Skip Navigation


Human Molecular Genetics Advance Access originally published online on July 5, 2007
Human Molecular Genetics 2007 16(18):2199-2208; doi:10.1093/hmg/ddm171
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
16/18/2199    most recent
ddm171v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (6)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Belbin, O.
Right arrow Articles by Morgan, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Belbin, O.
Right arrow Articles by Morgan, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© The Author 2007. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oxfordjournals.org

Regulatory region single nucleotide polymorphisms of the apolipoprotein E gene and the rate of cognitive decline in Alzheimer's disease

Olivia Belbin1, Janette L. Dunn2, Yan Ling1, Linda Morgan1, Sally Chappell1, Helen Beaumont3, Donald Warden3, David A. Smith3, Noor Kalsheker1 and Kevin Morgan1,*

1 Division of Clinical Chemistry, Institute of Genetics, Queen's Medical Centre, University of Nottingham, Nottingham NG7 2UH , UK, 2 School of Physics and Astronomy, University of Nottingham, University Park, Nottingham NG7 2RD, UK and 3 OPTIMA, Oxford Centre for Gene Function, Department of Physiology, Anatomy and Genetics, Oxford OX1 3PT, UK

* To whom correspondence should be addressed. Tel: +44 1158230724; Fax: +44 1159709167; Email: kevin.morgan{at}nottingham.ac.uk

Received May 11, 2007; Accepted June 28, 2007


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
The aim of this study was to investigate whether single nucleotide polymorphisms (SNPs) in the regulatory regions of the apolipoprotein E (APOE) gene modify the well-established {varepsilon}4-associated risk for Alzheimer's disease (AD). Sequencing of the APOE gene regulatory regions revealed four previously reported promoter SNPs and one novel SNP in the previously described macrophage enhancer (ME.1). In addition, we also studied the two classic allelic missense SNPs that define {varepsilon}2/{varepsilon}3/{varepsilon}4 status in a case–control association study. Analysis of pair-wise linkage disequilibrium (LD) of the five regulatory region SNPs with classic APOE SNPs revealed a previously unreported 7 kb LD block covering the entire APOE gene, part of the promoter and 3' enhancer region. We report here that in a case–control association study (N=719) of the seven SNPs, the genotype at codon 112 captures all the information required to assess disease risk. To explore correlations with quantitative traits, 169 patients were studied in whom rates of cognitive decline were available. In addition to the {varepsilon}4 allele, two regulatory region SNPs were associated with the rate of cognitive decline in AD patients. This study highlights the effect of APOE gene variation on risk of AD and rate of cognitive decline and demonstrates that a single SNP, which confers {varepsilon}4 status, captures all of the risk of developing AD but two SNPs in the regulatory region may affect the rate of cognitive decline in AD patients.


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
Familial early-onset Alzheimer's disease (AD) accounts for ~5–10% of AD cases and most can be explained by inheritance of known mutations in amyloid precursor protein (APP) and presenilins (PSEN) 1 and 2 genes in an autosomal dominant fashion (1). The remaining AD cases are associated with a later age at onset and represent ‘sporadic’ AD, although these sporadic cases may also have a significant genetic component. Several genes have been implicated as potential candidates for this late-onset AD (LOAD) but the apolipoprotein (APOE) gene is the only one to date that has been replicated in numerous studies (2–4). The protein exists in three major isoforms apoE2, apoE3 and apoE4 that arise from three APOE alleles: APOE {varepsilon}2, APOE {varepsilon}3 and APOE {varepsilon}4 as a result of two missense single nucleotide polymorphisms (SNPs), rs429358 and rs7412, which result in amino acid changes at residues 112 and 158, respectively (5). More detail on these two SNPs is shown in Table 1. Henceforth in this manuscript, these two polymorphisms are designated 112 and 158. Approximately 50% of LOAD cases are accounted for by individuals possessing the APOE {varepsilon}4 allele (3), although possession of this allele is not essential for predisposition to the disease. Possession of this allele not only increases the risk of developing AD (lifetime risk increases from 20% when no APOE {varepsilon}4 alleles are possessed to 90% when two copies are present) but it also decreases the age at onset [mean age at onset for individuals with no {varepsilon}4 allele was 84.3 years compared with 68.8 years for individuals possessing two copies (6)]. Interestingly, the APOE {varepsilon}2 allele is under-represented in AD cases (7) and is associated with a delayed age at onset of AD (2,6) suggesting a protective effect (8).


View this table:
[in this window]
[in a new window]

  		Table 1. Genotypes and the resulting amino acid changes associated with APOE allelic status

 
The APOE gene resides on chromosome 19q13.2 (9,10) and codes for the cholesterol transporter protein, apolipoprotein E (apoE), which represents the major lipoprotein within the CNS, where it is synthesized by astrocytes (11,12). It has been suggested that one role of apoE in the brain may be an involvement in neuronal homoestasis (13), particularly, mobilization of cholesterol in the CNS, where it is required for neuronal plasticity (14,15). ApoE is also postulated to play a role in neuronal repair by mediating the recycling of damaged cell membranes (13). Levels of apoE mRNA are increased in astrocytes following lesions to the entorhinal cortex (7), one of the first areas of brain to be affected by AD (16).

