Neuroscience Letters 584 (2015) 382–389
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Short communication
SORL1 gene polymorphism association with late-onset Alzheimer’s disease Xialu Feng, Deren Hou ∗ , Yanyao Deng, Wei Li, Mi Tian, Zhuling Yu Department of Neurology, The Third Xiangya Hospital, Central South University, 138 Tongzipo Road, Changsha HN410013, China
h i g h l i g h t s • • • •
SORL1 gene, as a candidate gene in lipid metabolic pathways is involved in AD onset. We examined on the relationship between SORL1 gene and Alzheimer’s disease onset. SORL1 gene SNPs at rs689021 and rs3824966 loci show no relationship with LOAD. The rs1784933 locus is associated with LOAD, with the A allele being a risk factor.
a r t i c l e
i n f o
Article history: Received 1 July 2014 Received in revised form 6 October 2014 Accepted 30 October 2014 Available online 4 November 2014 Keywords: Alzheimer’s disease SORL1 gene Polymorphism
a b s t r a c t We examined the relationship between loci polymorphisms (rs689021, rs3824966, and rs1784933) of the sortilin-related receptor 1 gene (SORL1) and late-onset Alzheimer’s disease (LOAD) in the Chinese Han population of the Hunan Changsha region. A case-control association analysis was used. Clinical data and peripheral blood were collected from 201 Alzheimer’s disease patients and 257 healthy controls. PCR and MALDI-TOF mass spectrometry detection technologies were used to identify single nucleotide polymorphism (SNP) distribution at SORL1 gene loci. Genotype and allele frequency differences were analyzed and compared between groups. No significant differences were found in genotype frequency distributions of the rs689021 and rs3824966 loci. Similarly, allele frequency distributions of the C and T alleles of rs689021, and the C and G alleles of rs3824966 showed no significant differences. However, the genotype frequency distribution of the rs1784933 locus was significantly different, and the allele frequency distribution of the A and G alleles were also significantly different. Multifactor logistic regression analysis showed that after correcting for confounding factors such as gender, age, and cholesterol, LOAD risk in rs1784933 AA genotype carriers was 1.803 times that in AG + GG genotype carriers. SORL1 gene SNPs at rs689021 and rs3824966 loci show no relationship with LOAD onset in the Chinese Han population of the Hunan Changsha region. Conversely, a SORL1 gene SNP at the rs1784933 locus is associated with LOAD onset, with the A allele being a risk factor. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Alzheimer’s disease (AD) is a degenerative disease of the nervous system, characterized by progressive cognitive dysfunction and memory impairment. AD pathogenesis is complex, and in recent years researchers have shown that disorders of lipid metabolism may be involved. Since Strittmatter et al. identified the
∗ Corresponding author. Tel.: +86 0731 88618007; fax: +86 0731 88618339. E-mail addresses: [email protected] (X. Feng), [email protected] (D. Hou), [email protected] (Y. Deng), [email protected] (W. Li), [email protected] (M. Tian), [email protected] (Z. Yu). http://dx.doi.org/10.1016/j.neulet.2014.10.055 0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
4 allele of the apolipoprotein E (APOE) gene as the main risk factor for AD in 1993, much research has been performed on the relationship between candidate genes in lipid metabolism pathways and AD onset [1]. Accordingly, alterations in the lipid metabolismrelated gene, sortilin-related receptor 1 (SORL1), may be AD risk factors. Thus, we examined single nucleotide polymorphism (SNPs) in SORL1, specifically, rs689021, rs3824966, and rs1784933, in lateonset Alzheimer’s disease (LOAD) patients from the Chinese Han population of the Hunan Changsha region, to determine the correlation between genotype and allele frequency at these loci and LOAD onset. Our results may provide a theoretical basis for further prevention and control of AD.
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
2. Materials and methods 2.1. Study subjects The LOAD group comprised outpatients and inpatients with AD, identified from the Neurology Department of the Third Xiangya Hospital of Changsha City of Hunan Province from July 2009 to 2013. The control group comprised healthy people without dementia, tested in the Health Management Center of the Third Xiangya Hospital at the same time. All participants are from the Chinese Han population of the Hunan Changsha region. Clinical data and peripheral blood specimens were collected from all subjects. Overall, there were 201 cases in the LOAD group (90 men and 111 women), with an age range of 66–90 years (mean age 76.79 ± 5.65 years). Inclusion criteria for the LOAD group were: (1) patients confirm with AD diagnostic criteria in the revised fourth edition of the “Diagnostic and Statistical Manual for Mental Illness” (DSM IV-R), and of the US National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Diseases and Related Disorders Associations (NINCDS-ADRDA); (2) patients confirm with minimental state examination (MMSE) scores for illiterate patients (≤17 points), patients with primary education (≤20 points), and patients educated to high school and above (≤24 points). Additionally, Clinical Dementia Rating (CDR) scores ≥1 point, Hamilton Depression Scale (HAMD) scores ≤7 points, Hachinski Ischemic Score (HIS) 20 points, and revised Hasegawa Dementia Scale (HDS-R) scores 0.05).
383
at 120 V for 30 min with a 1Kbp DNA ladder used as a marker. DNA band integrity was observed using an electrophoresis gel imaging system. DNA bands clearly visible with no other nonspecific bands showed good DNA integrity for SNP genotyping. 2.2.3. Primer design and synthesis Primer design was performed using Assay Design 3.1 software (Sequenom, Inc.) and synthesis performed by Gene-Cloud Biotechnology Co. (Beijing, China). Specific PCR amplification primers are shown in Table 1. 2.2.4. PCR and SNP genotyping Genomic DNA was standardized to 10–20 ng/L, and centrifuged at 1000 × g for 3 min. PCRs were performed in 384 well plates with 5 L reaction volumes. PCR conditions were: 94 ◦ C for 15 min, then 45 cycles of 94 ◦ C for 20 s, 56 ◦ C for 30 s, and 72 ◦ C for 1 min, with a final extension at 72 ◦ C for 30 min and then reactions held at 4 ◦ C. SAP reactions were performed to remove excess dNTPs from PCR products. Reactions (7 L) included 5 L PCR product and 2 L SAP reaction liquid. SAP reaction conditions were: 37 ◦ C for 40 min, 85 ◦ C for 5 min, and then reactions held at 4 ◦ C. Samples for SNP detection were obtained after PCR product purification and extension reactions to extend single primer bases. After desalination purification, resin, and chip spotting of samples was performed, and spotted chips analyzed using the MassArray Compact System (Sequenom, Inc.) to obtain genotyping data. 2.2.5. Statistical methods SPSS 19.0 statistical software was used for analysis. Measurement data are shown as mean ± SD, and count data as number of cases (relative numbers). T and chi-square tests were used to compare general characteristics between the two groups. Chi-square tests were also used to compare genotype and allele frequencies. The Hardy–Weinberg equilibrium test was used to determine group representativeness. The odds ratio (OR) and 95% confidence interval (CI) were used to assess relative risk. A logistic regression model was used to examine correlation between SORL1 SNPs and LOAD susceptibility after correcting for confounding factors. Test levels with ˛ = 0.05 and P < 0.05 were considered statistically significant. 3. Results
2.2. Methods
3.1. Mass spectrum genotype peaks detected by MassArray
2.2.1. Collection of clinical data and blood After selection using inclusion and exclusion criteria for LOAD and control groups, patient clinical data was collected using a unified questionnaire. Peripheral venous blood was extracted from all subjects. Blood was separated into two tubes with 3 mL in each. One tube was stored at −20 ◦ C with an EDTA anticoagulant, and used for extracting genomic DNA. The other tube was used for determination of biochemical indexes such as fasting blood glucose (FBG), total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDLC).
