Article Text
Abstract
Background: Sporadic Alzheimer’s disease (AD) is a common disabling disease of complex aetiology for which there are limited therapeutic options. We sought to investigate the role of the α7 nicotinic acetylcholine receptor gene (CHRNA7) in influencing risk of AD in a large population. CHRNA7 is a strong candidate gene for AD for several reasons: (1) its expression is altered differentially in the AD brain; (2) it interacts directly with β amyloid peptide (Aβ42); and (3) agonist activation induces several neuroprotective pathways.
Methods: In this study we used a genetic haplotype approach to assess the contribution of common variation at the CHRNA7 locus to risk of AD. Fourteen single nucleotide polymorphisms (SNPs) were genotyped in 764 AD patients and 314 controls.
Results: Three blocks of high linkage disequilibrium (LD) and low haplotype diversity were identified. The block 1 TCC haplotype was significantly associated with reduced odds of AD (p = 0.001) and was independent of apolipoprotein E (APOE) status. Individual SNPs were not associated with risk for AD.
Conclusions: We conclude that genetic variation in CHRNA7 influences susceptibility to AD. These results provide support for the development of α7nAChR agonists or modulators as potential drug treatments for AD. Further work is necessary to replicate the findings in other populations.
Statistics from Altmetric.com
Alzheimer’s disease (AD) is the most common form of dementia characterised by deficits in memory, attention and language leading to disabling functional decline. Behavioural and psychological symptoms are also common, affecting most individuals at some stage.1 Pathological features include extracellular senile plaques consisting of β amyloid (Aβ) peptide, neurofibrillary tangles composed of hyperphosphorylated τ protein, neuronal and synaptic loss, inflammation and oxidative stress. A selective reduction of cholinergic neurotransmission resulting from the loss of cholinergic neurons and neuronal nicotinic acetylcholine receptors (nAChRs) represents the key neurochemical change.2
Neuronal nicotinic acetylcholine receptors are a class of ligand-gated ion channels found throughout the nervous system and are implicated in a wide range of pathological states including AD as well as Parkinson’s disease, depression and schizophrenia.3 These receptors are formed using combinations of 12 potential subunits (α2–α10 and β2–β4) to produce a wide range of subtypes, each fulfilling different functions within the brain including roles in cognition, arousal, reward behaviour and central pain regulation.3 The most abundant nAChRs in the human brain are the α4β2 and α7 receptors.3
The α7 receptor subtype is homomeric, made up of five α7 subunits and is encoded by the CHRNA7 gene.4 CHRNA7 represents a strong candidate gene in AD association studies for several reasons. Reports have described decreases in α7nAChR expression in the hippocampus and temporal cortex of the AD brain.5 Studies using animal models have shown that administration of α7 agonists is beneficial to learning and memory, while antagonists produce the opposite effect.6 7 The α7nAChR is known to interact functionally with Aβ42,8 the consequences of which include blockade of the ligand binding site,9 down regulation of the ERK2/MAPK cascade leading to memory dysfunction,10 intraneuronal accumulation of Aβ4211 and τ phosphorylation.12 Conversely, the action of nicotine and other agonists on the α7nAChR activates neuroprotective pathways.13 Several α7nAChR agonists and modulators are currently under development for use in AD.14 One existing AD drug, galantamine, is both an acetylcholinesterase inhibitor and an allosteric modulator of α7nAChR.15 In contrast, memantine, an NMDA receptor antagonist, may act to inadvertently induce nicotinic receptor blockade.16
Sporadic late-onset AD is influenced by numerous postulated genetic and environmental risk factors.17 The apolipoprotein E (APOE) ∊4 allele is the only known genetic variant unequivocally associated with increased risk of late-onset AD.18 CHRNA7 maps to chromosome 15q13–q14, a region linked to several neuropsychiatric disorders.19 The gene is 138.5 kb long and contains 10 exons4 (fig 1). The human gene is partially duplicated (exons 5–10) and combined with five different exons to form a novel upstream gene termed dupCHRNA7.4 There are numerous intronic single nucleotide polymorphisms (SNPs) in CHRNA7 and three non-synonymous SNPs which are extremely rare. D15S1360, an intron 2 dinucleotide repeat polymorphism, has been genetically linked to the P50 sensory gating deficit found in most schizophrenics and 50% of their first degree relatives (LOD score 5.3).20 Two single SNP AD association studies of rs1514263 and an intron 3 polyT polymorphism were negative.21 22 D15S165, a 2 bp deletion in exon 6 of dupCHRNA7, was not associated with AD.23 In recent years there has been a shift from individual SNP association studies to the use of haplotypes to identify causal genetic variation underlying complex diseases. There are currently no studies using CHRNA7 haplotypes in association studies. The aim of the present study was to employ a genetic haplotype approach to assess the contribution of common variation at the CHRNA7 locus to risk of AD in a large case–control association study.
