The cerebral deposition of Aβ42, a neurotoxic proteolytic derivate of amyloid precursor protein (APP), is a central event in Alzheimer’s disease (AD)(Amyloid hypothesis). Given the key role of APP-Aβ metabolism in AD pathogenesis, we selected 29 genes involved in APP processing, Aβ degradation and clearance. We then used exome and genome sequencing to investigate the single independent (single-variant association test) and cumulative (gene-based association test) effect of coding variants in these genes as potential susceptibility factors for AD, in a cohort composed of 332 sporadic and mainly late-onset AD cases and 676 elderly controls from North America and the UK. Our study shows that common coding variability in these genes does not play a major role for the disease development. In the single-variant association analysis, the main hits, none of which statistically significant after multiple testing correction (1.9e-4<p-value<0.05), were found to be rare coding variants (0.009%<MAF<1.4%) with moderate to strong effect size (1.84<OR<Inf) that map to genes mainly involved in Aβ extracellular degradation (TTR, ACE), clearance (LRP1) and APP trafficking and recycling (SORL1). These results were partially replicated in the gene-based analysis (c-alpha and SKAT tests), that reports ECE1, LYZ and TTR as nominally associated to AD (1.7e-3 <p-value <0.05). In concert with previous studies, we suggest that 1) common coding variability in APP-Aβ genes is not a critical factor for AD development and 2) Aβ degradation and clearance, rather than Aβ production, may play a key role in the etiology of sporadic AD.
Citation: Sassi C, Ridge PG, Nalls MA, Gibbs R, Ding J, Lupton MK, et al. (2016) Influence of Coding Variability in APP-Aβ Metabolism Genes in Sporadic Alzheimer’s Disease. PLoS ONE 11(6): e0150079. doi:10.1371/journal.pone.0150079
Editor: Koichi M. Iijima, National Center for Geriatrics and Gerontology, JAPAN
Received: July 6, 2015; Accepted: February 9, 2016; Published: June 1, 2016
This is an open access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Data Availability: All relevant data are within the paper and its Supporting Information files.
Funding: This study was supported by the Alzheimer's Research UK, the Medical Research Council (MRC), the Wellcome Trust/MRC Joint Call in Neurodegeneration Award (WT089698) to the UK Parkinson's Disease Consortium (whose members are from the University College London Institute of Neurology, the University of Sheffield, and the MRC Protein Phosphorylation Unit at the University of Dundee), grants (P50 AG016574, U01 AG006786, and R01 AG18023), the National Institute for Health Research Biomedical Research Unit in Dementia at University College London Hospitals, University College London; the Big Lottery (to Dr. Morgan); a fellowship from Alzheimer's Research UK (to Dr. Guerreiro); and the Intramural Research Programs of the National Institute on Aging and the National Institute of Neurological Disease and Stroke, National Institutes of Health (Department of Health and Human Services Project number, ZO1 AG000950-10). The MRC London Neurodegenerative Diseases Brain Bank and the Manchester Brain Bank from Brains for Dementia Research are jointly funded from ARUK and AS. Tissue samples were supplied by The London Neurodegenerative Diseases Brain Bank, which receives funding from the MRC and as part of the Brains for Dementia Research programme, jointly funded by Alzheimer’s Research UK and Alzheimer’s Society.
Competing interests: The authors have declared that no competing interests exist.
The cerebral deposition of Aβ42 aggregates, insoluble neurotoxic derived of amyloid precursor protein (APP), is likely caused by an imbalance between Aβ production and clearance and represents a key event in Alzheimer’s disease (Amyloid hypothesis) .
A growing body of evidence has pointed to the critical role of APP-Aβ metabolism in AD pathogenesis. First, the discovery of APP, PSEN1 and PSEN2 mutations showed that familial Alzheimer’s disease is linked to Aβ42 overproduction [2–4]. Second, genome-wide association studies (GWASs) identified several susceptibility loci associated with AD (APOE, BIN1, PICALM, CD33, ABCA7, CLU, MS4A6A, EPHA1, CR1, CD2AP, SORL1, CASS4) and regulating APP-Aβ levels [5–12]. Third, next generation sequencing laid the ground for the discovery of TREM2 risk variants, highlighting the possible role of microglia in Aβ clearance [13–14]. Finally, although recent studies have shown that rare coding variability in PSEN1 may influence the susceptibility for apparently sporadic late-onset AD (LOAD) [15–16], increases in Aβ production currently explain a minority of AD cases. By contrast, it is very likely that the majority of AD cases are caused by impaired degradation and clearance of Aβ, which is produced at normal levels throughout life [17–18]. Despite the importance of APP-Aβ metabolism in AD, the role of genes taking part in Aβ production and catabolism as susceptibility factors for AD is still elusive and has not been extensively investigated. Therefore, in this study, we selected 29 genes known to be involved in APP and Aβ processing: ADAM9, ADAM10, ADAM17, MEP1B, BACE1, BACE2, NCSTN, PSENEN, APH1B, LRRTM3, APLP1, APBA1, SORL1, TTR, GPR3, ECE1, ECE2, IDE, CST3, CTSB, CTSD, LYZ, MME, ACE, MMP3, A2M, PLAT, KLK6 and LRP1. We then analyzed the single, independent and the cumulative effect of protein coding variants in these genes from exome and genome sequencing data, in a cohort composed of 332 sporadic and mainly late-onset AD cases and 676 elderly controls from North America and UK.