In LOAD and control groups, MassArray analysis detected CC, CT, and TT genotypes for rs689021, CC, CG, and GG genotypes for rs3824966, and AA, AG, and GG genotypes for rs1784933. Representative spectrometric genotype-feature peak charts for all three SNPs are shown (Figs. 1–3).
2.2.2. Whole blood genomic DNA extraction and sample identification Blood preserved with EDTA anticoagulant was removed from −20 ◦ C and thawed at room temperature. DNA was extracted using a blood genomic DNA extraction kit (ComWin Biotech Co., Beijing, China), according to the manufacturer’s instructions. After DNA extraction from all samples, DNA concentration and purity were determined. DNA integrity was confirmed by 2% agarose gel electrophoresis using 2 L DNA + 1 L 6 × loading buffer, and gels run
3.2. Clinical data comparison between LOAD and control groups Clinical data between LOAD and control groups were compared. Gender, age, widowhood, body mass index (BMI), history of head trauma, history of high blood pressure, triglycerides, LDL-C, HDL-C, and FBG showed no significant difference between groups (P > 0.05). However, aspects of education level and total cholesterol level were significantly different between the two groups (P < 0.05; Table 2). Low education level and high total cholesterol were risk factors for AD onset. 3.3. Testing for Hardy–Weinberg equilibrium Hardy–Weinberg equilibrium tests were performed in LOAD and control groups, on the three loci containing rs689021,
384
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
Table 1 Primer sequences. SNP rs689021
Forward Reverse Forward Reverse Forward Reverse
rs3824966 rs1784933
Nucleotide sequences (5 → 3 )
Fragment size (bp)
ACGTTGGATGACCTTACAGATGATGCAGCC ACGTTGGATGGGCCATAGTTTCCTAGCATC ACGTTGGATGCCAAGCTAATTCTCAGAGCC ACGTTGGATGTTGACAGCACTCATCCGTTC ACGTTGGATGTTTGAAGCAGTTCCAGGGTC ACGTTGGATGGAATGGAAGAGGACATCAGC
100 103 102
SNP: single nucleotide polymorphism; bp: base pairs.
Table 2 Clinical data comparison between LOAD and control groups.
3.5. C/G polymorphism distribution at the rs3824966 locus in LOAD and control groups
Index
LOAD group(n = 201)
Control group(n = 257)
P value
Gender (male/female) Age (years) Education level (≤primary school/>primary school) Loss of spouse (Y/N) HBP (Y/N) Head trauma (Y/N) BMI (kg/m2 ) TG (mmol/L) TC (mmol/L) LDL-C (mmol/L) HDL-C (mmol/L) FBG (mmol/L)
90/111 76.79 ± 5.65 117/84
121/136 75.88 ± 6.50 120/137
0.623 0.113 0.014*
89/112 62/139 8/193 22.65 ± 1.94 1.79 ± 1.15 5.11 ± 0.98 2.59 ± 1.09 1.36 ± 0.42 5.34 ± 0.90
92/165 65/192 11/246 22.35 ± 1.48 1.98 ± 1.40 4.87 ± 0.69 2.46 ± 1.13 1.35 ± 0.54 5.19 ± 0.71
0.065 0.188 0.873 0.068 0.123 0.002* 0.227 0.722 0.065
HBP: high blood pressure; BMI: body mass index; ≤primary school: illiterate or educated to primary school; >primary school: educated beyond primary school; TG: triglycerides; TC: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; FBG: fasting blood glucose. * P < 0.05 compared with the control group.
rs3824966, and rs1784933 genotypes. All loci had non-significant P values (>0.05), in accordance with Hardy–Weinberg equilibrium, and indicating that the samples used in this study show group representativeness and are suitable for genetic analysis.
3.4. C/T polymorphism distribution at the rs689021 locus in LOAD and control groups Detection and analysis results by MassArray identified CC, CT, and TT genotypes in both LOAD and control groups. In the AD group, the TT genotype had the highest frequency, followed by the CT genotype, and then the CC genotype. In the control group, the CT genotype had the highest frequency, followed by the TT genotype, and then the CC genotype. Group comparisons found no significant difference in genotype frequency distribution at the rs689021 locus. In addition, C and T allele frequency distribution showed no significant difference (Table 3).
Detection and analysis results by MassArray identified CC, CG, and GG genotypes in both LOAD and control groups. In both AD and control groups, the CG genotype had the highest frequency, followed by the GG genotype, and then the CC genotype. Group comparisons showed no significant difference in genotype frequency distribution at the rs3824966 locus. Additionally, C and G allele frequency distribution showed no significant difference (Table 3). 3.6. C/G polymorphism distribution at the rs1784933 locus in LOAD and control groups Detection and analysis results by MassArray identified GG, AG, and AA genotypes in both LOAD and control groups. In both AD and control groups, the AA genotype had the highest frequency, followed by the AG genotype, and then the GG genotype. Group comparisons found a significant difference in genotype frequency distribution at the rs1784933 locus. Moreover, C and G allele frequency distribution was also significantly different (Table 3). 3.7. Multifactor logistic regression analysis of SORL1 gene polymorphism distribution and LOAD correlation Multifactor logistic regression analysis showed that after correcting for influences such as gender, age, education level, and cholesterol, AD risk in rs1784933 AA genotype carriers was 1.803 times those with AG + GG genotypes (OR:1.803, 95%CI:1.225∼2.654,P = 0.003). 4. Discussion Amyloid- protein (A) deposition is a key pathology of AD. Present studies suggest that disorders of cholesterol metabolism are associated with A production and AD onset [2–4]. Raised cholesterol levels stimulate - and ␥-secretases, which catalyze amyloid precursor protein (APP) cleavage, generating many pathological A segments, of which, the A oligomer has the greatest impact. APP is a transmembrane glycoprotein with a
Table 3 Distribution of genotype and allele frequencies at the SORL1 SNPs. SNP
Group
rs689021 LOAD Control rs3824966 LOAD Control rs1784933 LOAD Control
Genotype (%cases) C/C (freq) 32 (15.9) 40 (15.6) C/C (freq) 22 (10.9) 27 (10.5) A/A (freq) 124 (61.7) 125 (48.6)
P C/T (freq) 84 (41.8) 121 (47.1) C/G (freq) 96 (47.8) 117 (45.5) A/G (freq) 66 (32.8) 107 (41.6)
T/T (freq) 85 (42.3) 96 (37.4) G/G (freq) 83 (41.3) 113 (44.0) G/G (freq) 11 (5.5) 25 (9.7)
0.494
0.848
0.015
Genotype and allele frequencies were compared between LOAD and control groups using 2 tests. P < 0.05, significant.