Although the APOE {varepsilon}4 allele confers a considerable risk for AD, not all individuals who possess this allele go on to develop the disease. This observation has led to the hypothesis that other genetic factors, possibly elsewhere within the APOE gene, may contribute to disease risk. One potential area of interest is the promoter region. To date, four promoter SNPs have been identified at positions –491, –427, –219 and +113 base pairs relative to the transcription start site (17–20).

The –491A/T polymorphism has been associated with an increased risk of developing AD independently from the {varepsilon}4-associated risk (18,21,22). The –427T/C, –219 T/G and +113G/C polymorphisms have also been associated with an increased risk for AD (21,23,24), although confirmation of these findings has been limited possibly due to the strong linkage disequilibrium (LD) with the {varepsilon}4 allele.

Functionally, these promoter SNPs have been proposed to affect the transcriptional activity of the gene in both the periphery and the CNS due to altered binding of nuclear proteins (13). Taking –491A, –427T and –219T as a reference haplotype, an A to T subsititution at –491 resulted in a 63% decrease and a T to G substitution at –219 resulted in a 169% increase in APOE promoter activity in HepG2 liver cells (17) and similar effects were reported in astrocytoma cells (23). The T to C substitution at –427 had no effect on promoter activity (17). Differential protein binding was observed for both –491 and –219 polymorphisms. The –491AA genotype is also associated with raised apoE protein levels in neuropathologically normal individuals (25). Since the risk of sporadic AD is associated with increased levels of apoE in plasma (26), it is possible that promoter polymorphisms could contribute to risk by increasing APOE expression.

In addition to the 5' proximal promoter (27–32), other control elements for APOE gene expression have been reported in the 3' flanking region of the gene. The functional peroxisome-proliferator-activated-receptor gamma response element (PPRE) is situated 2 kb downstream of the APOE gene (33), whereas a macrophage enhancer (ME.1) can be found 1.3 kb further downstream (34–36). The role of potential SNPs in these regions on APOE gene expression and any association with AD has yet to be determined.

In this paper, we report a case–control SNP association study of regulatory region SNPs in the promoter and ME.1 enhancer element as well as the established AD risk-associated missense SNPs. We also examine the effect of APOE SNPs on cognitive decline by non-linear mixed effects modelling.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 
SNP identification
In addition to the previously described promoter SNPs [–491 rs449647, –427 rs769446, –219 rs405509 (17) and +113 rs440446 (20)], one novel SNP was identified in the ME.1 region at position +7203 from the start of the APOE gene (Fig. 1). No SNPs were identified in the PPRE. The adjacent sequences for all SNPs and their minor allele frequencies are shown in Table 2. All SNPs had a minimum minor allele frequency of 5% during the mapping of 88 alleles. Since the genotypes for missense SNPs, 112 and 158, were already known for these samples, TaqMan assays were designed for the remaining five SNPs and genotyping was performed in the total 719 individuals. All polymorphisms were in Hardy–Weinberg equilibrium in controls and AD patients.


Figure 1
View larger version (21K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 1. SNP map of the APOE gene showing SNPs, regulatory and coding regions.

 


View this table:
[in this window]
[in a new window]

  		Table 2. Adjacent sequences for all SNPs genotyped in this study and the minor allele frequencies (N=44)

 
Haplotypes of the APOE gene
Using the five regulatory region SNPs, seven common haplotypes were identified, which accounted for over 86% APOE variation in controls (Table 3). Haplotypes are also presented for cases and controls in the total data set (Table 4) using the five regulatory SNPs plus the two SNPs that define {varepsilon}2/{varepsilon}3/{varepsilon}4 status. Ten common haplotypes (frequency >2% in controls) were apparent, of which, only one represented the {varepsilon}2 genotype (H1). This haplotype showed a trend for increased frequency in controls (4% in controls compared with 2% in AD patients) but was not statistically significant (P=0.32) Seven haplotypes (H2–H8) represented the {varepsilon}3 genotype. H2 and H3 were strongly protective (P=2x10–06 and 5x10–05, respectively), whereas the remaining five {varepsilon}3 haplotypes (H4–H8) were at best marginally protective or neutral. Two haplotypes (H9 and H10) represented the {varepsilon}4 genotype, both of which were at a statistically significant increased frequency in AD patients compared with controls (P=3x10–03 and 2x10–14, respectively). The statistically stronger association of H10 with AD than H9 is likely due to the higher frequency of the haplotype rather than a modifying effect of the regulatory SNPs on {varepsilon}4 status. However, the numbers of {varepsilon}4-positive controls are limited and studies with a larger data sets will be able to directly address this issue.


View this table:
[in this window]
[in a new window]

  		Table 3. Frequencies of the seven common haplotypes at five SNPs in the APOE gene regulatory regions in controls from Nottingham (N=93) and Oxford (N=174)

 


View this table:
[in this window]
[in a new window]

  		Table 4. Frequencies of the haplotypes at seven SNPs in the APOE gene in controls (N=261) and AD patients (N=317). Samples were only included when genotype information was available at all seven SNPs

 
LD of SNPs in the APOE gene
The pair-wise LD measures (r2 and D') for all seven SNPs are shown in Figure 2. Pair-wise r2 values ranged from 0.001 to 0.66 indicating that all seven SNPs provide a degree of unique haplotypic information. The D' values (Fig. 2C and D) revealed a strong degree of LD between the seven SNPs. In particular, an LD block (all D'>0.40) can be observed in the 3' 7 kb region that includes SNPs –219, +113, 112, 158 and ME.1 in both controls and AD patients. This LD within the APOE gene confounds attempts to differentiate potential promoter associations from the highly significant {varepsilon}4 effect.