METHODS
Study subjects
All subjects were Caucasian with grandparents born in Northern Ireland. Patients were recruited from outpatient memory clinics and informed written consent for the study was obtained from patients or their main carers. The sample consisted of 764 patients with a diagnosis of probable AD according to the National Institute of Neurological and Communicative Disorders and Stroke and the Alzheimer’s Disease and Related Disorders Association criteria (NINCDS-ADRDA).24 The control group consisted of 314 individuals with cognitive scores exceeding 28/30 using the Mini-Mental State Examination (MMSE).25 Control subjects with a family history of dementia were excluded from the study. Ethical approval was obtained from the Research Ethics Committee, Queen’s University Belfast.
SNP selection
ABI SNP Browser version 3.0 was used to select highly polymorphic SNPs across CHRNA7.
Genotyping
Polymerase chain reaction-restriction fragment length polymorphism (PCR-RFLP)
Genomic DNA was extracted from peripheral blood leucocytes by the salting-out method.26 Standard polymerase chain reaction (PCR) was performed using 100 ng genomic DNA with final concentrations of 1X buffer, 0.2 mM deoxynucleotide triphosphates, 2.5 mM magnesium chloride, 0.17 μM of each primer (Invitrogen, Life Technologies, Strathclyde, UK) and 0.5 units of Taq polymerase (Biotools, B & M labs, Spain) in a 15 μl volume. PCR amplification was performed with an initial denaturation at 95°C for 10 min, followed by 35 cycles with denaturation at 95°C for 30 s, annealing temperatures ranging from 55–59°C for 55 s and extension at 72°C for 90 s. Final extension was at 72°C for 10 min. The amplified products were digested with 2 units of the appropriate restriction enzyme (New England Biolabs, Hertfordshire, UK) by overnight incubation at the optimum temperature for the enzyme. The digested products were then electrophoresed on an agarose gel containing ethidium bromide (Sigma, Poole, UK) and visualised by an ultraviolet trans-illuminator. Table 1 provides full details of the primer sequences, annealing temperature, restriction enzyme, and PCR and restriction product sizes for each SNP genotyped by PCR-RFLP. APOE genotyping was performed essentially as described.27
Taqman SNP genotyping assays
Nine SNPs were genotyped using Taqman SNP Genotyping Assays (Applied Biosystems, Warrington, UK) on an Applied Biosystems 7500 Real Time PCR System. Taqman probes were VIC and FAM dye labelled. PCR was performed using 10 ng genomic DNA with final concentrations of 1X SNP genotyping assay (Applied Biosystems) and 1X Absolute QPCR ROX Mix (AbGene, Surrey, UK) in a 12.5 μl volume. No template controls were included in each run. A pre-read run (60°C for 1 min) was performed, followed by amplification and a post-read run (60°C for 1 min). Amplification consisted of an initial denaturation at 95°C for 10 min, followed by 40 cycles of 95°C for 15 s and 60°C for 1 min. Taqman SNP Genotyping Assay IDs were C_1515718_10, C_1515697_10, C_1515680_10, C_1483000_10, C_1483022_10, C_1483027_10, C_8867074_10, C_1483045_10 and C_12077939_10.
Statistics
For each SNP, χ2 analysis was used to test whether genotype frequencies deviated from Hardy–Weinberg equilibrium and to compare genotype and allele frequencies between cases and controls. Haploview28 was used to calculate pairwise linkage disequilibrium (LD). A D’ threshold of 0.9 was used to generate haplotype blocks. χ2 analysis was used to compare haplotype frequencies between cases and controls. The level of statistical significance was set at p = 0.05. Significant results were corrected for multiple testing using ×1000 permutation tests. χ2 analysis was also used to test for the potential confounding effects of APOE status.
RESULTS
The study consisted of 13 intronic SNPs and 1 SNP within the 5’ untranslated region of CHRNA7. Fig 1 shows the structure of CHRNA7 and the location of selected SNPs.
The AD group were on average older than the controls at sampling (t test, p<0.001) but there was no significant difference in the percentage of females between groups (χ2, p = 0.609).
We found that the APOE ∊4 allele was significantly associated with AD (χ2 = 73.6, p<0.001) in this population, confirming previous reports and suggesting that APOE should be evaluated as a potential confounding factor.