Allele (%cases) C (freq) 148 (36.8) 201 (39.1) C (freq) 140 (34.8) 171 (33.3) A (freq) 314 (78.1) 357 (69.5)
P T (freq) 254 (63.2) 313 (60.9) G (freq) 262 (65.2) 343 (66.7) G (freq) 88 (21.9) 157 (30.5)
0.479
0.621
0.003
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
385
689021
T
UEP.3087258
C
UEP.1143627
A
G
UEP.689021
20
Intensity
15 10 5 0 5200
5250
5300
5350
5400 Mass
5450
5500
5550
5600
CC genotypes
689021
T
UEP.3087258
C
UEP.1143627
A
G
UEP.689021
20
Intensity
15
10
5
0 5200
5250
5300
5350
5400 Mass
5450
5500
5550
5600
CT genotypes
689021
T
UEP.3087258
C
UEP.1143627
A
G
UEP.689021
18 16 14 12 Intensity
10 8 6 4 2 0 5200
5250
5300
5350
5400 Mass
5450
5500
5550
5600
TT genotypes Fig. 1. Mass spectrum of rs689021. MassArray analysis detected CC, CT, and TT genotypes in LOAD and control groups.
receptor-like structure, found on the surface of nerve cells inserted into the lipid bilayer, which consists of cholesterol, phospholipids, and local specialized structures known as the “raft area”. Moreover, - and ␥-secretases are present in lipid rafts, and normal protein metabolism may be disrupted, producing A oligomers by changes
in cholesterol content in rafts. The toxic effect of A oligomers in neurons is the basis for pathological changes in AD [5]. A can also affect the balance of cholesterol metabolism [6], and A-ester particles formed by A and cholesterol interfere with cholesterol metabolism in normal cells. It is this interaction of cholesterol
386
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
3824966 UEP.2383207
G
C
UEP.3824966
20
Intensity
15
10
5
0 7950
8000
8050
8100
8150
Mass
8200
8250
8300
CC genotypes 3824966 UEP.2383207
G
C
UEP.3824966
20
Intensity
15 10 5 0 7950
8000
8050
8100
8150
Mass
8200
8250
8300
CG genotypes
3824966 UEP.2383207
G
C
UEP.3824966
20
Intensity
15 10 5 0 7950
8000
8050
8100
Mass
8150
8200
8250
8300
GG genotypes Fig. 2. Mass spectrum of rs3824966. MassArray analysis detected CC, CG, and GG genotypes in LOAD and control groups.
with A that forms a vicious cycle, leading to excessive A deposition and eventually causing AD onset. Compared with the control group, total cholesterol in the LOAD group increased significantly, suggesting that high total cholesterol may be a risk factor for AD onset in this study, consistent with previous research. Because the APOE 4 allele is known to be an independent AD risk factor, many
studies have been performed on the relationship between candidate genes in lipid metabolic pathways and AD onset [1], to identify susceptibility genes in AD etiology. A study using “cholesterol” as a retrieval keyword in an AlzGene meta-analysis found five genes positively associated with AD; with SORL1 identified as one of them [7].
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
387
1784933 UEP.7412
A
G
UEP.4291
UEP.2304456
C
G
UEP.1784933
20
Intensity
15
10
5
0 6650
6700
6750
6800
Mass
6850
6900
6950
7000
AA genotypes
1784933 UEP.7412
A
G
UEP.4291
UEP.2304456
C
G
UEP.1784933
30 25
Intensity
20 15 10 5 0 6650
6700
6750
6800
Mass
6850
6900
6950
7000
AG genotypes
1784933 UEP.7412
A
G
UEP.4291
UEP.2304456
C
G
UEP.1784933
20
Intensity
15 10 5 0 6650
6700
6750
6800
Mass
6850
6900
6950
7000
GG genotypes Fig. 3. Mass spectrum of rs1784933. MassArray analysis detected AA, AG, and GG genotypes in LOAD and control groups.
388
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
Fig. 4. The location of the SNPs within the SORL1 gene. The SORL1 gene is located on chromosome 11.53160bp,130223bp,171456bp show accurately the location of rs689021,rs3824966,rs1784933 within the SORL1 gene.
In 2007, Rogaeva et al. [8] suggested that SORL1 may be the second risk factor for AD onset, reporting SNP results of the SORL1 gene sequence in six different species, and association between a SNP in SORL1 and LOAD. Following this, successive studies have indicated the SORL1 may be genetically associated with increased risk for amnestic mild cognitive impairment and Alzheimer’s disease in the Han Chinese population [9,10,11], although some studies have shown no obvious correlation [12,13]. In humans, the SORL1 gene is located on the long arm of chromosome 11 (q23.2 24.2) and is 177.49 kb in length. The receptor encoded by SORL1 is a type of membrane receptor that combines with lipoproteins, and is expressed in the central and peripheral nervous systems. It belongs to the family of LDL receptors, and is found in the plasma membrane, endosomes, and Golgi apparatus. As a selective APP protein receptor, SORL1 specifically binds to APP, assisting in cell transfer to the Golgi apparatus, and significantly decreasing the likelihood of APP being hydrolyzed to A by -hydrolase. SORL1 gene mutations increase the risk of nerve degeneration through a mechanism that may be related to gene activity inhibition [8]. SORL1 gene variants cause reductions in SORL1 receptor expression, leading to high A production levels in the brain, and subsequent AD onset. In this study, we selected SNP9 (rs689021) at the 5 end of the SORL1 gene, and SNP20 (rs3824966) and SNP26 (rs1784933) at the 3 end (Fig. 4). We found that in both LOAD and control groups, genotype and allele frequencies of rs689021 and rs3824966 were not significantly different. However, there was a significant difference in genotype and allele frequencies of the rs1784933 locus, suggesting an association with LOAD onset. Bettens et al. [14] studied 550 LOAD patients and 637 healthy controls in Belgium, and found that three genotypes at the rs689021 locus of the SORL1 gene showed significant differences between the two groups (P < 0.05). Ning et al. [15] evaluated the association of several SNPs, including rs689021, in the SORL1 gene with risk for LOAD in the Chinese population and observed that rs689021 in SORL1 showed no significantly different allele frequencies between patients and controls. Lee et al. [16] found a correlation between the rs3824966 G allele and AD onset among non-Hispanic whites and Nordic populations (P = 0.025). However, in this study, genotypes and alleles at the above two loci did not show a relationship with LOAD, and do not corroborate our results. This difference may be due to the influence of genetic heterogeneity in different races, populations, and regions, as well as sample size. As AD is a complex polygenic hereditary disease, it may also be caused by the synergistic effect of other genetic and environmental factors.