Figure 2
View larger version (40K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 2. Pair-wise LD measures; r2 in controls (A) and AD (B), and D' in controls (C) and AD (D). Boxes are coloured according to LD values: <0.01, white; 0.01–0.8, shades of grey; >0.8, black. SNPs within the 7 kb LD block are shown in bold.

 
Single SNP associations
As expected, the {varepsilon}4 effect was strong in this data set; OR for possession of one copy of the {varepsilon}4 allele was 4.6 (95% CI 3.3–6.3) and for possession of two copies was 12.5 (95% CI 3.8–40.7). Chi-squared tests of allele frequency versus disease status identified 112 as the major disease-associated SNP in the total data set (Table 5). The cytosine (C) allele at this locus (possessed by {varepsilon}4 bearers only) is the major risk allele, representing 12.4% of controls and 34.7% of AD patients (P=4x10–21). The other APOE allelic missense SNP, 158, shows a relatively weaker association; the thymine (T) allele (corresponding to the {varepsilon}2 allele) is at a higher frequency in controls, 8.9%, compared with AD, 5.1% (P=0.007).


View this table:
[in this window]
[in a new window]

  		Table 5. Single SNP allele frequencies and associated P-values for allelic associations

 
Association of the other five SNPs examined in this study also showed relatively weak associations when compared with 112 in the total data set (Table 5). Although, four SNPs (–491, –219, +113 and ME.1) resulted in P<0.05, because of the strong LD of these SNPs with 112 and 158, it cannot be assumed that these are independent associations without correcting for the {varepsilon}4 effect.

Correcting for the {varepsilon}4 effect
In this study, we have used two approaches for determining {varepsilon}4-independent associations. First, we split the data set by possession of the {varepsilon}4 allele, resulting in two separate subsets; those individuals who possess no {varepsilon}4 allele and those who possess one or two copies. Table 5 shows the minor allele frequency of the seven SNPs in non-{varepsilon}4 individuals and {varepsilon}4 bearers. Since non-{varepsilon}4 individuals always possess a T allele at 112, this SNP was not included in this part of the analysis. As shown in Table 5, no other single SNP showed association with control or AD in non-{varepsilon}4 individuals. The fact that 158 is not associated with AD in these individuals merely demonstrates that there is no significant difference in the frequency of the C allele (representing {varepsilon}3) between non-{varepsilon}4 controls and non-{varepsilon}4 AD patients.

The allelic association of 112 and 158 with AD also disappeared in {varepsilon}4 bearers (Table 5). The number of C alleles at 112 in {varepsilon}4 bearers is a direct measure of the number of {varepsilon}4 homozygotes and there was no significant difference in the frequency of the C allele between {varepsilon}4-bearing controls and {varepsilon}4-bearing AD patients. Likewise, the lack of association of 158 indicates that there is no significant difference in the number of {varepsilon}2/{varepsilon}4 individuals in {varepsilon}4-positive cases and controls. Two promoter SNPs demonstrated weak associations in {varepsilon}4 bearers: –491 (T allele demonstrated 16.0% frequency in controls compared with 8.5% in AD, P=0.016) and –219 (T allele demonstrated 63.6% frequency in AD compared with 54.1% in controls, P=0.05). The finding that the T allele at –491 has a protective effect in {varepsilon}4 bearers is in agreement with previous studies (18,21,22). However, this approach has limited power to detect associations due to the reduction of sample size and the relative lack of {varepsilon}4-positive controls (N=75) compared with {varepsilon}4 AD patients (N=225).

In addition, we have used a second approach, genotype logistic regression, to examine the OR of developing AD for each SNP while adjusting for the covariates 112, 158, gender and age at onset (AD patients) and age at examination (controls). This approach has the advantage of using the total data set resulting in greater power to detect an association and hence is likely to reveal effects that are not with the above analyses. As shown in Table 5, all SNPs except –427, demonstrated a significant association with AD when analysing each SNP independently but only 112 remained significant when correcting for covariates (OR 4.65, 95% CI 1.38–15.67). This suggests that the association of –491 and –219 reported in {varepsilon}4 positives could be an artefact due to the low number of {varepsilon}4 positive controls.

Association of APOE SNPs with cognitive decline
Here we used cognitive decline as a quantitative trait in AD patients, thus refining the phenotype. As has been previously described (37), cognitive decline data are more suited to a non-linear model rather than linear. We examined the effects of APOE SNPs on age and rate of cognitive decline using this previously described non-linear mixed effects model. Since there was no significant difference in age at midpoint of decline or rate of decline between males and females in this data set (data not shown), no corrections were made for gender in this study. Age at which patients reach half the asymptotic score is determined by the relative position of curves along the x-axis. The rate at which patients' cognitive decline falls from three-fourths to half the asymptotic score is determined by the slope of the curve. As shown in Figure 3A, there were no significant effects of the classical APOE genotypes on age at midpoint of cognitive decline but a strong effect of the {varepsilon}4 allele was apparent on the rate of decline. {varepsilon}4 homozygotes showed a significantly faster rate of decline (1.26 years to fall from a score of 71.27 to 47.5) than wild-type {varepsilon}3 homozygotes (1.62 years, P=0.002), whereas individuals with {varepsilon}2{varepsilon}3 or {varepsilon}3{varepsilon}4 genotypes showed no significant difference.