We then analysed the genotype data for each of the CHRNA7 SNPs individually. Genotype frequencies for rs883473 and rs4779563 deviated from Hardy–Weinberg equilibrium in the AD group and were omitted from further analysis. Minor allele frequencies for the CHRNA7 SNPs ranged from 0.147–0.467. Individually, none of the SNPs showed an association with risk for AD (data not shown). For rs1827294 there was a trend towards an association between the C allele and AD risk as the p value approached 0.05 (p = 0.053).
Figure 2 shows the LD plot generated from the genotype data using Haploview.28 A D’ threshold of 0.9 was used to define LD blocks. Three blocks of 31, 7 and 22 kb were generated, each with three major haplotypes. Fig 3 shows the frequencies of these haplotypes in the whole sample and the extent of LD between blocks. The SNPs marked with arrows correspond to the haplotype tag SNPs which can be used to represent each block.
We then compared haplotype frequencies between cases and controls using χ2 analysis (table 2). We found a significant difference in the frequency of the block 1 TCC haplotype between cases and controls (odds ratio (OR) 0.48, 95% confidence interval (CI) 0.30 to 0.75; p = 0.001). The result remained significant following correction for multiple testing using ×1000 permutation tests (p = 0.008). This suggests that the TCC haplotype is associated with reduced odds of AD. The three SNPs in this haplotype were not associated with the presence of APOE ∊4 (χ2 p values = 0.843, 0.607 and 0.733, respectively), hence APOE status does not confound this significant result.
DISCUSSION
As far as we are aware, this is the first study to use CHRNA7 haplotypes in an association study for AD and the first to characterise the pattern of linkage disequilibrium in CHRNA7 in the Northern Irish population. The pattern of LD was fragmented but showed pronounced similarity to the structure determined by the HapMap project in the Centre d’Etude du Polymorphisme Humain (CEPH) population (www.hapmap.org). We identified three blocks of high LD and low haplotype diversity. Six tag SNPs were identified to represent the major haplotypes in each block for use in future studies.
Key points
CHRNA7, the gene coding for the α7 nicotinic acetylcholine receptor, is a strong candidate gene for Alzheimer’s disease (AD).
We carried out an association study, genotyping 14 SNPs in CHRNA7 in a large Northern Irish case–control sample.
We identified haplotype blocks based on D’ and found that block 1 was associated with a reduced risk of AD in this population (p = 0.001).
The study revealed a significant association between the block 1 TCC haplotype and reduced risk for AD versus controls (OR 0.48, 95% CI 0.30 to 0.75; p = 0.001) and which was independent of APOE status. Such genetic variation may represent an adaptive response to aging which serves to reduce the risk of developing AD. Although functionality of the SNPs remains unexplored, this haplotype spans promoter and upstream untranslated regions (fig 1). A causal SNP in the promoter region may affect gene expression, conceivably increasing α7nAChR expression and resulting in improved cholinergic neurotransmission or increased neuroprotection. A causal intronic SNP may modulate genetic risk by altering RNA splicing or stability.
The findings may further stimulate the ongoing efforts to improve the pharmacological treatment options in AD. Augmentation of α7nAChR function is a key pharmacological strategy with exciting potential benefits. For example, CHRNA7 gene therapy resulted in improved spatial memory in mice.29 Recently, MEM 3454, an α7nAChR agonist, entered a randomised, double blind, placebo controlled human phase 2a clinical trial.30
The strengths of the present study include the large sample size, the use of highly polymorphic SNPs and the definition of haplotype blocks by LD before haplotype analysis. We have avoided random definition of haplotypes which overlooks the nature of LD across the gene. The average SNP density of 1 SNP/8.6 kb was adequate, although a higher SNP density would improve the accuracy of block definitions.
In conclusion, the results presented here suggest that haplotypic variation in CHRNA7 alters susceptibility to late-onset AD. Further work is necessary to replicate the findings in other populations. The study also provides a framework for haplotype based studies of CHRNA7 in relation to other neuropsychiatric illnesses such as Parkinson’s disease, schizophrenia and bipolar disorder.
Acknowledgments
This work was supported by funding assistance provided by the Alzheimer’s Society, Research and Development Office (Northern Ireland), Ulster Garden Villages and Alzheimer’s Research Trust; Pharma Shire, Eisai and Pfizer provided grant support relevant to analytical costs and research fellow involvement.
REFERENCES
Footnotes
Competing interests: The authors (DC, APP, CWR) have received honoraria from companies with likely interests in drug effects involving cholinergic pathways.
Ethics approval: Ethical approval was obtained from the Research Ethics Committee, Queen’s University Belfast.
Patient consent: Informed written consent was obtained from patients or their main carers for publication of this report.