The wild-type genotype at the rs1784933 locus of the SORL1 gene is GG, with GA and AA the two possible variants. Our results suggest that the A/G SNP at the rs1784933 locus is associated with LOAD in the Chinese Han population of the Hunan Changsha region, with the risk allele being the A allele (OR, 1.569; 95% CI, 1.16–2.122). Multifactor logistic regression analysis correcting for confounding factors such as gender, age, and cholesterol, found that the LOAD risk in AA genotype carriers was 1.803 times those carrying AG + GG genotypes (OR, 1.803; 95% CI, 1.225–2.654). Lee et al. [16] examined SORL1 genetic variation in 296 AD patients and 428 healthy elderly controls, and found that rs1784933 was significantly associated with AD in the African American population (P = 0.019). AD risk in G allele carriers was less than in A allele carriers, in accordance with our findings in this study. In conclusion, the relationship between rs689021, rs3824966, and rs1784933 loci of the SORL1 gene and AD onset may be influenced by many factors. Our study shows that polymorphism of the rs1784933 locus is associated with LOAD onset in AD patients from the Hunan Changsha region. There are SNPs at multiple SORL1 gene loci, therefore to further understand the correlation between SORL1 and AD, we need to identify more loci for studies of larger populations. This should provide a basis for clarifying AD pathogenesis and identifying new therapeutic targets. Acknowledgments This work was supported by the Community Development Supporting Plan Project of Hunan Provincial Science and Technology Department, China (Nos. 009SK3175 and 2012SK3218). References [1] T. Kanekiyo, H. Xu, G. Bu, ApoE and A in Alzheimer’ disease: accidental encounters or partners? Neuron 81 (2014) 740–754. [2] W.J. Davis, The ATP-binding cassette transporter-2 (ABCA2) regulates esterification of plasma membrane cholesterol by modulation of sphingolipid metabolism, Biochim. Biophys. Acta 1841 (2014) 168–179. [3] P. Gamba, G. Testa, B. Sottero, S. Gargiulo, G. Poli, G. Leonarduzzi, The link between altered cholesterol metabolism and Alzheimer’s disease, Ann. N. Y. Acad. Sci. 1259 (2012) 54–56. [4] M. Maulik, D. Westaway, J.H. Jhamandas, S. Kar, Role of cholesterol in APP metabolism and its significance in Alzheimer’s disease pathogenesis, Mol. Neurobiol. 47 (2013) 37–63. [5] G.M. Shankar, S. Li, T.H. Mehta, A. Garcia-Munoz, N.E. Shepardson, I. Smith, F.M. Brett, M.A. Farrell, M.J. Rowan, C.A. Lemere, C.M. Regan, D.M. Walsh, B.L. Sabatini, D.J. Selkoe, Amyloid-beta protein dimers isolated directly from Alzheimer’s brains impair synaptic plasticity and memory, Nat. Med. 14 (2008) 837–842.
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389 [6] A. Schneider, W. Schulz-Schaeffer, T. Hartmann, J.B. Schulz, M. Simons, Cholesterol depletion reduces aggregation of amyloid-beta peptide in hippocampal neurons, Neurobiol. Dis. 23 (2006) 573. [7] L. Bertram, M.B. McQueen, K. Mullin, D. Blacker, R.E. Tanzi, Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database, Nat. Genet. 39 (2007) 17–23. [8] E. Rogaeva, Y. Meng, J.H. Lee, Y. Gu, T. Kawarai, F. Zou, T. Katayama, C.T. Baldwin, R. Cheng, H. Hasegawa, F. Chen, N. Shibata, K.L. Lunetta, R. Pardossi-Piquard, C. Bohm, Y. Wakutani, L.A. Cupples, K.T. Cuenco, R.C. Green, L. Pinessi, I. Rainero, S. Sorbi, A. Bruni, R. Duara, R.P. Friedland, R. Inzelberg, W. Hampe, H. Bujo, Y.Q. Song, O.M. Andersen, T.E. Willnow, N. Graff-Radford, R.C. Petersen, D. Dickson, S.D. Der, P.E. Fraser, G. Schmitt-Ulms, S. Younkin, R. Mayeux, L.A. Farrer, S. George, P. Hyslop, The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease, Nat. Genet. 39 (2007) 168–177. [9] X. Xue, M. Zhang, Y. Lin, E. Xu, J. Jia, Association between the SORL1 rs2070045 polymorphism and late-onset Alzheimer’s disease: interaction with the ApoE genotype in the Chinese Han population, Neurosci. Lett. 559 (2014) 94–98. [10] C. Jin, L. Zhang, Y. Xian, X. Liu, Y. Wu, F. Zhang, J. Zhu, G. Zhang, C. Chen, R. Gong, L. Zhang, J. Yuan, F. Zhang, L. Tian, G. Wang, Z. Cheng, The SORL1 polymorphism rs985421 may confer the risk for amnestic mild cognitive impairment and Alzheimer’s disease in the Han Chinese population, Neurosci. Lett. 563 (2014) 80–84.
389
[11] X. Gao, M. Liu, L. Sun, B. Qin, H. Yu, Z. Yang, R. Qi, F. Gao, SORL1 genetic variants modulate risk of amnestic mild cognitive impairment in northern Han Chinese, Int. J. Neurosci. 124 (2014) 296–301. [12] F. Liu, M.A. Ikram, A.C. Janssens, M. Schuur, I.D. Koning, A. Isaacs, M. Struchalin, A.G. Uitterlinden, J.T. den Dunnen, K. Sleegers, K. Bettens, C.V. Broeckhoven, J.V. Swieten, A. Hofman, B.A. Oostra, Y.S. Aulchenko, M.M. Breteler, C.M. Duijn, A study of the SORL 1 genein Alzheimer’s disease and cognitive function, J. Alzheimer Dis. 18 (2009) 51–64. [13] R.L. Minster, S.T. DeKosk, M.I. Kamboh, No association of SORL 1 SNPs with Alzheimer’s disease, Neurosci. Lett. 440 (2008) 190–192. [14] K. Bettens, N. Brouwers, S. Engelborghs, P.P.D. Deyn, C.V. Broeckhoven, K. Sleegers, SORL1 is genetically associated with increased risk for late-onset Alzheimer disease in the Belgian population, Hum. Mutat. 29 (2008) 769–770. [15] M. Ning, Y. Yang, Z. Zhang, Z. Chen, T. Zhao, D. Zhang, D. Zhou, J. Xu, Z. Liu, Y. Wang, Y. Liu, X. Zhao, W. Li, S. Li, L. He, Amyloid--related genes SORL1 and ACE are genetically associated with risk for late-onset Alzheimer disease in the Chinese population, Alzheimer Dis. Assoc. Disord. 24 (2010) 390–396. [16] J.H. Lee, R. Cheng, N. Schupf, J. Manly, R. Lantigua, Y. Stern, E. Rogaeva, Y. Wakutani, L. Farrer, P.S. George, R. Mayeux, The association between genetic variants in SORL1 and Alzheimer’s disease in an urban, multiethnic, community-based cohort, Arch. Neurol. 64 (2007) 501–506.