Figure 3
View larger version (24K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Figure 3. Non-linear model of cognitive decline demonstrating the effect of (A) classic APOE genotypes and (B-F) APOE regulatory region SNPs. Asterisks indicate significant effects.

 
In the next stage of analysis, we added the five regulatory region APOE SNPS to the model (Fig. 3B–F). Although no SNP had a significant effect on the age at midpoint of cognitive decline, +113 and ME.1 demonstrated significant genotype-specific effects on rate of cognitive decline. For the +113 SNP, the GG genotype demonstrated significantly faster cognitive decline (1.45 years) than the reference GC genotype (1.68 years; P=0.04). For the ME.1 SNP, the TT genotype demonstrated significantly faster cognitive decline (0.94 years) compared with the reference CC genotype (1.49 years; P=0.0005). The ME.1 SNP therefore demonstrates the strongest association of any SNP in the APOE gene on rate of cognitive decline. In an attempt to correct for {varepsilon}4 status on the effect of +113 and ME.1 SNPs on cognitive decline, the data were split into {varepsilon}4 positive and negative patients; however, it must be noted that this approach reduces the power as the sample numbers decrease. Nevertheless, when the modelling was repeated on these data sets, the +113 effect was seen in both {varepsilon}4-positive and -negative patients with the effect on rate of cognitive decline strengthening in the {varepsilon}4-negative patients (P=0.0002; data not shown). The ME.1 effect became non-significant in both {varepsilon}4-positive and -negative patients. These data suggest that the +113 effect on cognitive decline appears not to be dependent on {varepsilon}4.


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
MATERIALS AND METHODS
 REFERENCES
 
Here we report an investigation of seven SNPs within the regulatory and coding regions of the APOE gene, providing a detailed report of their LD, a large AD case–control study and quantitative trait analysis of cognitive decline in AD patients. The fact that possession of the {varepsilon}4 allele is not essential for predisposition to the disease has led researchers to believe that other risk factors are yet to be discovered, possibly in the APOE gene itself. However, the strong LD within this region has, in the past, confounded attempts to identify susceptibility alleles other than {varepsilon}4. Of the seven SNPs studied, only frequencies for –219 (rs405509) and 112 (rs429358) are reported in the current version of HapMap (www.hapmap.org, data release 21a, phase II, January 2007) but all except the novel ME.1 SNP are listed in dbSNP (www.ncbi.nlm.nih.gov/SNP). Here we report a 7 kb linkage block stretching from the –219 promoter SNP to the 3' macrophage enhancer region 4 kb downstream of the APOE gene using genotype frequency data generated in this study. When examining these SNPs in a case–control study, we revealed no association that exceeds, or is independent of, the previously reported {varepsilon}4-associated risk.

It is possible that by looking at such a hugely variable phenotype as AD, we lack the power to isolate promoter associations from the {varepsilon}4 effect. Here we demonstrate that splitting the data set into {varepsilon}4-positive and -negative individuals in an attempt to do so is insufficient for two reasons. First, this is ignoring the {varepsilon}4 dose effect that is apparent in such studies. Potential associations in {varepsilon}4 positives that have been reported in the past could merely be reflecting the effect of {varepsilon}4 homozygotes versus {varepsilon}4 heterozygotes. Secondly, the use of such small subsets for a hugely variable phenotype results in a significant loss of power to detect associations and the lack of {varepsilon}4 positive controls compared with {varepsilon}4 positive AD patients could shed doubt on positive associations reported in the {varepsilon}4-positive data set.

For these reasons, we examined the role of the APOE SNPs in a more refined AD phenotype of cognitive decline (N=169). Here we report an association of {varepsilon}4 homozygosity with a significantly faster rate of cognitive decline and a trend towards earlier age at midpoint of decline in AD patients, a finding that is consistent with previous observations (37). These authors studied 218 AD patients and found that {varepsilon}4 associated with a faster rate and earlier decline. Interestingly, in another previous study, the APOE {varepsilon}4 allele has been shown to influence conversion of mild cognitive impairment to AD (38). This study failed to find any evidence for an {varepsilon}4 effect on the rate of cognitive decline in AD patients. However, the cognitive decline data were partitioned into groups of fast and slow decline as oppposed to implementing non-linear effects modelling as conducted in our study.

Further to this effect, we found two regulatory region SNPs that were also associated with an altered rate of cognitive decline, one of which (+113) demonstrated association in both {varepsilon}4-positive and -negative individuals, possibly indicating that it is not tracking with {varepsilon}4 status. This quantitative trait approach provides greater power than case–control studies and has indicated that regulatory region variation in the APOE gene may contribute to a faster cognitive decline in AD patients. The ME.1 SNP result exemplifies this point since, although its association with disease risk becomes non-significant after correction for confounders, the quantitative trait studies demonstrate that its effect on the rate of cognitive decline is readily apparent.

