Association Between Genetic Traits for Immune-Mediated Diseases and Alzheimer Disease | Allergy and Clinical Immunology | JAMA Neurology | JAMA Network
[Skip to Navigation]
Sign In
Figure 1.  Single-Nucleotide Polymorphism Enrichment for Alzheimer Disease (AD) Across Different Levels of Significance
Single-Nucleotide Polymorphism Enrichment for Alzheimer Disease (AD) Across Different Levels of Significance

Conditional quantile-quantile (Q-Q) plots of nominal −log10(P) vs empirical −log10(q) (corrected for inflation) in AD below the standard genome-wide association study threshold of P < 5 × 10−8 as a function of significance of association with Crohn disease (CD) (A), ulcerative colitis (UC) (B), type 1 diabetes (T1D) (C), rheumatoid arthritis (RA) (D), celiac disease (CeD) (E), and psoriasis (PSOR) (F) at the level of −log10(P) ≥ 0.0, −log10(P) ≥ 1.0, −log10(P) ≥ 2.0, and −log10(P) ≥ 3.0 corresponding to P ≤ 1, P ≤ .1, P ≤ .01, and P ≤ .001, respectively. Dashed line indicates all single-nucleotide polymorphisms.

Figure 2.  Shared Genetic Risk Between Alzheimer Disease (AD) and Immune-Mediated Diseases
Shared Genetic Risk Between Alzheimer Disease (AD) and Immune-Mediated Diseases

Conjunction Manhattan plot of conjunctional –log10(false discovery rate P value) for AD given Crohn disease (CD), ulcerative colitis (UC), type 1 diabetes (T1D), rheumatoid arthritis (RA), celiac disease (CeD), and psoriasis (PSOR). Single-nucleotide polymorphisms with conjunctional –log10(false discovery rate P value) > 1.3 (ie, false discovery rate P < .05) are shown with large points. A black line around the large points indicates the most significant single-nucleotide polymorphism in each linkage disequilibrium block; these single-nucleotide polymorphisms were each annotated with the closest gene, which is listed with the symbol in each locus. For additional details, see eAppendix 2 in the Supplement.