Materials and Methods
We used exome and genome sequencing data to identify common, low frequency, and rare coding variants in 29 genes involved in: Aβ production (ADAM9, ADAM10, ADAM17, BACE1, BACE2, NCSTN, PSENEN, APH1B, MEP1B, LRRTM3, GPR3), APP stabilization (APLP1, APBA1), APP recycling (SORL1), Aβ deposition (TTR), intracellular degradation (ECE1, ECE2, IDE, CST3, CTSB, CTSD, LYZ, MME), extracellular degradation and clearance (ACE, MMP3, A2M, PLAT, KLK6, LRP1).
These genes were chosen on the basis of PubMed based literature search and/or based on predicted protein interactions using STRING (http://string.embl.de/).
The discovery cohort was composed of 332 apparently sporadic AD cases and 676 elderly controls, neuropathologically and clinically confirmed, originating from the UK and North America. The mean age at disease onset was 71.66 years (range 41–94 years) for cases and the mean age of ascertainment was 78.15 years (range 60–102 years) for controls (Table 1). The majority of the cases (77%) were late-onset (> 65 years at onset).
Among the cases and controls, 42% and 51% were female, respectively. 58% and 47% of the cases and controls carried the APOE ε4 allele, respectively. The APOE ε4 allele was significantly associated to the disease status in the NIH and ADNI series (p-value = 0.02 and 1.19x10-9, respectively). Importantly, all the BYU controls from the Cache County Study on Memory in Aging were heterozygous for APOE ɛ4 allele. However, given the fact that 1) they were elderly (mean age 80.8 years old [range: 75–94.59]) and without any clinical sign of dementia and 2) APOE ɛ4 allele is a risk but not a causative factor, we still considered them as controls and included in the study.
Written consent for participation was obtained in accordance with institutional review board standards. All samples had fully informed consent for retrieval and were authorized for ethically approved scientific investigation. The UCLH Research Ethics Committee number 10/H0716/3, BYU IRB, Cardiff REC for Wales 08/MRE09/38+5, REC Reference 04/Q2404/130, National Research Ethics Service (NRES) specifically approved this study.
We performed whole exome sequencing on a cohort of 332 sporadic and mainly late-onset AD cases and 477 elderly controls. DNA was extracted from blood or brain both for cases and controls using standard protocols. Library preparation for next generation sequencing used DNA (between 1 μg and 3 μg) fragmented in a Covaris E210 (Covaris Inc.). Following fragmentation, DNA was end-repaired by 5’phosphorylation, using the Klenow polymerase. A poly-adenine tail was added to the 3’end of the phosphorylated fragment and ligated to Illumina adapters. After purification using an AMPure DNA Purification kit (Beckman Coulter, Inc), adapter-ligated products were amplified. The DNA library was then hybridized to an exome capture library (NimbleGen SeqCap EZ Exome v2.0, Roche Nimblegen Inc. or TruSeq, Illumina Inc.) and precipitated using streptavidin-coated magnetic beads (Dynal Magnetic Beads, Invitrogen). Exome-enriched libraries were PCR-amplified, and then DNA hybridized to paired-end flow cells using a cBot (Illumina, Inc.) cluster generation system. Samples were sequenced on the Illumina HiSeq™ 2000 using 2x100 paired end reads cycles.
Whole Genome sequencing
Genome sequencing was performed in 199 elderly, clinically healthy controls, from the Cache County Study on Memory in Aging. DNA was extracted from blood using standard protocols. All samples were sequenced with the use of Illumina HiSeq technology. Alignment was performed with the use of CASAVA software and variant calling was performed with the use of SAMtools  and the Genome Analysis Toolkit GATK . This sequencing and variant calling were performed by our collaborators at Brigham Young University.
Sequence alignment and variant calling were performed against the reference human genome (UCSC hg19). Paired end sequence reads (2x100bp paired end read cycles) were aligned using the Burrows-Wheeler aligner . Format conversion and indexing were performed with Picard (www.picard.sourceforge.net/index.shtml). GATK was used to recalibrate base quality scores, perform local re-alignments around indels and to call and filter the variants . VCFtools was used to annotate gene information for the remaining novel variants. We used ANNOVAR software to annotate the variants . Variants were checked against established databases (1000 Genomes Project and dbSNP v.134). The protein coding effects of variants was predicted using SIFT, Polyphen2 and SeattleSeq Annotation (gvs.gs.washington.edu/SeattleSeqAnnotation). All variants within the coding regions of 29 candidate genes (A2M [NM_000014], ACE [NM_000789], ADAM9 [NM_003816], ADAM10 [NM_001110], ADAM17 [NM_003183], APBA1 [NM_001163], APH1B [NM_031301], APLP1 [NM_001024807], BACE1 [NM_012104], BACE2 [NM_012105], CST3 [NM_000099], CTSB [NM_001908], CTSD [NM_001909], ECE1 [NM_001397], ECE2 [NM_014693], GPR3 [NM_005281], IDE [NM_004969], LRP1 [NM_002332], KLK6 [NM_001012964], LRRTM3 [NM_178011], LYZ [NM_000239], MEP1B [NM_005925], MME [NM_000902], MMP3 [NM_002422], NCSTN [NM_015331], PLAT [NM_000930], PSENEN [NM_172341], SORL1 [NM_003105], TTR [NM_000371]) have been collected and analyzed. (S1 Table) (Further details are provided in the supplementary materials)
For each variant, allele frequencies were calculated in cases and controls and Fisher’s exact test on allelic association was performed. All computations were performed in R (version x64 3.0.2, http://www.r-project.org/). The threshold call rate for inclusion of both subjects and variants in the analysis was 95%.
