.1%, OR = 1.66, 95% CI = 1.33–2.07, P b 0.001), suggesting that it is possible risk factors for LOAD, while CCC haplotype was at lower frequencies in LOAD vs control (43.8% vs 52.1%, OR = 0.70, 95% CI = 0.57–0.81, P = 0.001) (Table 4).
4. Discussion CLU is a single copy gene, located on human chromosome 8p21-p12 [14], a chromosomal region of interest in LOAD defined by the genomewide linkage study [15]. Butler et al. have pooled linkage results from five independent genome scans and applied the genome search meta-analysis (GSMA) method to linkage studies for LOAD [4]. The GSMA finding suggests that the chromosome 8p region achieved the lowest summed rank P value of 0.001 as a locus harbouring susceptibility genes for LOAD. Recently, two large genome-wide association studies in Alzheimer's Disease (AD) have identified variants in three different genes (CLU,
Table 3 Adjusted OR estimated by multivariate logistic regression analysis. SNPs
Model
OR (95% CI)
Wald
P value
rs2279590
Dom Rec Add Dom Rec Add Dom Rec Add
0.87(0.61–1.24) 0.36(0.08–1.67) 0.84(0.61–1.16) 0.89(0.63–1.25) 0.36(0.08–1.67) 0.85(0.62–1.17) 1.65(1.12–2.42) 1.65(1.14–2.37) 1.44(1.15–1.81)
0.61 1.71 1.14 0.47 1.71 0.97 6.49 7.16 10.03
0.437 0.191 0.285 0.492 0.191 0.325 0.011 0.007 0.002
rs11136000
rs9331888
Adjusted for age, gender, and the carriage of at least one APOE ε4 allele. Dom, dominant model; Rec, recessive model; Add, additive model; OR, odds ratio; CI, confidence interval.
PICALM and CR1) as being associated with the risk of developing AD [5,6]. The strongest association was reported for SNPS in CLU. In Lambert's study, three SNPs (rs2279590, rs11136000, rs9331888) within CLU gene showed statistically significant association with AD (OR: 0.86, 0.86 and 1.12, respectively) [5]. In Harold's study, they identified that the intronic SNP (rs11136000) is associated with the development of AD (OR=0.84) [6]. The association between rs111360001 polymorphisms was replicated by a new GWAS conducted by Seshadri et al. (0R=0.90) [7]. However, the other two case–control studies failed to find evidence of association between CLU and AD risk [9,10]. As a smaller real effect size than that reported from GWAS, the disparities possibly arise from that these two case–control studies have no sufficient power to detect very small genetic effects. Here, we report for the first time the significant association of the rs9331888 SNP in CLU with an increased risk of LOAD in a Han Chinese population. This association was evident in non-carriers of the APOE ε4 allele, but not in its carriers. Logistic regression analysis revealed that the rs9331888 polymorphism presented strong associations with LOAD in the dominant, recessive and additive models. Haplotype analysis identified a risk haplotype (CCG) and a protective haplotype (CCC). In Caucasian historical controls [5–7], the minor allele frequencies of the rs2279590, rs11136000 and rs9331888 SNPs within CLU were 41%, 36%–40% and 28%, respectively, while in our study, the minor allele
Table 4 Estimated haplotype frequencies of CLU in LOAD cases and controls. Haplotypes
Cases
Controls
χ2
OR (95% CI)
P
TTG CCC CCG Global test
0.154 0.438 0.395
0.178 0.521 0.281
1.573 10.753 19.888
0.84(0.63–1.11) 0.70(0.57–0.81) 1.66(1.33–2.07)
0.210 0.001 b 0.001 b 0.001
The SNPs as ordered from left to right (5′ to 3′) are rs2279590, rs11136000 and rs9331888. Minor alleles are underlined.