Table 1.  Summary Data From All Genome-wide Association Studies Used in the Current Study
Summary Data From All Genome-wide Association Studies Used in the Current Study
Table 2.  Top SNPs Associated With AD and Immune-Mediated Diseases
Top SNPs Associated With AD and Immune-Mediated Diseases
1.
Brookmeyer  R, Johnson  E, Ziegler-Graham  K, Arrighi  HM.  Forecasting the global burden of Alzheimer’s disease.  Alzheimers Dement. 2007;3(3):186-191.PubMedGoogle ScholarCrossref
2.
Eikelenboom  P, Hoozemans  JJ, Veerhuis  R, van Exel  E, Rozemuller  AJ, van Gool  WA.  Whether, when and how chronic inflammation increases the risk of developing late-onset Alzheimer’s disease.  Alzheimers Res Ther. 2012;4(3):15.PubMedGoogle ScholarCrossref
3.
Schmidt  R, Schmidt  H, Curb  JD, Masaki  K, White  LR, Launer  LJ.  Early inflammation and dementia: a 25-year follow-up of the Honolulu-Asia Aging Study.  Ann Neurol. 2002;52(2):168-174.PubMedGoogle ScholarCrossref
4.
Koyama  A, O’Brien  J, Weuve  J, Blacker  D, Metti  AL, Yaffe  K.  The role of peripheral inflammatory markers in dementia and Alzheimer’s disease: a meta-analysis.  J Gerontol A Biol Sci Med Sci. 2013;68(4):433-440.PubMedGoogle ScholarCrossref
5.
Jonsson  T, Stefansson  H, Steinberg  S,  et al.  Variant of TREM2 associated with the risk of Alzheimer’s disease.  N Engl J Med. 2013;368(2):107-116.PubMedGoogle ScholarCrossref
6.
Guerreiro  R, Wojtas  A, Bras  J,  et al; Alzheimer Genetic Analysis Group.  TREM2 variants in Alzheimer’s disease.  N Engl J Med. 2013;368(2):117-127.PubMedGoogle ScholarCrossref
7.
Bradshaw  EM, Chibnik  LB, Keenan  BT,  et al; Alzheimer Disease Neuroimaging Initiative.  CD33 Alzheimer’s disease locus: altered monocyte function and amyloid biology.  Nat Neurosci. 2013;16(7):848-850.PubMedGoogle ScholarCrossref
8.
Desikan  RS, Schork  AJ, Wang  Y,  et al; Inflammation Working Group and International Genomics of Alzheimer’s Disease Project (IGAP) and DemGene Investigators.  Polygenic overlap between C-reactive protein, plasma lipids, and Alzheimer disease.  Circulation. 2015;131(23):2061-2069.PubMedGoogle ScholarCrossref
9.
Karch  CM, Cruchaga  C, Goate  AM.  Alzheimer’s disease genetics: from the bench to the clinic.  Neuron. 2014;83(1):11-26.PubMedGoogle ScholarCrossref
10.
Andreassen  OA, Djurovic  S, Thompson  WK,  et al; International Consortium for Blood Pressure GWAS; Diabetes Genetics Replication and Meta-analysis Consortium; Psychiatric Genomics Consortium Schizophrenia Working Group.  Improved detection of common variants associated with schizophrenia by leveraging pleiotropy with cardiovascular-disease risk factors.  Am J Hum Genet. 2013;92(2):197-209.PubMedGoogle ScholarCrossref
11.
Andreassen  OA, Thompson  WK, Schork  AJ,  et al; Psychiatric Genomics Consortium; Bipolar Disorder and Schizophrenia Working Groups.  Improved detection of common variants associated with schizophrenia and bipolar disorder using pleiotropy-informed conditional false discovery rate.  PLoS Genet. 2013;9(4):e1003455.PubMedGoogle ScholarCrossref
12.
Lambert  JC, Ibrahim-Verbaas  CA, Harold  D,  et al; European Alzheimer’s Disease Initiative; Genetic and Environmental Risk in Alzheimer’s Disease; Alzheimer’s Disease Genetic Consortium; Cohorts for Heart and Aging Research in Genomic Epidemiology.  Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease.  Nat Genet. 2013;45(12):1452-1458.PubMedGoogle ScholarCrossref
13.
Franke  A, McGovern  DPB, Barrett  JC,  et al.  Genome-wide meta-analysis increases to 71 the number of confirmed Crohn’s disease susceptibility loci.  Nat Genet. 2010;42(12):1118-1125.PubMedGoogle ScholarCrossref
14.
Stahl  EA, Raychaudhuri  S, Remmers  EF,  et al; BIRAC Consortium; YEAR Consortium.  Genome-wide association study meta-analysis identifies seven new rheumatoid arthritis risk loci.  Nat Genet. 2010;42(6):508-514.PubMedGoogle ScholarCrossref
15.
Barrett  JC, Clayton  DG, Concannon  P,  et al; Type 1 Diabetes Genetics Consortium.  Genome-wide association study and meta-analysis find that over 40 loci affect risk of type 1 diabetes.  Nat Genet. 2009;41(6):703-707.PubMedGoogle ScholarCrossref
16.
Dubois  PCA, Trynka  G, Franke  L,  et al.  Multiple common variants for celiac disease influencing immune gene expression.  Nat Genet. 2010;42(4):295-302.PubMedGoogle ScholarCrossref
17.
Ellinghaus  D, Ellinghaus  E, Nair  RP,  et al.  Combined analysis of genome-wide association studies for Crohn disease and psoriasis identifies seven shared susceptibility loci.  Am J Hum Genet. 2012;90(4):636-647.PubMedGoogle ScholarCrossref
18.
Anderson  CA, Boucher  G, Lees  CW,  et al.  Meta-analysis identifies 29 additional ulcerative colitis risk loci, increasing the number of confirmed associations to 47.  Nat Genet. 2011;43(3):246-252.PubMedGoogle ScholarCrossref
19.
Zhang  B, Gaiteri  C, Bodea  L-G,  et al.  Integrated systems approach identifies genetic nodes and networks in late-onset Alzheimer’s disease.  Cell. 2013;153(3):707-720.PubMedGoogle ScholarCrossref
20.
Beecham  GW, Hamilton  K, Naj  AC,  et al; Alzheimer’s Disease Genetics Consortium.  Genome-wide association meta-analysis of neuropathologic features of Alzheimer’s disease and related dementias.  PLoS Genet. 2014;10(9):e1004606.PubMedGoogle ScholarCrossref
21.
Andreassen  OA, Zuber  V, Thompson  WK,  et al; PRACTICAL Consortium; CRUK GWAS.  Shared common variants in prostate cancer and blood lipids.  Int J Epidemiol. 2014;43(4):1205-1214.PubMedGoogle ScholarCrossref
22.
Andreassen  OA, McEvoy  LK, Thompson  WK,  et al; International Consortium for Blood Pressure Genome-Wide Association Studies, Genetic Factors for Osteoporosis Consortium.  Identifying common genetic variants in blood pressure due to polygenic pleiotropy with associated phenotypes.  Hypertension. 2014;63(4):819-826.PubMedGoogle ScholarCrossref
23.
Andreassen  OA, Desikan  RS, Wang  Y,  et al.  Abundant genetic overlap between blood lipids and immune-mediated diseases indicates shared molecular genetic mechanisms.  PLoS One. 2015;10(4):e0123057.PubMedGoogle ScholarCrossref
24.
Nichols  T, Brett  M, Andersson  J, Wager  T, Poline  JB.  Valid conjunction inference with the minimum statistic.  Neuroimage. 2005;25(3):653-660.PubMedGoogle ScholarCrossref
25.
Kim  S, Kim  SF, Maag  D,  et al.  Amino acid signaling to mTOR mediated by inositol polyphosphate multikinase.  Cell Metab. 2011;13(2):215-221.PubMedGoogle ScholarCrossref
26.
Xu  R, Paul  BD, Smith  DR,  et al.  Inositol polyphosphate multikinase is a transcriptional coactivator required for immediate early gene induction.  Proc Natl Acad Sci U S A. 2013;110(40):16181-16186.PubMedGoogle ScholarCrossref
27.
Xu  R, Sen  N, Paul  BD,  et al.  Inositol polyphosphate multikinase is a coactivator of p53-mediated transcription and cell death.  Sci Signal. 2013;6(269):ra22.PubMedGoogle Scholar
28.
Wickramasinghe  VO, Savill  JM, Chavali  S,  et al.  Human inositol polyphosphate multikinase regulates transcript-selective nuclear mRNA export to preserve genome integrity.  Mol Cell. 2013;51(6):737-750.PubMedGoogle ScholarCrossref
29.
Ahmed  I, Sbodio  JI, Harraz  MM,  et al.  Huntington’s disease: neural dysfunction linked to inositol polyphosphate multikinase.  Proc Natl Acad Sci U S A. 2015;112(31):9751-9756.PubMedGoogle ScholarCrossref
30.
Jolles  J, Bothmer  J, Markerink  M, Ravid  R.  Phosphatidylinositol kinase is reduced in Alzheimer’s disease.  J Neurochem. 1992;58(6):2326-2329.PubMedGoogle ScholarCrossref
31.
Heppner  FL, Ransohoff  RM, Becher  B.  Immune attack: the role of inflammation in Alzheimer disease.  Nat Rev Neurosci. 2015;16(6):358-372.PubMedGoogle ScholarCrossref
Original Investigation
June 2016