In this study, we have sufficient power (≥80%) to detect common SNVs (MAF > 5%) with modest effect (OR = 2) through single-variant association analysis; however, we had limited power (<80%) to detect very rare SNVs (MAF < 0.1%), even those with strong effect (OR > 4).
A p-value of 0.05 was set as a nominal significance threshold. Based on simple Bonferroni correction for multiple testing, the thresholds for single variant and gene-based association are defined by p-value = 1.9e-4 (0.05/256 coding variants) and 1.7e-3 (0.05/29 genes), respectively.
For the gene-based analysis we have pooled together coding and non-coding variants with a MAF ≤0.05 and studied their cumulative effect on the AD trait.
C-alpha test and SKAT are closely related, being both non-burden tests, analyzing and collapsing the effect of genetic variants of different frequency (common and rare), effect (protective, damaging and neutral) and effect size (modest, moderate, strong). SKAT can be considered an expansion of the c-alpha test because overcomes some of its limits. Indeed, SKAT 1) can be applied also to the study of continuous traits; 2) does not need any permutation; 3) applies covariates to the study. In addition, we have used a set of 5–10 size matched genes not linked to APP-Aβ metabolism (based on a Pubmed and STRING search) as negative controls for the gene-based association analysis. Moreover, we have assessed the reliability of our results, comparing the total variant frequency of the selected genes in our study with the one reported for the European-American cohort in the Exome Variant Server (EVS)(http://evs.gs.washington.edu/EVS/).
The study population consisted of a total of 332 sporadic and mainly late-onset AD cases and 676 elderly controls of British and North American ancestry (Table 1).
We do not report any pathogenic mutation in APP, PSEN1 and PSEN2 in our cohort. However, one of the controls was an heterozygous carrier of the protective variant APP p.A673T (MAF 7x10e-4 in our cohort and MAF 5x10e-4 among the European non-finnish, ExAC database, released 13 January 2015).
We performed a single-variant and a single-gene association analysis in a pre-defined set of genes involved in APP processing (ADAM9, ADAM10, ADAM17, MEP1B, BACE1, BACE2, NCSTN, PSENEN, APH1B, APLP1, APBA1), Aβ metabolism and catabolism (LRRTM3, LRP1, TTR, GPR3, SORL1, ECE1, ECE2, IDE, CST3, CTSB, CTSD, LYZ, MME, ACE, MMP3, A2M, PLAT, KLK6), including 68 Megabase pairs (Mbs) of coding sequence.
A total of 960 single nucleotide variants (SNVs) has been identified. Among these, 256 (26.6%) were nonsynonymous, 194 (20.20%) were synonymous, 97 (10.1%) were intronic and 413 (43%) UTR variants. Among the missense variants, 192 (75%) were very rare (MAF<1%), 16 (6.25%) were low frequency (1%<MAF<5%) and 12(4.68%) were common (MAF>5%). In addition, we report 36 novel coding variants. Variant minor allele frequency and novel variants were based on ExAC database, released 13 January 2015, or dbSNP 137(S2 Table). The overall variant frequency in our cohort was in line with the variant frequency reported in the American-European cohort in the Exome Variant server database (S3 Table).
Moreover, 120 missense variants (46.8%) were described as damaging variants by at least 2 out of 3 in silico prediction softwares (SIFT, Polyphen and Mutation Taster). Importantly, genes involved in Aβ degradation and clearance harbor the highest relative frequency of rare coding and damaging variants (mean = 4.73 rare coding variants/Kbp of coding sequence and 3.16 damaging coding variants/Kbp of coding sequence, respectively). By contrast, genes taking part in APP processing and Aβ production, present the lowest relative frequency of rare coding and damaging variants (mean = 3.59 rare coding variants/Kbp of coding sequence and 1.5 damaging coding variants/Kbp of coding sequence, respectively), suggesting a higher degree of conservation of this last cluster of genes (S4 and S5 Tables).
Single coding variant association test
We identified 3 nominally significant variants, clustering in genes involved in Aβ catabolism (TTR [p.T139M], ACE [p.T916M]) and APP cleavage (APH1B [p.T27I]). Overall, the main hits (variants with the lowest p-values) mainly cluster to genes predominantly involved in Aβ degradation and clearance (6 out of 8 genes [75%]), with ACE, SORL1 and LRP1 harboring multiple variants, compared to APP processing (2 out of 8 genes [25%]). Moreover, most of these top genes are highly expressed in the brain (http://biogps.org/) and harbor very rare coding variants (0.009%<MAF<1.4%) with moderate to strong effect size (1.84<OR<Inf). (Table 2).
However, none of the coding variants detected in the studied genes reached the statistical significance, based on a corrected p-value (p-value <1.9e-4), in the single-variant association test.
For all these main variants, with the APH1B (p.T27I) and SORL1 (p.D2065V) exceptions, the minor allele was substantially more frequent in cases compared to controls, suggesting a possible role as a risk factor for AD. The study possessed relatively low power to detect any significant association between cases and controls for low frequency and rare variants. Nevertheless, we analyzed these variants because we could not preclude the possibility that high effect risk alleles were present.