J.-T. Yu et al. / Clinica Chimica Acta 411 (2010) 1516–1519
frequencies of these three SNPs within CLU in controls were 18.3%, 19.3% and 47.9%, respectively. The minor allele frequencies of the rs2279590, rs11136000 and rs9331888 within CLU in our healthy controls were similar to those in the Han Chinese population from the dbSNP database (21.1%, 23.3% and 41.1%, respectively) (http://www.ncbi.nlm.nih.gov/ SNP/) and significantly different from the Caucasian historical controls. This disparity between our findings and those of the historical studies might also be attributable to the different genetic backgrounds of the Chinese Han and Caucasian populations. Moreover, the study does not have sufficient power to detect very small genetic effects. A smaller real effect size than the originally reported from GWAS, (which in turn would have provoked an underestimation of sample size calculations of replication studies), could also partially explain the observed heterogeneity in CLU association between our study and the studies with Caucasian patients. CLU encodes ‘clusterin’ (also known as apolipoprotein J), which is a lipoprotein expressed in most mammalian tissues with higher levels in the brain than in many other tissues [16]. Clusterin possesses several characteristic properties that can connect it to the pathology of Alzheimer's Disease [8], i.e. (i) it binds avidly to amyloid-β oligomers and prevent their fibrillization and is present in neuritic plaques, (ii) it is involved in the clearance of amyloid-β peptides and fibrils by binding to megalin receptors and enhancing their endocytosis within glial cells in the brain, (iii) it is a complement inhibitor and can suppress complement activation observed in AD, (iv) it is included in lipid particles and thus affects cholesterol transport in conjunction with apolipoprotein E, (v) it is a stress-induced chaperone which can be transported to cytoplasm where it can bind to Bax protein and inhibit neuronal apoptosis, and (vi) it can also bind to Smad2/3 proteins and potentiate the neuroprotective TGFβ signaling. These biological properties indicate that clusterin could have beneficial, preventive effects on the AD pathology. Here, our data demonstrated that the CLU rs9331888 polymorphism is significantly associated with increased risk of AD in Han Chinese. The associated SNP (rs9331888) is located in the five prime untranslated regions (5′-UTR) of CLU gene. Thus, this polymorphism does not modify the encoded protein directly. However, 5′-UTR variants may potentially influence the protein expression levels and therefore the disease susceptibility [17]. This is potentially very important as it suggests that a therapy designed to counteract the functional susceptibility of CLU may be beneficial even when administered to mature individuals. So, further experimental research is necessary to define the direct functional association between CLU rs9331888 polymorphism and the occurrence of LOAD. In summary, our study suggests that genetic variations in the CLU gene might modify the risk for LOAD in Han Chinese. Future study
1519
should focus on detailed functional research of rs9331888, which will provide stronger evidence to support our findings. Acknowledgements We are grateful to all of the subjects who kindly agreed to participate in this study. This study was supported by grants from the National Natural Science Foundation of China (30870884) and the Shandong Provincial Outstanding Medical Academic Professional Program. References [1] Querfurth HW, LaFerla FM. Alzheimer's disease. N Engl J Med 2010;362:329–44. [2] Bertram L, Tanzi RE. Thirty years of Alzheimer's disease genetics: the implications of systematic meta-analyses. Nat Rev Neurosci 2008;9:768–78. [3] Bertram L, McQueen MB, Mullin K, et al. Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database. Nat Genet 2007;39:17–23. [4] Butler AW, Ng MY, Hamshere ML, et al. Meta-analysis of linkage studies for Alzheimer's disease — a web resource. Neurobiol Aging 2009;30:1037–47. [5] Lambert JC, Heath S, Even G, et al. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer's disease. Nat Genet 2009;41:1094–9. [6] Harold D, Abraham R, Hollingworth P, et al. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer's disease. Nat Genet 2009;41:1088–93. [7] Seshadri S, Fitzpatrick AL, Ikram MA, et al. Genome-wide analysis of genetic loci associated with Alzheimer disease. JAMA 2010;303:1832–40. [8] Nuutinen T, Suuronen T, Kauppinen A, et al. Clusterin: a forgotten player in Alzheimer's disease. Brain Res Rev 2009;61:89–104. [9] Guerreiro RJ, Beck J, Gibbs JR, et al. Genetic variability in CLU and its association with Alzheimer's disease. PLoS ONE 2010;5:e9510. [10] Tycko B, Feng L, Nguyen L, et al. Polymorphisms in the human apolipoprotein-J/ clusterin gene: ethnic variation and distribution in Alzheimer's disease. Hum Genet 1996;98:430–6. [11] McKhann G, Drachman D, Folstein M, et al. Clinical diagnosis of Alzheimer's disease: report of the NINCDSADRDA work group under the auspices of Department of Health and Human Services task force on Alzheimer's disease. Neurology 1984;34:939–44. [12] Buetow KH, Edmonson M, MacDonald R, et al. High-throughput development and characterization of a genomewide collection of gene-based single nucleotide polymorphism markers by chip-based matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. Proc Natl Acad Sci USA 2001;98:581–4. [13] Donohoe GG, Salomäki A, Lehtimäki T, et al. Rapid identification of apolipoprotein E genotypes by multiplex amplification refractory mutation system PCR and capillary gel electrophoresis. Clin Chem 1999;45:143–6. [14] Wong P, Taillefer D, Lakins J, et al. Molecular characterization of human TRPM-2/ clusterin, a gene associated with sperm maturation, apoptosis and neurodegeneration. Eur J Biochem 1994;221:917–25. [15] Lee JH, Cheng R, Graff-Radford N, et al. Analyses of the National Institute on Aging Late-Onset Alzheimer's Disease Family Study: implication of additional loci. Arch Neurol 2008;65:1518–26. [16] de Silva HV, Harmony JA, Stuart WD, et al. Apolipoprotein J: structure and tissue distribution. Biochemistry 1990;29:5380–9. [17] Yan H, Yuan W, Velculescu VE, et al. Allelic variation in human gene expression. Science 2002;297:1143.