Association Between Genetic Traits for Immune-Mediated Diseases and Alzheimer Disease

Author Affiliations
  • 1Department of Neurology, University of California, San Francisco
  • 2Department of Psychiatry, Washington University, St Louis, Missouri
  • 3Department of Neurosciences, University of California, San Diego, La Jolla
  • 4Department of Cognitive Sciences, University of California, San Diego, La Jolla
  • 5Department of Psychiatry, University of California, San Diego, La Jolla
  • 6Department of Radiology, University of California, San Diego, La Jolla
  • 7Norwegian Centre for Mental Disorders Research, Institute of Clinical Medicine, University of Oslo, Oslo, Norway
  • 8Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway
  • 9Institute of Clinical Molecular Biology, Christian-Albrechts-University of Kiel, Kiel, Germany
  • 10Neuroradiology Section, Department of Radiology and Biomedical Imaging, University of California, San Francisco
  • 11Department of Neurology, Massachusetts General Hospital, Boston
  • 12Department of Pathology and Laboratory Medicine, University of Pennsylvania Perelman School of Medicine, Philadelphia
  • 13Norwegian PSC Research Center and KG Jebsen Inflammation Research Centre, Research Institute of Internal Medicine, Division of Cancer Medicine, Surgery and Transplantation, Oslo University Hospital Rikshospitalet, Oslo, Norway
  • 14Division of Gastroenterology, Institute of Medicine, University of Bergen, Bergen, Norway
JAMA Neurol. 2016;73(6):691-697. doi:10.1001/jamaneurol.2016.0150
Abstract

Importance  Late-onset Alzheimer disease (AD), the most common form of dementia, places a large burden on families and society. Although epidemiological and clinical evidence suggests a relationship between inflammation and AD, their relationship is not well understood and could have implications for treatment and prevention strategies.