Gene-based association test
In addition to single-marker analysis, we performed gene-wide analysis to combine the joint signal from multiple coding and non-coding variants with a MAF≤ 0.05 within a gene and to provide greater statistical power than that for single-marker tests. All the variants (nonsynonymous, synonymous, UTRs) located within the studied genes and their exon-intron flanking regions were collapsed together and their joint effect has been studied and compared with 5 to 10 size and variant matched gene controls (S6A and S6B Table). Genes involved in Aβ degradation and clearance were enriched for the lowest p-values. The combined effect of variants in ECE1, LYZ and in TTR reached the nominal significance in the c-alpha and SKAT tests, respectively (1.7e-3 <p-value <0.05) (Tables 3 and 4). There was a partial overlap between genes identified in the single-marker analysis and those with SKAT and c-alpha tests. Importantly, TTR was the main finding both in the SKAT and single-marker association analysis. Therefore, suggesting TTR as a promising potential candidate risk gene for AD.
The Amyloid cascade hypothesis is the main accepted hypothesis underlying AD pathology. Several genes within the APP-Aβ metabolism pathway have been reported as potential candidate genes for AD. However, coding variability among these has not been extensively investigated. The vast majority of reported studies are based on candidate gene approaches using array-based SNP genotyping and are focused mainly on genes involved in Aβ catabolism (http://www.alzgene.org/). Thus, leaving low frequency and rare coding variants and genes involved in Aβ production largely unexplored.
GWASs and chip-based candidate gene approaches have shown that common and generally non-coding variability within these genes does not play a critical role for AD development. The only exceptions to this general rule are represented by SORL1 and ABCA7, which have been reported associated with late-onset apparently sporadic and familial AD both with GWASs, candidate gene approaches and exome sequencing [25–28].
In this study, we report a screening of genes known to be involved in the APP-Aβ metabolism (APP processing, Aβ production, degradation and clearance). We applied single-marker and gene-based association analyses, to investigate the independent and joint effect of coding variability within these genes in a cohort composed of 332 apparently sporadic and mainly late-onset AD cases and 676 elderly controls from North America and the UK.
In our cohort, genes involved in Aβ degradation and clearance harbor the highest relative frequency of rare and predicted damaging variants (mean = 4.73 rare coding variants/Kbp of coding sequence and 3.16 damaging coding variants/Kbp of coding sequence, respectively). Conversely, genes encoding for proteins regulating Aβ production presented the lowest relative frequency of coding and likely damaging variability (mean = 3.59 rare coding variants/Kbp of coding sequence and 1.5 damaging coding variants/Kbp of coding sequence, respectively) (S4 and S5 Tables), suggesting a higher degree of conservation.
In the single-variant association analysis, the main hits were very rare coding variants (MAF 0.009%< MAF<1.4%) with strong effect sizes (1.84<OR<Inf) mapping to genes involved in Aβ extracellular degradation (ACE, TTR, CTSB, MMP3), APP trafficking and recycling (SORL1), and clearance (LRP1). Among the main single-marker hits only 2 genes (APH1B, NCSTN) were component of the γ secretase complex and therefore pivotal for APP processing. Our study was underpowered for the detection of rare and low frequency variants and these variants were nominally significant after Bonferroni correction (1.9e-4<p-value <0.05). Importantly, in our cohort, TREM2 p.R47H, the second most common risk factor for sporadic AD, has been detected in 6 cases (1.8%) and 4 controls (0.59%) and, given our small sample size, with a MAF = 0.2%, was not significantly associated to AD (p-value = 0.09). Therefore, we suggest that the main variants detected in our study may be functional and warrant a follow up in an extended sample size.
Transthyretin (TTR) is a 55-kDa protein, particularly abundant in the CSF and human plasma, where it transports thyroxin from the peripheral blood circulation to the brain. TTR has been already reported linked to AD dementia. Particularly, it has been hypothesized that TTR may act as a scaffold protein, binding to amyloid and preventing its deposition and aggregation in plaques . Moreover, TTR is a well established biomarker for AD: 1) a decrese in the CSF and serum TTR levels has been associated to an increased AD severity and faster rate of disease progression and 2) 5 SNPs and different TTR haplotypes have been related to hippocampal atrophy . By contrast, TTR overexpression decreases Aβ deposition, protects against Aβ plaques formation and improves the cognitive function in different AD mouse model strains [31–32]. Finally, TTR, likewise other well conserved secreted proteins critically involved in dementia (PRP and PGRN), has no homologous proteins (http://string-db.org/), therefore implying that even subtle changes to its epitopes and/or domains may be functionally relevant and phenotypically manifest. In line with this hypothesis, TTR is the main hit both in the single-variant analysis and, with only 6 variants detected (4 coding and 2 3´ UTR variants), in the gene-based analysis (SKAT). Thus, suggesting that either the joint effect of the coding and non-coding variability at the TTR locus is likely to be functional or that the signal is driven by TTR p.T139M, that has a very strong effect size (OR = 6.19, 95% CI = 1.099–63.091).
The second strongest signal in the single variant analysis maps to ACE (p.T916M), encoding angiotensin I converting enzyme 1 (ACE1), a zinc metalloprotease, which regulates blood pressure. Multiple lines of evidence have shown that ACE1 may ameliorate the cognitive decline either regulating the cerebral blood flow and/or converting Aβ42 to Aβ40, a more soluble isoform and its activity is increased in AD brain, in proportion to the parenchymal Aβ load [33–35].