Contents lists available at ScienceDirect
Neuroscience Letters journal homepage: www.elsevier.com/locate/neulet
Short communication
SORL1 gene polymorphism association with late-onset Alzheimer’s disease Xialu Feng, Deren Hou ∗ , Yanyao Deng, Wei Li, Mi Tian, Zhuling Yu Department of Neurology, The Third Xiangya Hospital, Central South University, 138 Tongzipo Road, Changsha HN410013, China
h i g h l i g h t s • • • •
SORL1 gene, as a candidate gene in lipid metabolic pathways is involved in AD onset. We examined on the relationship between SORL1 gene and Alzheimer’s disease onset. SORL1 gene SNPs at rs689021 and rs3824966 loci show no relationship with LOAD. The rs1784933 locus is associated with LOAD, with the A allele being a risk factor.
a r t i c l e
i n f o
Article history: Received 1 July 2014 Received in revised form 6 October 2014 Accepted 30 October 2014 Available online 4 November 2014 Keywords: Alzheimer’s disease SORL1 gene Polymorphism
a b s t r a c t We examined the relationship between loci polymorphisms (rs689021, rs3824966, and rs1784933) of the sortilin-related receptor 1 gene (SORL1) and late-onset Alzheimer’s disease (LOAD) in the Chinese Han population of the Hunan Changsha region. A case-control association analysis was used. Clinical data and peripheral blood were collected from 201 Alzheimer’s disease patients and 257 healthy controls. PCR and MALDI-TOF mass spectrometry detection technologies were used to identify single nucleotide polymorphism (SNP) distribution at SORL1 gene loci. Genotype and allele frequency differences were analyzed and compared between groups. No significant differences were found in genotype frequency distributions of the rs689021 and rs3824966 loci. Similarly, allele frequency distributions of the C and T alleles of rs689021, and the C and G alleles of rs3824966 showed no significant differences. However, the genotype frequency distribution of the rs1784933 locus was significantly different, and the allele frequency distribution of the A and G alleles were also significantly different. Multifactor logistic regression analysis showed that after correcting for confounding factors such as gender, age, and cholesterol, LOAD risk in rs1784933 AA genotype carriers was 1.803 times that in AG + GG genotype carriers. SORL1 gene SNPs at rs689021 and rs3824966 loci show no relationship with LOAD onset in the Chinese Han population of the Hunan Changsha region. Conversely, a SORL1 gene SNP at the rs1784933 locus is associated with LOAD onset, with the A allele being a risk factor. © 2014 Elsevier Ireland Ltd. All rights reserved.
1. Introduction Alzheimer’s disease (AD) is a degenerative disease of the nervous system, characterized by progressive cognitive dysfunction and memory impairment. AD pathogenesis is complex, and in recent years researchers have shown that disorders of lipid metabolism may be involved. Since Strittmatter et al. identified the
∗ Corresponding author. Tel.: +86 0731 88618007; fax: +86 0731 88618339. E-mail addresses: [email protected] (X. Feng), [email protected] (D. Hou), [email protected] (Y. Deng), [email protected] (W. Li), [email protected] (M. Tian), [email protected] (Z. Yu). http://dx.doi.org/10.1016/j.neulet.2014.10.055 0304-3940/© 2014 Elsevier Ireland Ltd. All rights reserved.
4 allele of the apolipoprotein E (APOE) gene as the main risk factor for AD in 1993, much research has been performed on the relationship between candidate genes in lipid metabolism pathways and AD onset [1]. Accordingly, alterations in the lipid metabolismrelated gene, sortilin-related receptor 1 (SORL1), may be AD risk factors. Thus, we examined single nucleotide polymorphism (SNPs) in SORL1, specifically, rs689021, rs3824966, and rs1784933, in lateonset Alzheimer’s disease (LOAD) patients from the Chinese Han population of the Hunan Changsha region, to determine the correlation between genotype and allele frequency at these loci and LOAD onset. Our results may provide a theoretical basis for further prevention and control of AD.
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
2. Materials and methods 2.1. Study subjects The LOAD group comprised outpatients and inpatients with AD, identified from the Neurology Department of the Third Xiangya Hospital of Changsha City of Hunan Province from July 2009 to 2013. The control group comprised healthy people without dementia, tested in the Health Management Center of the Third Xiangya Hospital at the same time. All participants are from the Chinese Han population of the Hunan Changsha region. Clinical data and peripheral blood specimens were collected from all subjects. Overall, there were 201 cases in the LOAD group (90 men and 111 women), with an age range of 66–90 years (mean age 76.79 ± 5.65 years). Inclusion criteria for the LOAD group were: (1) patients confirm with AD diagnostic criteria in the revised fourth edition of the “Diagnostic and Statistical Manual for Mental Illness” (DSM IV-R), and of the US National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Diseases and Related Disorders Associations (NINCDS-ADRDA); (2) patients confirm with minimental state examination (MMSE) scores for illiterate patients (≤17 points), patients with primary education (≤20 points), and patients educated to high school and above (≤24 points). Additionally, Clinical Dementia Rating (CDR) scores ≥1 point, Hamilton Depression Scale (HAMD) scores ≤7 points, Hachinski Ischemic Score (HIS) 20 points, and revised Hasegawa Dementia Scale (HDS-R) scores 0.05).
383
at 120 V for 30 min with a 1Kbp DNA ladder used as a marker. DNA band integrity was observed using an electrophoresis gel imaging system. DNA bands clearly visible with no other nonspecific bands showed good DNA integrity for SNP genotyping. 2.2.3. Primer design and synthesis Primer design was performed using Assay Design 3.1 software (Sequenom, Inc.) and synthesis performed by Gene-Cloud Biotechnology Co. (Beijing, China). Specific PCR amplification primers are shown in Table 1. 2.2.4. PCR and SNP genotyping Genomic DNA was standardized to 10–20 ng/L, and centrifuged at 1000 × g for 3 min. PCRs were performed in 384 well plates with 5 L reaction volumes. PCR conditions were: 94 ◦ C for 15 min, then 45 cycles of 94 ◦ C for 20 s, 56 ◦ C for 30 s, and 72 ◦ C for 1 min, with a final extension at 72 ◦ C for 30 min and then reactions held at 4 ◦ C. SAP reactions were performed to remove excess dNTPs from PCR products. Reactions (7 L) included 5 L PCR product and 2 L SAP reaction liquid. SAP reaction conditions were: 37 ◦ C for 40 min, 85 ◦ C for 5 min, and then reactions held at 4 ◦ C. Samples for SNP detection were obtained after PCR product purification and extension reactions to extend single primer bases. After desalination purification, resin, and chip spotting of samples was performed, and spotted chips analyzed using the MassArray Compact System (Sequenom, Inc.) to obtain genotyping data. 2.2.5. Statistical methods SPSS 19.0 statistical software was used for analysis. Measurement data are shown as mean ± SD, and count data as number of cases (relative numbers). T and chi-square tests were used to compare general characteristics between the two groups. Chi-square tests were also used to compare genotype and allele frequencies. The Hardy–Weinberg equilibrium test was used to determine group representativeness. The odds ratio (OR) and 95% confidence interval (CI) were used to assess relative risk. A logistic regression model was used to examine correlation between SORL1 SNPs and LOAD susceptibility after correcting for confounding factors. Test levels with ˛ = 0.05 and P < 0.05 were considered statistically significant. 3. Results
2.2. Methods
3.1. Mass spectrum genotype peaks detected by MassArray
2.2.1. Collection of clinical data and blood After selection using inclusion and exclusion criteria for LOAD and control groups, patient clinical data was collected using a unified questionnaire. Peripheral venous blood was extracted from all subjects. Blood was separated into two tubes with 3 mL in each. One tube was stored at −20 ◦ C with an EDTA anticoagulant, and used for extracting genomic DNA. The other tube was used for determination of biochemical indexes such as fasting blood glucose (FBG), total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDLC).