Objective  To determine whether a subset of genes involved with increased risk of inflammation are also associated with increased risk for AD.

Design, Setting, and Participants  In a genetic epidemiology study conducted in July 2015, we systematically investigated genetic overlap between AD (International Genomics of Alzheimer’s Project stage 1) and Crohn disease, ulcerative colitis, rheumatoid arthritis, type 1 diabetes, celiac disease, and psoriasis using summary data from genome-wide association studies at multiple academic clinical research centers. P values and odds ratios from genome-wide association studies of more than 100 000 individuals were from previous comparisons of patients vs respective control cohorts. Diagnosis for each disorder was previously established for the parent study using consensus criteria.

Main Outcomes and Measures  The primary outcome was the pleiotropic (conjunction) false discovery rate P value. Follow-up for candidate variants included neuritic plaque and neurofibrillary tangle pathology; longitudinal Alzheimer’s Disease Assessment Scale cognitive subscale scores as a measure of cognitive dysfunction (Alzheimer’s Disease Neuroimaging Initiative); and gene expression in AD vs control brains (Gene Expression Omnibus data).

Results  Eight single-nucleotide polymorphisms (false discovery rate P < .05) were associated with both AD and immune-mediated diseases. Of these, rs2516049 (closest gene HLA-DRB5; conjunction false discovery rate P = .04 for AD and psoriasis, 5.37 × 10−5 for AD, and 6.03 × 10−15 for psoriasis) and rs12570088 (closest gene IPMK; conjunction false discovery rate P = .009 for AD and Crohn disease, P = 5.73 × 10−6 for AD, and 6.57 × 10−5 for Crohn disease) demonstrated the same direction of allelic effect between AD and the immune-mediated diseases. Both rs2516049 and rs12570088 were significantly associated with neurofibrillary tangle pathology (P = .01352 and .03151, respectively); rs2516049 additionally correlated with longitudinal decline on Alzheimer’s Disease Assessment Scale cognitive subscale scores (β [SE], 0.405 [0.190]; P = .03). Regarding gene expression, HLA-DRA and IPMK transcript expression was significantly altered in AD brains compared with control brains (HLA-DRA: β [SE], 0.155 [0.024]; P = 1.97 × 10−10; IPMK: β [SE], −0.096 [0.013]; P = 7.57 × 10−13).

Conclusions and Relevance  Our findings demonstrate genetic overlap between AD and immune-mediated diseases and suggest that immune system processes influence AD pathogenesis and progression.

Introduction

Alzheimer disease (AD) is the most common neurodegenerative disease, with an estimated prevalence of 30 million people worldwide, a number that is expected to quadruple in the next 40 years.1 Currently, there is no effective treatment that delays onset or slows progression of AD. With the aging of the US population and high costs associated with caring for cognitively impaired elderly individuals, identifying strategies that prevent AD is of utmost importance.

Inflammation, a core feature of many immune-mediated diseases, is being increasingly recognized as an important etiologic characteristic of AD. Complement factors and activated microglia are established histopathologic features in brains of patients with AD,2 and observational studies and meta-analyses suggest that elevated C-reactive protein levels are associated with increased risk for developing AD.3,4 Importantly, recent associations between AD and genetic variants encoding triggering receptor expressed on myeloid cells 2 (TREM2)5,6 and myeloid cell surface antigen CD337 suggest that immune system–associated mechanisms may contribute to and drive AD pathogenesis.

Recent work8,9 indicates that common genetic variants associated with inflammation may also increase the risk for developing AD. Herein, building on our prior work and applying a recently validated approach for investigating polygenic pleiotropy,10,11 we systematically investigated genetic overlap between AD and immune-mediated diseases. Taking advantage of several large genome-wide association studies (GWASs),12-17 we sought to identify single-nucleotide polymorphisms (SNPs) jointly associated with AD and 1 or more of the immune-mediated traits including Crohn disease (CD), ulcerative colitis (UC), rheumatoid arthritis (RA), type 1 diabetes (T1D), celiac disease (CeD), and psoriasis (PSOR).