SORL1 encodes for Sortilin related receptor (SORLA, also known as LR11), which binds to APOE and mediates several intracellular sorting and trafficking functions through a VPS10 (vacuolar protein sorting protein 10) domain . SORLA is highly expressed in the brain and modulates APP recycling through the retromer complex, thereby influencing levels of Aβ [37–38]. A growing body of evidence suggested SORL1 as an excellent positional and functional candidate for AD [12, 25, 39–42, 25–26]. Moreover, decreased expression and DNA methylation changes at the SORL1 locus have been reported associated to AD [43–44]. SORL1 harbors 3 of the top variants identified (p.E270K, p.D2065V and p.R11P). Importantly, p.E270K has been reported in 7 Carribean Hispanic families and in vitro studies showed that SORL1 p.E270K is a functional variant, leading to increased secretion of Aβ, when transfected in HEK293 cell lines and weakening the binding to APP . Therefore, it is plausible that SORL1 (p.E270K) may increase the susceptibility also for apparently sporadic LOAD. Thus, this finding, with the other main SORL1 variants detected (p.D2065V and p.R11P) should be further investigated.
Finally, four of the main hits map to the low-density lipoprotein receptor related protein 1 (LRP1) gene. Importantly, LRP1 is a major Aβ clearance receptor in cerebral vascular smooth muscle cells and disturbance of this pathway contributes to Aβ accumulation in the brain . In addition, LRP1 locus was identified as an AD candidate locus in consanguineous Israeli-Arab community .
The c-alpha and SKAT tests reported ECE1 and LYZ and TTR, respectively, as main genes associated with LOAD, although after the Bonferroni correction, the association was only nominally significant. Very interestingly, the main genes identified both with the single-marker and gene-based analysis play a pivotal role in the cardiovascular system and have been already signaled as potential risk factors for cerebral amyloid angiopathy (CAA) or vascular dementia . First, ECE1 and ACE are key components of the renin-angiotensin cascade, that controls blood pressure . Second, ECE1, ACE and TTR are mostly expressed by endothelial cells in the CNS (http://web.stanford.edu/group/barres_lab/brain_rnaseq.html). Third, acute and chronic hypoxia, through the release of the hypoxia inducible factor 1α (HIF-1α), exerts a critical epigenetic regulation on APP processing and Aβ catabolism key genes. Notably, HIF-1α has been reported to increase γ and β cleavage of APP and impair Aβ degradation and clearance mainly through the down regulation of pivotal proteins such as MME, ECE1 and TTR [50–53].
In summary, our study shows that 1) common coding variability within genes involved in APP-Aβ metabolism does not play a critical role for AD development; 2) genes regulating Aβ production are more conserved than genes playing a key role in Aβ degradation and clearance, thus less frequently involved in sporadic AD; 3) TTR, ACE, SORL1, CTSB and LRP1 harbor rare coding variants with strong effect size, likely to be functional and warrant further investigation in an extended cohort; 4) ECE1, LYZ and TTR play a critical role in the cardiovascular system and the joint effect of their variants may increase the susceptibility to AD. Finally, in concert with previous studies, our results support a potential overlapping biology with shared risk factors between CAA, vascular dementia and AD.
S1 Table. List of the 29 genes selected in our study.
S2 Table. Coding variants detected in APP-Aß metabolism genes in our study.
Aa, aminoacid; OR, odds ratio; CI, confidence interval.
S3 Table. Comparison between total variants and relative frequency of variants detected in our study and Exome Variant Server database (EVS) in the European-American cohort.
S4 Table. Relative frequency of rare variants in the 29 genes analyzed.Kbp, kilobasepair.
S5 Table. Relative frequency of predicted damaging variants in the 29 genes analyzed.
The variants have been predicted as damaging by at least 2 out 3 in silico prediction softwares (MUTATION TASTER, SIFT, POLYPHEN2). Kbp, kilobasepair.
A. Size-matched negative controls for the gene-based association analysis (c-alpha test). B. Size-matched negative controls for the gene-based association analysis (SKAT).
This study was supported by the Alzheimer's Research UK, the Medical Research Council (MRC), the Wellcome Trust/MRC Joint Call in Neurodegeneration Award (WT089698) to the UK Parkinson's Disease Consortium (whose members are from the University College London Institute of Neurology, the University of Sheffield, and the MRC Protein Phosphorylation Unit at the University of Dundee), grants (P50 AG016574, U01 AG006786, and R01 AG18023), the National Institute for Health Research Biomedical Research Unit in Dementia at University College London Hospitals, University College London; the Big Lottery (to Dr. Morgan); a fellowship from Alzheimer's Research UK (to Dr. Guerreiro); and the Intramural Research Programs of the National Institute on Aging and the National Institute of Neurological Disease and Stroke, National Institutes of Health (Department of Health and Human Services Project number, ZO1 AG000950-10). The MRC London Neurodegenerative Diseases Brain Bank and the Manchester Brain Bank from Brains for Dementia Research are jointly funded from ARUK and AS. The Alzheimer’s Research UK (ARUK) Consortium: Peter Passmore, David Craig, Janet Johnston, Bernadette McGuinness, Stephen Todd, Queen’s University Belfast, UK; Reinhard Heun, Royal Derby Hospital, UK; Heike Kölsch, University of Bonn, Germany; Patrick G. Kehoe, University of Bristol, UK; Nigel M. Hooper, University of Leeds, UK; Emma R.L.C. Vardy, University of Newcastle, UK; David M. Mann, Stuart Pickering-Brown, University of Manchester, UK; Kristelle Brown, James Lowe, Kevin Morgan, University of Nottingham, UK; A. David Smith, Gordon Wilcock, Donald Warden, University of Oxford (OPTIMA), UK; Clive Holmes, University of Southampton, UK.