In LOAD and control groups, MassArray analysis detected CC, CT, and TT genotypes for rs689021, CC, CG, and GG genotypes for rs3824966, and AA, AG, and GG genotypes for rs1784933. Representative spectrometric genotype-feature peak charts for all three SNPs are shown (Figs. 1–3).
2.2.2. Whole blood genomic DNA extraction and sample identification Blood preserved with EDTA anticoagulant was removed from −20 ◦ C and thawed at room temperature. DNA was extracted using a blood genomic DNA extraction kit (ComWin Biotech Co., Beijing, China), according to the manufacturer’s instructions. After DNA extraction from all samples, DNA concentration and purity were determined. DNA integrity was confirmed by 2% agarose gel electrophoresis using 2 L DNA + 1 L 6 × loading buffer, and gels run
3.2. Clinical data comparison between LOAD and control groups Clinical data between LOAD and control groups were compared. Gender, age, widowhood, body mass index (BMI), history of head trauma, history of high blood pressure, triglycerides, LDL-C, HDL-C, and FBG showed no significant difference between groups (P > 0.05). However, aspects of education level and total cholesterol level were significantly different between the two groups (P < 0.05; Table 2). Low education level and high total cholesterol were risk factors for AD onset. 3.3. Testing for Hardy–Weinberg equilibrium Hardy–Weinberg equilibrium tests were performed in LOAD and control groups, on the three loci containing rs689021,
384
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
Table 1 Primer sequences. SNP rs689021
Forward Reverse Forward Reverse Forward Reverse
rs3824966 rs1784933
Nucleotide sequences (5 → 3 )
Fragment size (bp)
ACGTTGGATGACCTTACAGATGATGCAGCC ACGTTGGATGGGCCATAGTTTCCTAGCATC ACGTTGGATGCCAAGCTAATTCTCAGAGCC ACGTTGGATGTTGACAGCACTCATCCGTTC ACGTTGGATGTTTGAAGCAGTTCCAGGGTC ACGTTGGATGGAATGGAAGAGGACATCAGC
100 103 102
SNP: single nucleotide polymorphism; bp: base pairs.
Table 2 Clinical data comparison between LOAD and control groups.
3.5. C/G polymorphism distribution at the rs3824966 locus in LOAD and control groups
Index
LOAD group(n = 201)
Control group(n = 257)
P value
Gender (male/female) Age (years) Education level (≤primary school/>primary school) Loss of spouse (Y/N) HBP (Y/N) Head trauma (Y/N) BMI (kg/m2 ) TG (mmol/L) TC (mmol/L) LDL-C (mmol/L) HDL-C (mmol/L) FBG (mmol/L)
90/111 76.79 ± 5.65 117/84
121/136 75.88 ± 6.50 120/137
0.623 0.113 0.014*
89/112 62/139 8/193 22.65 ± 1.94 1.79 ± 1.15 5.11 ± 0.98 2.59 ± 1.09 1.36 ± 0.42 5.34 ± 0.90
92/165 65/192 11/246 22.35 ± 1.48 1.98 ± 1.40 4.87 ± 0.69 2.46 ± 1.13 1.35 ± 0.54 5.19 ± 0.71
0.065 0.188 0.873 0.068 0.123 0.002* 0.227 0.722 0.065
HBP: high blood pressure; BMI: body mass index; ≤primary school: illiterate or educated to primary school; >primary school: educated beyond primary school; TG: triglycerides; TC: total cholesterol; LDL-C: low-density lipoprotein cholesterol; HDL-C: high-density lipoprotein cholesterol; FBG: fasting blood glucose. * P < 0.05 compared with the control group.
rs3824966, and rs1784933 genotypes. All loci had non-significant P values (>0.05), in accordance with Hardy–Weinberg equilibrium, and indicating that the samples used in this study show group representativeness and are suitable for genetic analysis.
3.4. C/T polymorphism distribution at the rs689021 locus in LOAD and control groups Detection and analysis results by MassArray identified CC, CT, and TT genotypes in both LOAD and control groups. In the AD group, the TT genotype had the highest frequency, followed by the CT genotype, and then the CC genotype. In the control group, the CT genotype had the highest frequency, followed by the TT genotype, and then the CC genotype. Group comparisons found no significant difference in genotype frequency distribution at the rs689021 locus. In addition, C and T allele frequency distribution showed no significant difference (Table 3).
Detection and analysis results by MassArray identified CC, CG, and GG genotypes in both LOAD and control groups. In both AD and control groups, the CG genotype had the highest frequency, followed by the GG genotype, and then the CC genotype. Group comparisons showed no significant difference in genotype frequency distribution at the rs3824966 locus. Additionally, C and G allele frequency distribution showed no significant difference (Table 3). 3.6. C/G polymorphism distribution at the rs1784933 locus in LOAD and control groups Detection and analysis results by MassArray identified GG, AG, and AA genotypes in both LOAD and control groups. In both AD and control groups, the AA genotype had the highest frequency, followed by the AG genotype, and then the GG genotype. Group comparisons found a significant difference in genotype frequency distribution at the rs1784933 locus. Moreover, C and G allele frequency distribution was also significantly different (Table 3). 3.7. Multifactor logistic regression analysis of SORL1 gene polymorphism distribution and LOAD correlation Multifactor logistic regression analysis showed that after correcting for influences such as gender, age, education level, and cholesterol, AD risk in rs1784933 AA genotype carriers was 1.803 times those with AG + GG genotypes (OR:1.803, 95%CI:1.225∼2.654,P = 0.003). 4. Discussion Amyloid- protein (A) deposition is a key pathology of AD. Present studies suggest that disorders of cholesterol metabolism are associated with A production and AD onset [2–4]. Raised cholesterol levels stimulate - and ␥-secretases, which catalyze amyloid precursor protein (APP) cleavage, generating many pathological A segments, of which, the A oligomer has the greatest impact. APP is a transmembrane glycoprotein with a
Table 3 Distribution of genotype and allele frequencies at the SORL1 SNPs. SNP
Group
rs689021 LOAD Control rs3824966 LOAD Control rs1784933 LOAD Control
Genotype (%cases) C/C (freq) 32 (15.9) 40 (15.6) C/C (freq) 22 (10.9) 27 (10.5) A/A (freq) 124 (61.7) 125 (48.6)
P C/T (freq) 84 (41.8) 121 (47.1) C/G (freq) 96 (47.8) 117 (45.5) A/G (freq) 66 (32.8) 107 (41.6)
T/T (freq) 85 (42.3) 96 (37.4) G/G (freq) 83 (41.3) 113 (44.0) G/G (freq) 11 (5.5) 25 (9.7)
0.494
0.848
0.015
Genotype and allele frequencies were compared between LOAD and control groups using 2 tests. P < 0.05, significant.