Box Section Ref ID

Key Points

  • Question Is there genetic overlap between Alzheimer disease and immune-mediated diseases that promote inflammation?

  • Findings In a genetic epidemiology study using summary data from multiple genome-wide association studies of more than 100 000 people and validated methods to assess genetic pleiotropy, we identified 8 variants that were statistically significantly associated with both AD and immune-mediated disease. Two of these, representing the HLA locus and IPMK, demonstrated the same direction of allelic effect.

  • Meaning The findings demonstrate genetic overlap between Alzheimer disease and immune-mediated diseases and suggest that immune system processes influence Alzheimer disease pathogenesis and progression.

Methods
Participant Samples

We evaluated complete GWAS results in the form of summary statistics (P values and odds ratios) for clinically diagnosed AD12 and immune-mediated diseases including CD,13 UC,18 RA,14 T1D,15 CeD,16 and PSOR17 (Table 1). We obtained publicly available AD GWAS summary statistics data from the International Genomics of Alzheimer’s Disease Project (IGAP stages 1 and 2; for additional details, see eAppendix 2 in the Supplement and the article by Lambert et al12). In this study, we used the IGAP stage 1 cohort, which consisted of 17 008 patients with AD (mean [SD] age, 74.7 [7.7 years; 59.4% female) and 37 154 controls (mean [SD] age, 76.3 [8.1] years; 58.6% female) drawn from 4 different consortia across North America and Europe with genotyped or imputed data at 7 055 881 SNPs (for a description of the AD cases and controls within the IGAP stage 1 substudies, see the article by Lambert et al12). The relevant institutional review boards or ethics committees approved the research protocol of the individual GWASs used in the current analysis, and all participants gave written informed consent. The current analysis was finalized on July 21, 2015.

For gene expression analyses, we used publicly available total RNA expression data from 1647 autopsied brain tissues (from dorsolateral prefrontal cortex, visual cortex, and cerebellum) in a total of 549 brains of 376 patients with AD and 173 nondemented healthy controls from the Gene Expression Omnibus data set GSE44772.19 As previously described,19 all participants were diagnosed at intake and each brain underwent extensive neuropathology examination. Tissues were profiled on a custom-made Agilent 44K array of 40 638 DNA probes.

For neuropathology analyses, we used publicly available summary statistics genetic data (major allele, P values, and directionality of effect) of neuropathological measures from 4914 brain autopsies.20 This data set consists of meta-analysis results from a GWAS performed to identify genetic loci correlating with established AD neuropathologic phenotypes. We focused on neurofibrillary tangles (NFTs) and neuritic plaques (NPs). The NFTs were classified as 4 Braak groups (none, transentorhinal, limbic, isocortical) and the NPs were classified as present or absent (any NPs vs no NPs).20

Longitudinal cognitive decline data were obtained from 723 individuals (196 older controls, 355 individuals with mild cognitive impairment, and 172 patients with clinical AD) from the initial cohort of the Alzheimer’s Disease Neuroimaging Initiative (see eAppendix 2 in the Supplement for additional description). We restricted our analyses to participants with available genotype and baseline and follow-up neuropsychological assessments with the Alzheimer’s Disease Assessment Scale cognitive subscale (range, 6 months to 3.5 years; mean [SD], 1.56 [0.75] years; total of 3399 assessments).

Statistical Analysis
Conditional Quantile-Quantile plots

Using recently developed statistical methods to evaluate pleiotropic effects,8,11,21-23 we evaluated SNPs associating with AD (IGAP stage 1) and the 6 immune-mediated diseases. For given associated phenotypes A and B, pleiotropic enrichment of phenotype A with phenotype B exists if the proportion of SNPs or genes associated with phenotype A increases as a function of increased association with phenotype B. To assess for enrichment, we constructed conditional quantile-quantile (Q-Q) plots of nominal −log10(P) values for all AD SNPs and for subsets of SNPs determined by the significance of their association with the 6 immune-mediated diseases. In conditional Q-Q plots, the presence of enrichment is reflected as successive leftward deflections of the curve for phenotype A if the degree of deflection from the expected null line is dependent on the degree of association with phenotype B. To assess for polygenic effects below the standard GWAS significance threshold, we focused the Q-Q plots on SNPs with nominal −log10(P) < 7.3 (corresponding to P > 5 × 10−8). The enrichment seen can be directly interpreted in terms of true discovery rate (true discovery rate = 1 − false discovery rate [FDR]) (eAppendix 2 in the Supplement).