Tissue samples were supplied by The London Neurodegenerative Diseases Brain Bank, which receives funding from the MRC and as part of the Brains for Dementia Research programme, jointly funded by Alzheimer’s Research UK and Alzheimer’s Society.
Conceived and designed the experiments: CS R. Guerreiro ABS JH. Performed the experiments: CS MKL R. Guerreiro KSB. Analyzed the data: CS PGR MAN R. Gibbs R. Guerreiro JD ABS JH. Contributed reagents/materials/analysis tools: ABS JH AMG CC JSK KM JFP. Wrote the paper: CS R. Guerreiro ABS JH. Samples collection: ARUK Consortium SA MKL KL CM KSB CC AMG R. Guerreiro JSK CT JL JB JT ABS JH.
- 1. Hardy J, Selkoe DJ. The amyloid hypothesis of Alzheimer’s disease: progress and problems on the road to therapeutics. Science. 2002 Jul 19;297(5580):353–6. pmid:12130773
- 2. Goate A, Chartier-Harlin MC, Mullan M, Brown J, Crawford F, Fidani L, et al. Segregation of a missense mutation in the amyloid precursor protein gene with familial Alzheimer’s disease. Nature. 1991 Feb 21;349(6311):704–6. pmid:1671712
- 3. Sherrington R, Rogaev EI, Liang Y, Rogaeva EA, Levesque G, Ikeda M, et al. Cloning of a gene bearing missense mutations in early-onset familial Alzheimer’s disease. Nature. 1995 Jun 29;375(6534):754–60. pmid:7596406
- 4. Rogaev EI, Sherrington R, Rogaeva EA, Levesque G, Ikeda M, Liang Y, et al. Familial Alzheimer’s disease in kindreds with missense mutations in a gene on chromosome 1 related to the Alzheimer’s disease type 3 gene. Nature. 1995 Aug 31;376(6543):775–8. pmid:7651536
- 5. Harold D, Abraham R, Hollingworth P, Sims R, Gerrish A, Hamshere ML, et al. Genome-wide association study identifies variants at CLU and PICALM associated with Alzheimer’s disease. Nat Genet. 2009 Oct;41(10):1088–93. doi: 10.1038/ng.440. pmid:19734902
- 6. Lambert J-C, Heath S, Even G, Campion D, Sleegers K, Hiltunen M, et al. Genome-wide association study identifies variants at CLU and CR1 associated with Alzheimer’s disease. Nat Genet. 2009 Oct;41(10):1094–9. doi: 10.1038/ng.439. pmid:19734903
- 7. Seshadri S, Fitzpatrick AL, Ikram MA, DeStefano AL, Gudnason V, Boada M, et al. Genome-wide analysis of genetic loci associated with Alzheimer disease. JAMA J Am Med Assoc. 2010 May 12;303(18):1832–40.
- 8. Hollingworth P, Harold D, Sims R, Gerrish A, Lambert J-C, Carrasquillo MM, et al. Common variants at ABCA7, MS4A6A/MS4A4E, EPHA1, CD33 and CD2AP are associated with Alzheimer’s disease. Nat Genet. 2011 May;43(5):429–35. doi: 10.1038/ng.803. pmid:21460840
- 9. Naj AC, Jun G, Beecham GW, Wang L-S, Vardarajan BN, Buros J, et al. Common variants at MS4A4/MS4A6E, CD2AP, CD33 and EPHA1 are associated with late-onset Alzheimer’s disease. Nat Genet. 2011 May;43(5):436–41. doi: 10.1038/ng.801. pmid:21460841
- 10. Kamboh MI, Demirci FY, Wang X, Minster RL, Carrasquillo MM, Pankratz VS, et al. Genome-wide association study of Alzheimer’s disease. Transl Psychiatry. 2012;2:e117. doi: 10.1038/tp.2012.45. pmid:22832961
- 11. Reitz C, Jun G, Naj A, Rajbhandary R, Vardarajan BN, Wang L-S, et al. Variants in the ATP-binding cassette transporter (ABCA7), apolipoprotein E ϵ4,and the risk of late-onset Alzheimer disease in African Americans. JAMA J Am Med Assoc. 2013 Apr 10;309(14):1483–92.
- 12. Lambert J-C, Ibrahim-Verbaas CA, Harold D, Naj AC, Sims R, Bellenguez C, et al. Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease. Nat Genet. 2013 Dec;45(12):1452–8. doi: 10.1038/ng.2802. pmid:24162737
- 13. Guerreiro R, Wojtas A, Bras J, Carrasquillo M, Rogaeva E, Majounie E, et al. TREM2 variants in Alzheimer’s disease. N Engl J Med. 2013 Jan 10;368(2):117–27. doi: 10.1056/NEJMoa1211851. pmid:23150934
- 14. Jonsson T, Stefansson H, Steinberg S, Jonsdottir I, Jonsson PV, Snaedal J, et al. Variant of TREM2 associated with the risk of Alzheimer’s disease. N Engl J Med. 2013 Jan 10;368(2):107–16. doi: 10.1056/NEJMoa1211103. pmid:23150908
- 15. Benitez BA, Karch CM, Cai Y, Jin SC, Cooper B, Carrell D, et al. The PSEN1, p.E318G variant increases the risk of Alzheimer’s disease in APOE-ε4 carriers. PLoS Genet. 2013;9(8):e1003685. doi: 10.1371/journal.pgen.1003685. pmid:23990795
- 16. Sassi C, Guerreiro R, Gibbs R, Ding J, Lupton MK, Troakes C, et al. Investigating the role of rare coding variability in Mendelian dementia genes (APP, PSEN1, PSEN2, GRN, MAPT, and PRNP) in late-onset Alzheimer’s disease. Neurobiol Aging. 2014 Jun 16;
- 17. Miners JS, Baig S, Palmer J, Palmer LE, Kehoe PG, Love S. Abeta-degrading enzymes in Alzheimer’s disease. Brain Pathol Zurich Switz. 2008 Apr;18(2):240–52.