It is interesting to note that all associations reported here for the APOE gene affected the rate but not age at midpoint of cognitive decline in AD patients. We have previously reported that possession of a GG genotype of an intronic SNP of the alpha-1 antichymotrypsin (ACT) gene associated with a significantly earlier age at midpoint of cognitive decline than the other two genotypes at this SNP (39). This demonstrates that SNPs in two AD candidate genes (ACT and APOE) affect different parameters (rate and age at midpoint) of cognitive decline. Further larger studies need to be conducted investigating these parameters in order to reveal any potential epistatic interactions between the APOE and ACT genes; currently, we do not have the power to discern these effects.

In conclusion, the results of this study demonstrate that the association of the APOE gene with AD can be captured by just looking at a single SNP (112), which defines carrier status for the {varepsilon}4 allele. This is in agreement with a recently published paper, which also alluded to the value of 112 as a ‘tag’ for AD risk (40). The haplotype data again tracks with {varepsilon}4 status, and from the curent data, it would appear that the regulatory variants have at best minimal interaction with the {varepsilon}4 allele. It is likely that the previous associations reported for a number of other functional SNPs in the promoter area are due to the fact that they track with {varepsilon}4 status and as a result do not impart any additional information. However, when using cognitive decline as a quantitative trait to further refine the phenotype, association with APOE SNPs additional to {varepsilon}4 emerge.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
REFERENCES
 
Patient samples
All subjects gave informed consent to be included in the study, which was granted approval by the local Ethics Committee. For the case–control association study, the 719 samples were obtained from two UK centres: University of Nottingham Brain Bank and Oxford Project to Investigate Memory and Ageing (OPTIMA), University of Oxford. Details of patient samples used in this study are as follows: 324 controls (162 males, 158 females, four with undocumented gender, mean age 75.3±9.4 years) and 395 AD patients (161 males, 227 females, seven with undocumented gender, mean age at onset 75.6±10.5 years). There was no evidence for population stratification in this sample set in a previous study (39); thus, the samples were pooled for all analyses. Samples were histopathologically confirmed as definite disease (AD) (N=310) or control (N=324) using CERAD criteria (41). Probable AD patients from Oxford (N=85) were also included since there has been 100% concordance between patients diagnosed with probable AD and confirmation of the disease post-mortem (N=34) (42). All patients with evidence of an autosomal dominant AD trait, or where a first-degree relative had been diagnosed with familial AD, were excluded.

PCR and automated sequencing
Genomic DNA was extracted from whole blood or brain tissue using the QIAamp DNA blood mini kit (Qiagen, Crawley, West Sussex, UK) following manufacturer's procedure. All amplification protocols were performed with an annealing temperature of 60°C; primer sequences are shown in Figure 1.

SNP identification
Forty-four samples were used to identify common SNPs (minor allele frequency >5%) within the APOE gene by sequencing known regulatory regions (promoter, 3'PPRE and 3'ME.1 enhancer region). This approach has 99% power for the detection of polymorphisms that are present at a frequency of ≥5%. Potential polymorphic sites were identified by multiple sequence alignment using ClustalW software. Polymorphisms were confirmed by repeat PCR and sequencing of the opposite strand.

Genotyping of polymorphisms
All 719 samples were genotyped at five sites using fluorescently labelled TaqMan probes (Vic or Fam) by Geneservice (Cambridge, UK). Fifteen percent of the samples assayed were of known genotype, determined by sequencing, and 10% were genotyped in duplicate as quality assurance measures. The data were only accepted when there was 100% concordance between duplicate samples. The two APOE allelic SNPs, 112 and 158, were genotyped by RFLP using a previously described protocol (43). Samples in which genotyping failed at more than three sites were not included in the analyses. For each of the SNPs, the failure rates were as follows; –491:4.6%, –427:7.2%, –219:4.0%, +113:4.3%, 112:2.8%, 158:3.1% and ME.1:3.1%.

Haplotypes, linkage and case–control association
The five regulatory region SNPs were used to estimate the haplotype frequencies found in controls and AD patients using the haplo.em function of haplo.stats v1.2.2 in the R programming language (44). This method utilizes a maximum likelihood analysis approach, which we have previously used (39). LD measures were calculated using haploview software (www.broad.mit.edu). Chi-squared tests and multivariate binary logistic regression of individual SNPs with disease status were performed using Statistical Package for Social Sciences (SPSS) v12.0.1.

Association of APOE SNPs with cognitive decline
Cognitive scores were obtained using the CAMCOG score system (45). In two previous studies using samples from the same OPTIMA collection, the CAMCOG score data could be fitted using a non-linear mixed effects model (37,39). Analysis of the CAMCOG data in this study was performed using the same methods. Briefly, we used a three-parameter logistic (S-shaped) function: CAMCOG=asymptote/{1+exp[(age–xmid)/scale]}. The asymptotic score was set at a CAMCOG score of 95, in common with the previous work. The xmid parameter is the age at which patients reach 50% of the asymptotic score (CAMCOG=47.5) and scale is the time taken to fall from three-fourths (CAMCOG=71.25) to half the asymptotic score. We first modelled the interactive effects of age with {varepsilon}2/{varepsilon}3/{varepsilon}4 status (fixed effects) upon the xmid and scale parameters. Random effects were also included for xmid and scale, recognizing that different patients will obtain a score of 50% of the asymptotic value at different ages and fall from three-fourths to half the asymptotic score at different rates. We then modelled the five regulatory region SNPs using a similar model. The model was implemented using the non-linear mixed-effects function ‘nlme’ of the nlme library v3.1-77 in the R programming suite (46). P-values for the effect of each genotype compared with a designated reference genotype were obtained for each SNP. In the case of {varepsilon}2/{varepsilon}3/{varepsilon}4 status, this was the wild-type {varepsilon}3{varepsilon}3 genotype, and for the five regulatory region SNPs, it was the most frequent homozygous genotype. Genotypes present in less than 10 individuals were excluded from the study.