Allele (%cases) C (freq) 148 (36.8) 201 (39.1) C (freq) 140 (34.8) 171 (33.3) A (freq) 314 (78.1) 357 (69.5)
P T (freq) 254 (63.2) 313 (60.9) G (freq) 262 (65.2) 343 (66.7) G (freq) 88 (21.9) 157 (30.5)
0.479
0.621
0.003
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
385
689021
T
UEP.3087258
C
UEP.1143627
A
G
UEP.689021
20
Intensity
15 10 5 0 5200
5250
5300
5350
5400 Mass
5450
5500
5550
5600
CC genotypes
689021
T
UEP.3087258
C
UEP.1143627
A
G
UEP.689021
20
Intensity
15
10
5
0 5200
5250
5300
5350
5400 Mass
5450
5500
5550
5600
CT genotypes
689021
T
UEP.3087258
C
UEP.1143627
A
G
UEP.689021
18 16 14 12 Intensity
10 8 6 4 2 0 5200
5250
5300
5350
5400 Mass
5450
5500
5550
5600
TT genotypes Fig. 1. Mass spectrum of rs689021. MassArray analysis detected CC, CT, and TT genotypes in LOAD and control groups.
receptor-like structure, found on the surface of nerve cells inserted into the lipid bilayer, which consists of cholesterol, phospholipids, and local specialized structures known as the “raft area”. Moreover, - and ␥-secretases are present in lipid rafts, and normal protein metabolism may be disrupted, producing A oligomers by changes
in cholesterol content in rafts. The toxic effect of A oligomers in neurons is the basis for pathological changes in AD [5]. A can also affect the balance of cholesterol metabolism [6], and A-ester particles formed by A and cholesterol interfere with cholesterol metabolism in normal cells. It is this interaction of cholesterol
386
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
3824966 UEP.2383207
G
C
UEP.3824966
20
Intensity
15
10
5
0 7950
8000
8050
8100
8150
Mass
8200
8250
8300
CC genotypes 3824966 UEP.2383207
G
C
UEP.3824966
20
Intensity
15 10 5 0 7950
8000
8050
8100
8150
Mass
8200
8250
8300
CG genotypes
3824966 UEP.2383207
G
C
UEP.3824966
20
Intensity
15 10 5 0 7950
8000
8050
8100
Mass
8150
8200
8250
8300
GG genotypes Fig. 2. Mass spectrum of rs3824966. MassArray analysis detected CC, CG, and GG genotypes in LOAD and control groups.
with A that forms a vicious cycle, leading to excessive A deposition and eventually causing AD onset. Compared with the control group, total cholesterol in the LOAD group increased significantly, suggesting that high total cholesterol may be a risk factor for AD onset in this study, consistent with previous research. Because the APOE 4 allele is known to be an independent AD risk factor, many
studies have been performed on the relationship between candidate genes in lipid metabolic pathways and AD onset [1], to identify susceptibility genes in AD etiology. A study using “cholesterol” as a retrieval keyword in an AlzGene meta-analysis found five genes positively associated with AD; with SORL1 identified as one of them [7].
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
387
1784933 UEP.7412
A
G
UEP.4291
UEP.2304456
C
G
UEP.1784933
20
Intensity
15
10
5
0 6650
6700
6750
6800
Mass
6850
6900
6950
7000
AA genotypes
1784933 UEP.7412
A
G
UEP.4291
UEP.2304456
C
G
UEP.1784933
30 25
Intensity
20 15 10 5 0 6650
6700
6750
6800
Mass
6850
6900
6950
7000
AG genotypes
1784933 UEP.7412
A
G
UEP.4291
UEP.2304456
C
G
UEP.1784933
20
Intensity
15 10 5 0 6650
6700
6750
6800
Mass
6850
6900
6950
7000
GG genotypes Fig. 3. Mass spectrum of rs1784933. MassArray analysis detected AA, AG, and GG genotypes in LOAD and control groups.
388
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389
Fig. 4. The location of the SNPs within the SORL1 gene. The SORL1 gene is located on chromosome 11.53160bp,130223bp,171456bp show accurately the location of rs689021,rs3824966,rs1784933 within the SORL1 gene.
In 2007, Rogaeva et al. [8] suggested that SORL1 may be the second risk factor for AD onset, reporting SNP results of the SORL1 gene sequence in six different species, and association between a SNP in SORL1 and LOAD. Following this, successive studies have indicated the SORL1 may be genetically associated with increased risk for amnestic mild cognitive impairment and Alzheimer’s disease in the Han Chinese population [9,10,11], although some studies have shown no obvious correlation [12,13]. In humans, the SORL1 gene is located on the long arm of chromosome 11 (q23.2 24.2) and is 177.49 kb in length. The receptor encoded by SORL1 is a type of membrane receptor that combines with lipoproteins, and is expressed in the central and peripheral nervous systems. It belongs to the family of LDL receptors, and is found in the plasma membrane, endosomes, and Golgi apparatus. As a selective APP protein receptor, SORL1 specifically binds to APP, assisting in cell transfer to the Golgi apparatus, and significantly decreasing the likelihood of APP being hydrolyzed to A by -hydrolase. SORL1 gene mutations increase the risk of nerve degeneration through a mechanism that may be related to gene activity inhibition [8]. SORL1 gene variants cause reductions in SORL1 receptor expression, leading to high A production levels in the brain, and subsequent AD onset. In this study, we selected SNP9 (rs689021) at the 5 end of the SORL1 gene, and SNP20 (rs3824966) and SNP26 (rs1784933) at the 3 end (Fig. 4). We found that in both LOAD and control groups, genotype and allele frequencies of rs689021 and rs3824966 were not significantly different. However, there was a significant difference in genotype and allele frequencies of the rs1784933 locus, suggesting an association with LOAD onset. Bettens et al. [14] studied 550 LOAD patients and 637 healthy controls in Belgium, and found that three genotypes at the rs689021 locus of the SORL1 gene showed significant differences between the two groups (P < 0.05). Ning et al. [15] evaluated the association of several SNPs, including rs689021, in the SORL1 gene with risk for LOAD in the Chinese population and observed that rs689021 in SORL1 showed no significantly different allele frequencies between patients and controls. Lee et al. [16] found a correlation between the rs3824966 G allele and AD onset among non-Hispanic whites and Nordic populations (P = 0.025). However, in this study, genotypes and alleles at the above two loci did not show a relationship with LOAD, and do not corroborate our results. This difference may be due to the influence of genetic heterogeneity in different races, populations, and regions, as well as sample size. As AD is a complex polygenic hereditary disease, it may also be caused by the synergistic effect of other genetic and environmental factors.