- 18. Selkoe DJ. Clearing the brain’s amyloid cobwebs. Neuron. 2001 Oct 25;32(2):177–80. pmid:11683988
- 19. Ramasamy A, Trabzuni D, Guelfi S, Varghese V, Smith C, Walker R, et al. Genetic variability in the regulation of gene expression in ten regions of the human brain. Nat Neurosci. 2014 Oct;17(10):1418–28. doi: 10.1038/nn.3801. pmid:25174004
- 20. Gibbs JR, van der Brug MP, Hernandez DG, Traynor BJ, Nalls MA, Lai S-L, et al. Abundant quantitative trait loci exist for DNA methylation and gene expression in human brain. PLoS Genet. 2010 May;6(5):e1000952. doi: 10.1371/journal.pgen.1000952. pmid:20485568
- 21. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence Alignment/Map format and SAMtools. Bioinforma Oxf Engl. 2009 Aug 15;25(16):2078–9.
- 22. McKenna A, Hanna M, Banks E, Sivachenko A, Cibulskis K, Kernytsky A, et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010 Sep;20(9):1297–303. doi: 10.1101/gr.107524.110. pmid:20644199
- 23. Li H, Durbin R. Fast and accurate short read alignment with Burrows-Wheeler transform. Bioinforma Oxf Engl. 2009 Jul 15;25(14):1754–60.
- 24. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. 2010 Sep;38(16):e164. doi: 10.1093/nar/gkq603. pmid:20601685
- 25. Rogaeva E, Meng Y, Lee JH, Gu Y, Kawarai T, Zou F, et al. The neuronal sortilin-related receptor SORL1 is genetically associated with Alzheimer disease. Nat Genet. 2007 Feb;39(2):168–77. pmid:17220890
- 26. Pottier C, Hannequin D, Coutant S, Rovelet-Lecrux A, Wallon D, Rousseau S, et al. High frequency of potentially pathogenic SORL1 mutations in autosomal dominant early-onset Alzheimer disease. Mol Psychiatry. 2012 Sep;17(9):875–9. doi: 10.1038/mp.2012.15. pmid:22472873
- 27. Liu G, Li F, Zhang S, Jiang Y, Ma G, Shang H, et al. Analyzing large-scale samples confirms the association between the ABCA7 rs3764650 polymorphism and Alzheimer’s disease susceptibility. Mol Neurobiol. 2014 Dec;50(3):757–64. doi: 10.1007/s12035-014-8670-4. pmid:24643655
- 28. Steinberg S, Stefansson H, Jonsson T, Johannsdottir H, Ingason A, Helgason H, et al. Loss-of-function variants in ABCA7 confer risk of Alzheimer’s disease. Nat Genet. 2015 Mar 25;
- 29. Costa R, Gonçalves A, Saraiva MJ, Cardoso I. Transthyretin binding to A-Beta peptide—impact on A-Beta fibrillogenesis and toxicity. FEBS Lett. 2008 Mar 19;582(6):936–42. doi: 10.1016/j.febslet.2008.02.034. pmid:18295603
- 30. Cuenco KT, Friedland R, Baldwin CT, Guo J, Vardarajan B, Lunetta KL, et al. Association of TTR polymorphisms with hippocampal atrophy in Alzheimer disease families. Neurobiol Aging. 2011 Feb;32(2):249–56. doi: 10.1016/j.neurobiolaging.2009.02.014. pmid:19328595
- 31. Choi SH, Leight SN, Lee VM-Y, Li T, Wong PC, Johnson JA, et al. Accelerated Abeta deposition in APPswe/PS1deltaE9 mice with hemizygous deletions of TTR (transthyretin). J Neurosci Off J Soc Neurosci. 2007 Jun 27;27(26):7006–10.
- 32. Wati H, Kawarabayashi T, Matsubara E, Kasai A, Hirasawa T, Kubota T, et al. Transthyretin accelerates vascular Abeta deposition in a mouse model of Alzheimer’s disease. Brain Pathol Zurich Switz. 2009 Jan;19(1):48–57.
- 33. Zou K, Maeda T, Watanabe A, Liu J, Liu S, Oba R, et al. Abeta42-to-Abeta40- and angiotensin-converting activities in different domains of angiotensin-converting enzyme. J Biol Chem. 2009 Nov 13;284(46):31914–20. doi: 10.1074/jbc.M109.011437. pmid:19773553
- 34. Hu J, Igarashi A, Kamata M, Nakagawa H. Angiotensin-converting enzyme degrades Alzheimer amyloid beta-peptide (A beta); retards A beta aggregation, deposition, fibril formation; and inhibits cytotoxicity. J Biol Chem. 2001 Dec 21;276(51):47863–8. pmid:11604391
- 35. Hemming ML, Selkoe DJ. Amyloid beta-protein is degraded by cellular angiotensin-converting enzyme (ACE) and elevated by an ACE inhibitor. J Biol Chem. 2005 Nov 11;280(45):37644–50. pmid:16154999
- 36. Yamazaki H, Bujo H, Kusunoki J, Seimiya K, Kanaki T, Morisaki N, et al. Elements of neural adhesion molecules and a yeast vacuolar protein sorting receptor are present in a novel mammalian low density lipoprotein receptor family member. J Biol Chem. 1996 Oct 4;271(40):24761–8. pmid:8798746
- 37. Andersen OM, Reiche J, Schmidt V, Gotthardt M, Spoelgen R, Behlke J, et al. Neuronal sorting protein-related receptor sorLA/LR11 regulates processing of the amyloid precursor protein. Proc Natl Acad Sci U S A. 2005 Sep 20;102(38):13461–6. pmid:16174740
- 38. Offe K, Dodson SE, Shoemaker JT, Fritz JJ, Gearing M, Levey AI, et al. The lipoprotein receptor LR11 regulates amyloid beta production and amyloid precursor protein traffic in endosomal compartments. J Neurosci Off J Soc Neurosci. 2006 Feb 1;26(5):1596–603.