   ACKNOWLEDGEMENTS
 
We would like to thank the Alzheimer's Research Trust for funding this collaboration and the Medical Research Council for funding J.L.D. with a discipline hopping award. We would also like to thank the OPTIMA (Oxford Project to Investigate Memory and Ageing) for providing the Oxford sample collection used in the haplotype analyses phase of this study.

Conflict of Interest statement. All authors declare that they have no conflict of interest towards the work presented in this manuscript.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 RESULTS
 DISCUSSION
 MATERIALS AND METHODS
 REFERENCES
 

  1. Bertram L., Tanzi R.E. Alzheimer's disease: one disorder, too many genes? Hum Mol Genet (2004) 13(1):R135–R141.[Abstract/Free Full Text]

  2. Chartier-Harlin M.C., Parfitt M., Legrain S., Perez-Tur J., Brousseau T., Evans A., Berr C., Vidal O., Roques P., Gourlet V., et al. Apolipoprotein E, epsilon 4 allele as a major risk factor for sporadic early and late-onset forms of Alzheimer's disease: analysis of the 19q13.2 chromosomal region. Hum. Mol. Genet. (1994) 3:569–574.[Abstract/Free Full Text]

  3. Saunders A.M., Strittmatter W.J., Schmechel D., George-Hyslop P.H., Pericak-Vance M.A., Joo S.H., Rosi B.L., Gusella J.F., Crapper-MacLachlan D.R., Alberts M.J., et al. Association of apolipoprotein E allele epsilon 4 with late-onset familial and sporadic Alzheimer's disease. Neurology (1993) 43:1467–1472.[Abstract/Free Full Text]

  4. Strittmatter W.J., Saunders A.M., Schmechel D., Pericak-Vance M., Enghild J., Salvesen G.S., Roses A.D. Apolipoprotein E: high-avidity binding to beta-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc. Natl Acad. Sci. USA (1993) 90:1977–1981.[Abstract/Free Full Text]

  5. Weisgraber K.H., Innerarity T.L., Mahley R.W. Abnormal lipoprotein receptor-binding activity of the human E apoprotein due to cysteine–arginine interchange at a single site. J. Biol. Chem. (1982) 257:2518–2521.[Free Full Text]

  6. Corder E.H., Saunders A.M., Strittmatter W.J., Schmechel D.E., Gaskell P.C., Small G.W., Roses A.D., Haines J.L., Pericak-Vance M.A. Gene dose of apolipoprotein E type 4 allele and the risk of Alzheimer's disease in late onset families. Science (1993) 261:921–923.[Abstract/Free Full Text]

  7. Poirier J., Hess M., May P.C., Finch C.E. Astrocytic apolipoprotein E mRNA and GFAP mRNA in hippocampus after entorhinal cortex lesioning. Brain Res. Mol. Brain Res. (1991) 11:97–106.[Medline]

  8. Corder E.H., Saunders A.M., Risch N.J., Strittmatter W.J., Schmechel D.E., Gaskell P.C. Jr, Rimmler J.B., Locke P.A., Conneally P.M., Schmader K.E., et al. Protective effect of apolipoprotein E type 2 allele for late onset Alzheimer disease. Nat. Genet. (1994) 7:180–184.[CrossRef][Web of Science][Medline]

  9. Das H.K., McPherson J., Bruns G.A., Karathanasis S.K., Breslow J.L. Isolation, characterization, and mapping to chromosome 19 of the human apolipoprotein E gene. J. Biol. Chem. (1985) 260:6240–6247.[Abstract/Free Full Text]

  10. Olaisen B., Teisberg P., Gedde-Dahl T. Jr. The locus for apolipoprotein E (apoE) is linked to the complement component C3 (C3) locus on chromosome 19 in man. Hum. Genet. (1982) 62:233–236.[CrossRef][Web of Science][Medline]

  11. Mahley R.W. Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science (1988) 240:622–630.[Abstract/Free Full Text]

  12. Mahley R.W., Innerarity T.L., Rall S.C. Jr, Weisgraber K.H. Plasma lipoproteins: apolipoprotein structure and function. J. Lipid Res. (1984) 25:1277–1294.[Abstract]

  13. Laws S.M., Hone E., Gandy S., Martins R.N. Expanding the association between the APOE gene and the risk of Alzheimer's disease: possible roles for APOE promoter polymorphisms and alterations in APOE transcription. J. Neurochem. (2003) 84:1215–1236.[CrossRef][Web of Science][Medline]

  14. Moestrup S.K., Gliemann J., Pallesen G. Distribution of the alpha 2-macroglobulin receptor/low density lipoprotein receptor-related protein in human tissues. Cell Tissue Res. (1992) 269:375–382.[CrossRef][Web of Science][Medline]