Conjunction FDRs

To identify specific loci involved with both AD and the 6 immune-mediated diseases, we computed conjunction FDRs.10,11Conjunction FDR is defined as the posterior probability that an SNP is null for either phenotype or both simultaneously, given that the P values for both traits are as small as or smaller than the observed P values. A conservative estimate of the conjunction FDR is given by the maximum statistic24 in taking the maximum of FDRtrait1|trait2 and FDRtrait2|trait1. Unlike the conditional FDR, which ranks disease- or primary phenotype–associated SNPs based on genetic “relatedness” with secondary phenotypes,8,11 the conjunction FDR minimizes the effect of a single phenotype driving the common association signal and pinpoints pleiotropic loci between the traits of interest (eAppendix 2 in the Supplement). We used an overall FDR threshold of less than 0.05, which means 5 expected false discoveries per 100 reported. Additionally, we constructed Manhattan plots based on the ranking of conjunction FDR to illustrate the genomic location of the pleiotropic loci.10,11 In all analyses, we controlled for the effects of genomic inflation by using intergenic SNPs (eAppendix 2 in the Supplement). Detailed information on conjunction Q-Q plots, Manhattan plots, and conjunction FDR can be found in eAppendix 2 in the Supplement and prior reports.10,11,21-23

Results
Conditional Q-Q plots

We observed SNP enrichment for AD across different levels of significance with CD, UC, T1D, RA, CeD, and PSOR (Figure 1). For progressively stringent P value thresholds for AD SNPs, ie, increasing values of nominal −log10(P), we found moderate enrichment, ie, leftward shift or decreasing values of empirical −log10(q), using CD, UC, T1D, and RA and minimal enrichment using CeD and PSOR (Figure 1). Removing the major histocompatibility complex (MHC)–associated SNPs did not result in attenuation of genetic enrichment (eFigure 1 in the Supplement), indicating that the observed pleiotropy between AD and the immune-mediated diseases was not confined to the MHC region. We observed a similar pattern of enrichment for CD, UC, T1D, RA, CeD, and PSOR SNPs conditional on AD SNPs, suggesting symmetric genetic enrichment between AD and the immune-mediated diseases (eFigure 2 in the Supplement).

Conjunction FDRs

At a conjunction FDR P < .05, we identified 8 SNPs that were associated with both AD and the immune-mediated diseases: (1) rs2516049 with PSOR; (2) rs12679874 with CD; (3) rs12570088 with CD; (4) rs2280231 with T1D; (5) rs8055533 with T1D; (6) rs7258465 with T1D; (7) rs16980051 with CD; and (8) rs2298428 with CD (Figure 2 and Table 2). Of these, rs2516049 (chromosome 6; closest gene HLA-DRB5; reference allele T; conjunction trait PSOR; conjunction FDR P = .04 for AD and PSOR, 5.37 × 10−5 for AD, and P = 6.03 × 10−15 for PSOR) and rs12570088 (chromosome 10; closest gene IPMK; reference allele A; conjunction trait CD; conjunction FDR P = .009 for AD and CD, 5.73 × 10−6 for AD, and 6.57 × 10−5 for CD) demonstrated the same direction of allelic effect between AD and the immune-mediated diseases (Table 2).

Gene Expression Analyses

First, we investigated whether HLA-DRB5 and IPMK RNA levels are altered in AD brains vs control brains (Gene Expression Omnibus data set GSE44772).19 Using logistic regression, we found significantly decreased IPMK transcript expression in AD brains compared with control brains (β [SE], −0.096 [0.013]; P = 7.57 × 10−13). Although we did not observe a significant effect for HLA-DRB5 in AD brains, we found significantly increased transcript expression for HLA-DRA (β [SE], 0.155 [0.024]; P = 1.97 × 10−10).

Association With Neuropathology and Cognitive Decline

Using publicly available data,20 we next evaluated the relationship between rs2516049 and rs12570088 and neuropathological measures from 4914 brain autopsies. We found that both rs2516049 (reference allele T; P = .01352; direction of effect of the reference allele of this SNP is the same in the GWAS pleiotropy analysis and the analysis of NFT pathology [++]) and rs12570088 (reference allele A; P = .03151; direction of effect of the reference allele of this SNP is the same in the GWAS pleiotropy analysis and the analysis of NFT pathology [++]) were significantly associated with worse NFT pathology. We did not find a significant effect between rs2516049 (reference allele T; P = .07144; direction of effect of the reference allele of this SNP is in the opposite direction in the GWAS pleiotropy analysis vs the analysis of NP pathology [+−]) and rs12570088 (reference allele A; P = .2168; direction of effect of the reference allele of this SNP is the same in the GWAS pleiotropy analysis and the analysis of NP pathology [++]) and NP pathology.