- 39. Feulner TM, Laws SM, Friedrich P, Wagenpfeil S, Wurst SHR, Riehle C, et al. Examination of the current top candidate genes for AD in a genome-wide association study. Mol Psychiatry. 2010 Jul;15(7):756–66. doi: 10.1038/mp.2008.141. pmid:19125160
- 40. Miyashita A, Koike A, Jun G, Wang L-S, Takahashi S, Matsubara E, et al. SORL1 is genetically associated with late-onset Alzheimer’s disease in Japanese, Koreans and Caucasians. PloS One. 2013;8(4):e58618. doi: 10.1371/journal.pone.0058618. pmid:23565137
- 41. Tan EK, Lee J, Chen CP, Teo YY, Zhao Y, Lee WL. SORL1 haplotypes modulate risk of Alzheimer’s disease in Chinese. Neurobiol Aging. 2009 Jul;30(7):1048–51. pmid:18063222
- 42. Lee JH, Cheng R, Schupf N, Manly J, Lantigua R, Stern Y, et al. The association between genetic variants in SORL1 and Alzheimer disease in an urban, multiethnic, community-based cohort. Arch Neurol. 2007 Apr;64(4):501–6. pmid:17420311
- 43. Scherzer CR, Offe K, Gearing M, Rees HD, Fang G, Heilman CJ, et al. Loss of apolipoprotein E receptor LR11 in Alzheimer disease. Arch Neurol. 2004 Aug;61(8):1200–5. pmid:15313836
- 44. Yu L, Chibnik LB, Srivastava GP, Pochet N, Yang J, Xu J, et al. Association of Brain DNA methylation in SORL1, ABCA7, HLA-DRB5, SLC24A4, and BIN1 with pathological diagnosis of Alzheimer disease. JAMA Neurol. 2015 Jan;72(1):15–24. doi: 10.1001/jamaneurol.2014.3049. pmid:25365775
- 45. Vardarajan BN, Zhang Y, Lee JH, Cheng R, Bohm C, Ghani M, et al. Coding mutations in SORL1 and Alzheimer disease. Ann Neurol. 2015 Feb;77(2):215–27. doi: 10.1002/ana.24305. pmid:25382023
- 46. Kanekiyo T, Liu C-C, Shinohara M, Li J, Bu G. LRP1 in brain vascular smooth muscle cells mediates local clearance of Alzheimer’s amyloid-β. J Neurosci Off J Soc Neurosci. 2012 Nov 14;32(46):16458–65.
- 47. Farrer LA, Bowirrat A, Friedland RP, Waraska K, Korczyn AD, Baldwin CT. Identification of multiple loci for Alzheimer disease in a consanguineous Israeli-Arab community. Hum Mol Genet. 2003 Feb 15;12(4):415–22. pmid:12566388
- 48. Revesz T, Holton JL, Lashley T, Plant G, Frangione B, Rostagno A, et al. Genetics and molecular pathogenesis of sporadic and hereditary cerebral amyloid angiopathies. Acta Neuropathol (Berl). 2009 Jul;118(1):115–30.
- 49. Rossi GP, Sacchetto A, Cesari M, Pessina AC. Interactions between endothelin-1 and the renin-angiotensin-aldosterone system. Cardiovasc Res. 1999 Aug 1;43(2):300–7. pmid:10536660
- 50. Nalivaevaa NN, Fisk L, Kochkina EG, Plesneva SA, Zhuravin IA, Babusikova E, et al. Effect of hypoxia/ischemia and hypoxic preconditioning/reperfusion on expression of some amyloid-degrading enzymes. Ann N Y Acad Sci. 2004 Dec;1035:21–33. pmid:15681798
- 51. Nalivaeva NN, Belyaev ND, Kerridge C, Turner AJ. Amyloid-clearing proteins and their epigenetic regulation as a therapeutic target in Alzheimer’s disease. Front Aging Neurosci. 2014;6:235. doi: 10.3389/fnagi.2014.00235. pmid:25278875
- 52. Li L, Zhang X, Yang D, Luo G, Chen S, Le W. Hypoxia increases Abeta generation by altering beta- and gamma-cleavage of APP. Neurobiol Aging. 2009 Jul;30(7):1091–8. pmid:18063223
- 53. Wang Z, Yang D, Zhang X, Li T, Li J, Tang Y, et al. Hypoxia-induced down-regulation of neprilysin by histone modification in mouse primary cortical and hippocampal neurons. PloS One. 2011;6(4):e19229. doi: 10.1371/journal.pone.0019229. pmid:21559427