  15. Rebeck G.W., Reiter J.S., Strickland D.K., Hyman B.T. Apolipoprotein E in sporadic Alzheimer's disease: allelic variation and receptor interactions. Neuron (1993) 11:575–580.[CrossRef][Web of Science][Medline]

  16. Morris J.C. The challenge of characterizing normal brain aging in relation to Alzheimer's disease. Neurobiol. Aging (1997) 18:388–389. discussion.[CrossRef][Web of Science][Medline]

  17. Artiga M.J., Bullido M.J., Sastre I., Recuero M., Garcia M.A., Aldudo J., Vazquez J., Valdivieso F. Allelic polymorphisms in the transcriptional regulatory region of apolipoprotein E gene. FEBS Lett. (1998) 421:105–108.[CrossRef][Web of Science][Medline]

  18. Bullido M.J., Artiga M.J., Recuero M., Sastre I., Garcia M.A., Aldudo J., Lendon C., Han S.W., Morris J.C., Frank A., et al. A polymorphism in the regulatory region of APOE associated with risk for Alzheimer's dementia. Nat. Genet. (1998) 18:69–71.[CrossRef][Web of Science][Medline]

  19. Lambert J.C., Pasquier F., Cottel D., Frigard B., Amouyel P., Chartier-Harlin M.C. A new polymorphism in the APOE promoter associated with risk of developing Alzheimer's disease. Hum. Mol. Genet. (1998) 7:533–540.[Abstract/Free Full Text]

  20. Mui S., Briggs M., Chung H., Wallace R.B., Gomez-Isla T., Rebeck G.W., Hyman B.T. A newly identified polymorphism in the apolipoprotein E enhancer gene region is associated with Alzheimer's disease and strongly with the epsilon 4 allele. Neurology (1996) 47:196–201.[Abstract/Free Full Text]

  21. Lambert J.C., Berr C., Pasquier F., Delacourte A., Frigard B., Cottel D., Perez-Tur J., Mouroux V., Mohr M., Cecyre D., et al. Pronounced impact of Th1/E47cs mutation compared with -491 AT mutation on neural APOE gene expression and risk of developing Alzheimer's disease. Hum. Mol. Genet. (1998) 7:1511–1516.[Abstract/Free Full Text]

  22. Wang J.C., Kwon J.M., Shah P., Morris J.C., Goate A. Effect of APOE genotype and promoter polymorphism on risk of Alzheimer's disease. Neurology (2000) 55:1644–1649.[Abstract/Free Full Text]

  23. Artiga M.J., Bullido M.J., Frank A., Sastre I., Recuero M., Garcia M.A., Lendon C.L., Han S.W., Morris J.C., Vazquez J., et al. Risk for Alzheimer's disease correlates with transcriptional activity of the APOE gene. Hum. Mol. Genet. (1998) 7:1887–1892.[Abstract/Free Full Text]

  24. Zurutuza L., Verpillat P., Raux G., Hannequin D., Puel M., Belliard S., Michon A., Pothin Y., Camuzat A., Penet C., et al. APOE promoter polymorphisms do not confer independent risk for Alzheimer's disease in a French population. Eur. J. Hum. Genet. (2000) 8:713–716.[CrossRef][Web of Science][Medline]

  25. Laws S.M., Hone E., Taddei K., Harper C., Dean B., McClean C., Masters C., Lautenschlager N., Gandy S.E., Martins R.N. Variation at the APOE -491 promoter locus is associated with altered brain levels of apolipoprotein E. Mol. Psychiatry (2002) 7:886–890.[CrossRef][Web of Science][Medline]

  26. Taddei K., Clarnette R., Gandy S.E., Martins R.N. Increased plasma apolipoprotein E (apoE) levels in Alzheimer's disease. Neurosci. Lett. (1997) 223:29–32.[CrossRef][Web of Science][Medline]

  27. Berg D.T., Calnek D.S., Grinnell B.W. The human apolipoprotein E gene is negatively regulated in human liver HepG2 cells by the transcription factor BEF-1. J. Biol. Chem. (1995) 270:15447–15450.[Abstract/Free Full Text]

  28. Chang D.J., Paik Y.K., Leren T.P., Walker D.W., Howlett G.J., Taylor J.M. Characterization of a human apolipoprotein E gene enhancer element and its associated protein factors. J. Biol. Chem. (1990) 265:9496–9504.[Abstract/Free Full Text]

  29. Do Carmo S., Seguin D., Milne R., Rassart E. Modulation of apolipoprotein D and apolipoprotein E mRNA expression by growth arrest and identification of key elements in the promoter. J. Biol. Chem. (2002) 277:5514–5523.[Abstract/Free Full Text]

  30. Paik Y.K., Chang D.J., Reardon C.A., Walker M.D., Taxman E., Taylor J.M. Identification and characterization of transcriptional regulatory regions associated with expression of the human apolipoprotein E gene. J. Biol. Chem. (1988) 263:13340–13349.[Abstract/Free Full Text]

  31. Salero E., Perez-Sen R., Aruga J., Gimenez C., Zafra F. Transcription factors Zic1 and Zic2 bind and transactivate the apolipoprotein E gene promoter. J. Biol. Chem. (2001) 276:1881–1888.[Abstract/Free Full Text]

  32. Smith J.D., Melian A., Leff T., Breslow J.L. Expression of the human apolipoprotein E gene is regulated by multiple positive and negative elements. J. Biol. Chem. (1988) 263:8300–8308.[Abstract/Free Full Text]