The wild-type genotype at the rs1784933 locus of the SORL1 gene is GG, with GA and AA the two possible variants. Our results suggest that the A/G SNP at the rs1784933 locus is associated with LOAD in the Chinese Han population of the Hunan Changsha region, with the risk allele being the A allele (OR, 1.569; 95% CI, 1.16–2.122). Multifactor logistic regression analysis correcting for confounding factors such as gender, age, and cholesterol, found that the LOAD risk in AA genotype carriers was 1.803 times those carrying AG + GG genotypes (OR, 1.803; 95% CI, 1.225–2.654). Lee et al. [16] examined SORL1 genetic variation in 296 AD patients and 428 healthy elderly controls, and found that rs1784933 was significantly associated with AD in the African American population (P = 0.019). AD risk in G allele carriers was less than in A allele carriers, in accordance with our findings in this study. In conclusion, the relationship between rs689021, rs3824966, and rs1784933 loci of the SORL1 gene and AD onset may be influenced by many factors. Our study shows that polymorphism of the rs1784933 locus is associated with LOAD onset in AD patients from the Hunan Changsha region. There are SNPs at multiple SORL1 gene loci, therefore to further understand the correlation between SORL1 and AD, we need to identify more loci for studies of larger populations. This should provide a basis for clarifying AD pathogenesis and identifying new therapeutic targets. Acknowledgments This work was supported by the Community Development Supporting Plan Project of Hunan Provincial Science and Technology Department, China (Nos. 009SK3175 and 2012SK3218). References [1] T. Kanekiyo, H. Xu, G. Bu, ApoE and A in Alzheimer’ disease: accidental encounters or partners? Neuron 81 (2014) 740–754. [2] W.J. Davis, The ATP-binding cassette transporter-2 (ABCA2) regulates esterification of plasma membrane cholesterol by modulation of sphingolipid metabolism, Biochim. Biophys. Acta 1841 (2014) 168–179. [3] P. Gamba, G. Testa, B. Sottero, S. Gargiulo, G. Poli, G. Leonarduzzi, The link between altered cholesterol metabolism and Alzheimer’s disease, Ann. N. Y. Acad. Sci. 1259 (2012) 54–56. [4] M. Maulik, D. Westaway, J.H. Jhamandas, S. Kar, Role of cholesterol in APP metabolism and its significance in Alzheimer’s disease pathogenesis, Mol. Neurobiol. 47 (2013) 37–63. [5] G.M. Shankar, S. Li, T.H. Mehta, A. Garcia-Munoz, N.E. Shepardson, I. Smith, F.M. Brett, M.A. Farrell, M.J. Rowan, C.A. Lemere, C.M. Regan, D.M. Walsh, B.L. Sabatini, D.J. Selkoe, Amyloid-beta protein dimers isolated directly from Alzheimer’s brains impair synaptic plasticity and memory, Nat. Med. 14 (2008) 837–842.
X. Feng et al. / Neuroscience Letters 584 (2015) 382–389 [6] A. Schneider, W. Schulz-Schaeffer, T. Hartmann, J.B. Schulz, M. Simons, Cholesterol depletion reduces aggregation of amyloid-beta peptide in hippocampal neurons, Neurobiol. Dis. 23 (2006) 573. [7] L. Bertram, M.B. McQueen, K. Mullin, D. Blacker, R.E. Tanzi, Systematic meta-analyses of Alzheimer disease genetic association studies: the AlzGene database, Nat. Genet. 39 (2007) 17–23. [8] E. Rogaeva, Y. Meng, J.H. Lee, Y. Gu, T. Kawarai, F. Zou, T. Katayama, C.T. Baldwin, R. Cheng, H. Hasegawa, F. Chen, N. Shibata, K.L. Lunetta, R. Pardossi-Piquard, C. Bohm, Y. Wakutani, L.A. Cupples, K.T. Cuenco, R.C. Green, L. Pinessi, I. Rainero, S. Sorbi, A. Bruni, R. Duara, R.P. Friedland, R. Inzelberg, W. Hampe, H. Bujo, Y.Q. Song, O.M. Andersen, T.E. Willnow, N. Graff-Radford, R.C. Petersen, D. Dickson, S.D. Der, P.E. Fraser, G. Schmitt-Ulms, S. Younkin, R. Mayeux, L.A. Farrer, S. George, P. Hyslop, The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease, Nat. Genet. 39 (2007) 168–177. [9] X. Xue, M. Zhang, Y. Lin, E. Xu, J. Jia, Association between the SORL1 rs2070045 polymorphism and late-onset Alzheimer’s disease: interaction with the ApoE genotype in the Chinese Han population, Neurosci. Lett. 559 (2014) 94–98. [10] C. Jin, L. Zhang, Y. Xian, X. Liu, Y. Wu, F. Zhang, J. Zhu, G. Zhang, C. Chen, R. Gong, L. Zhang, J. Yuan, F. Zhang, L. Tian, G. Wang, Z. Cheng, The SORL1 polymorphism rs985421 may confer the risk for amnestic mild cognitive impairment and Alzheimer’s disease in the Han Chinese population, Neurosci. Lett. 563 (2014) 80–84.
389
[11] X. Gao, M. Liu, L. Sun, B. Qin, H. Yu, Z. Yang, R. Qi, F. Gao, SORL1 genetic variants modulate risk of amnestic mild cognitive impairment in northern Han Chinese, Int. J. Neurosci. 124 (2014) 296–301. [12] F. Liu, M.A. Ikram, A.C. Janssens, M. Schuur, I.D. Koning, A. Isaacs, M. Struchalin, A.G. Uitterlinden, J.T. den Dunnen, K. Sleegers, K. Bettens, C.V. Broeckhoven, J.V. Swieten, A. Hofman, B.A. Oostra, Y.S. Aulchenko, M.M. Breteler, C.M. Duijn, A study of the SORL 1 genein Alzheimer’s disease and cognitive function, J. Alzheimer Dis. 18 (2009) 51–64. [13] R.L. Minster, S.T. DeKosk, M.I. Kamboh, No association of SORL 1 SNPs with Alzheimer’s disease, Neurosci. Lett. 440 (2008) 190–192. [14] K. Bettens, N. Brouwers, S. Engelborghs, P.P.D. Deyn, C.V. Broeckhoven, K. Sleegers, SORL1 is genetically associated with increased risk for late-onset Alzheimer disease in the Belgian population, Hum. Mutat. 29 (2008) 769–770. [15] M. Ning, Y. Yang, Z. Zhang, Z. Chen, T. Zhao, D. Zhang, D. Zhou, J. Xu, Z. Liu, Y. Wang, Y. Liu, X. Zhao, W. Li, S. Li, L. He, Amyloid--related genes SORL1 and ACE are genetically associated with risk for late-onset Alzheimer disease in the Chinese population, Alzheimer Dis. Assoc. Disord. 24 (2010) 390–396. [16] J.H. Lee, R. Cheng, N. Schupf, J. Manly, R. Lantigua, Y. Stern, E. Rogaeva, Y. Wakutani, L. Farrer, P.S. George, R. Mayeux, The association between genetic variants in SORL1 and Alzheimer’s disease in an urban, multiethnic, community-based cohort, Arch. Neurol. 64 (2007) 501–506.