  33. Galetto R., Albajar M., Polanco J.I., Zakin M.M., Rodriguez-Rey J.C. Identification of a peroxisome-proliferator-activated-receptor response element in the apolipoprotein E gene control region. Biochem. J. (2001) 357:521–527.[CrossRef][Web of Science][Medline]

  34. Laffitte B.A., Repa J.J., Joseph S.B., Wilpitz D.C., Kast H.R., Mangelsdorf D.J., Tontonoz P. LXRs control lipid-inducible expression of the apolipoprotein E gene in macrophages and adipocytes. Proc. Natl Acad. Sci. USA (2001) 98:507–512.[Abstract/Free Full Text]

  35. Mazzone T., Gump H., Diller P., Getz G.S. Macrophage free cholesterol content regulates apolipoprotein E synthesis. J. Biol. Chem. (1987) 262:11657–11662.[Abstract/Free Full Text]

  36. Shih S.J., Allan C., Grehan S., Tse E., Moran C., Taylor J.M. Duplicated downstream enhancers control expression of the human apolipoprotein E gene in macrophages and adipose tissue. J. Biol. Chem. (2000) 275:31567–31572.[Abstract/Free Full Text]

  37. Martins C.A., Oulhaj A., de Jager C.A., Williams J.H. APOE alleles predict the rate of cognitive decline in Alzheimer disease: a nonlinear model. Neurology (2005) 65:1888–1893.[Abstract/Free Full Text]

  38. Chiappelli M., Borroni B., Archetti S., Calabrese E., Corsi M.M., Franceschi M., Padovani A., Licastro F. VEGF gene and phenotype relation with Alzheimer's disease and mild cognitive impairment. Rejuvenation Res. (2006) 9:485–493.[CrossRef][Web of Science][Medline]

  39. Belbin O., Dunn J.L., Chappell S., Ritchie A.E., Ling Y., Morgan L., Pritchard A., Warden D.R., Lendon C.L., Lehmann D.J., et al. A single nucleotide polymorphism in the alpha-l-anticchymotrypsin gene is associated with severe astrocytosis and more rapid cognitive decline in Alzheimer's disease patients. Neurobiol. Aging (2007) doi:10.1016/j.neurobiolaging.2007.02.021.

  40. Blomqvist M.E., Reynolds C., Katzov H., Feuk L., Andreasen N., Bogdanovic N., Blennow K., Brookes A.J., Prince J.A. Towards compendia of negative genetic association studies: an example for Alzheimer disease. Hum. Genet. (2006) 119:29–37.[CrossRef][Web of Science][Medline]

  41. Mirra S.S., Hart M.N., Terry R.D. Making the diagnosis of Alzheimer's disease. A primer for practicing pathologists. Arch. Pathol. Lab. Med. (1993) 117:132–144.[Web of Science][Medline]

  42. Nagy Z., Esiri M.M., Hindley N.J., Joachim C., Morris J.H., King E.M., McDonald B., Litchfield S., Barnetson L., Jobst K.A., et al. Accuracy of clinical operational diagnostic criteria for Alzheimer's disease in relation to different pathological diagnostic protocols. Dement. Geriatr. Cogn. Disord. (1998) 9:219–226.[CrossRef][Web of Science][Medline]

  43. Wenham P.R., Price W.H., Blandell G. Apoliporotein E genotyping by one-stage PCR. Lancet (1991) 337:1158–1159.[Web of Science][Medline]

  44. Schaid D.J., Rowland C.M., Tines D.E., Jacobson R.M., Poland G.A. Score tests for association between traits and haplotypes when linkage phase is ambiguous. Am J Hum Genet (2002) 70:425–434.[CrossRef][Web of Science][Medline]

  45. Roth M.H.F., Huppert F.A., Tym E., Mountjoy C.Q., Diffident-Brown A., Shoesmith D.J. CAMDEX: The Cambridge Examination for Mental Disorders of the Elderly (1988) Cambridge, UK: Cambridge University Press.

  46. Pinheiro J., Bates D.M. Mixed-effects Models in S and S-Plus (2000) New York, USA: Springer-Verlag New York Inc.


Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 NeurologyHome page
A. van der Walt, J. Stankovich, M. Bahlo, B. V. Taylor, I.A.F. van der Mei, S. J. Foote, T. J. Kilpatrick, J. P. Rubio, and H. Butzkueven
Apolipoprotein genotype does not influence MS severity, cognition, or brain atrophy
Neurology, September 29, 2009; 73(13): 1018 - 1025.
[Abstract] [Full Text] [PDF]

Home page
 Hum Mol GenetHome page
L. Xiong, H. Catoire, P. Dion, C. Gaspar, R. G. Lafreniere, S. L. Girard, A. Levchenko, J.-B. Riviere, L. Fiori, J. St-Onge, et al.
MEIS1 intronic risk haplotype associated with restless legs syndrome affects its mRNA and protein expression levels
Hum. Mol. Genet., March 15, 2009; 18(6): 1065 - 1074.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
16/18/2199    most recent
ddm171v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (6)
Right arrowRequest Permissions
Google Scholar
Right arrow Articles by Belbin, O.
Right arrow Articles by Morgan, K.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Belbin, O.
Right arrow Articles by Morgan, K.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?