Finally, we assessed the relationship between rs2516049 and rs12570088 and longitudinal cognitive decline in 723 individuals (196 older controls, 355 individuals with mild cognitive impairment, and 172 patients with clinical AD). Using linear mixed-effects models, covarying for age, sex, baseline Clinical Dementia Rating sum of box score, and APOE ε4 carrier status (carrier vs noncarrier), we found a significant relationship between rs2516049 and longitudinal Alzheimer’s Disease Assessment Scale cognitive subscale scores (β [SE], 0.405 [0.190]; P = .03); we found no relationship between rs12570088 and longitudinal Alzheimer’s Disease Assessment Scale cognitive subscale scores (β [SE], 0.114 [0.450]; P = .75).

Discussion

In this study, we found that a subset of genetic variants that were associated with increased risk for immune-mediated diseases also conferred increased risk for AD (genetic pleiotropy). Two polymorphisms, namely rs2516049 (chromosome 6; closest gene HLA-DRB5) and rs12570088 (chromosome 10; closest gene IPMK), demonstrated the same direction of allelic effect between AD and immune-mediated diseases and were associated with greater intracranial burden of NFTs. We additionally observed that HLA-DRA (a paralog for HLA-DRB5) and IPMK transcript expression was different in AD brains compared with control brains. Considered together, these findings suggest that immune- and inflammation-associated genes, particularly the HLA locus and IPMK, likely influence AD pathogenesis and progression.

Our results illustrate that combining GWASs from multiple diseases and phenotypes provides important insights into shared genetic risk. By leveraging genetic signal in one phenotype, we were able to identify variants in a second phenotype that would otherwise not be detected using a single-phenotype approach. Whereas the conditional FDR improves statistical power for gene discovery by conditioning on the association with secondary phenotypes, the conjunction FDR identifies loci jointly associated with 2 traits. Using this pleiotropic approach, we found 8 polymorphisms that were associated with both AD and the 6 immune-mediated diseases. We found 2 directionally consistent pleiotropic loci (closest genes HLA-DRB5 and IPMK) between the 6 immune-mediated diseases and AD.

The human leukocyte antigen (HLA) region encodes proteins that are responsible for immune system regulation in humans; HLA-DR is an MHC class II cell surface receptor encoded by the HLA complex. Association with the SNP near HLA-DRB5 builds on previous work by identifying this locus as a risk factor for AD.12 We found evidence supporting an association between rs2516049 and greater NFT burden. We also found that HLA-DRA had greater expression in AD pathological brain samples, which suggests potential for increased intracranial inflammation in AD. It is important to note that we were unable to define the causal gene(s) on chromosome 6 responsible for our pleiotropic signal given the complexity of the MHC region. However, our findings suggest the need for finer mapping of the HLA region in AD.

Inositol polyphosphate multikinase (IPMK) catalyzes inositol polyphosphate and phosphoinositide turnover. It has also been implicated in many noncatalytic activities, including regulation of amino acid signaling to mTOR complex 1,25 as a transcriptional activator for immediate early genes26 and p53-mediated transcription, which facilitates cell death27 and nuclear messenger RNA export.28 In our study, the IPMK locus was associated with NFT pathology, and IPMK expression was lower in postmortem AD samples. Expression of IPMK has been shown to be lower in pathological samples from patients with Huntington disease as well as in cellular and animal models of Huntington disease.29 Phosphatidylinositol kinase activity (one of the enzymatic functions of IPMK) has also been shown to be lower in postmortem tissue from patients with AD.30 Considered together with our results, these studies suggest an important role for IPMK in neurodegeneration.

Although the primary focus of this work was on pleiotropic loci demonstrating the same direction of allelic effect, we also found 6 loci that showed an opposite direction of allelic effect between AD and the immune-mediated diseases. One hypothesis for these findings is that the observed pleiotropy between AD and these 6 loci could be due to different haplotypes or gene alleles involving the same SNPs. Another equally plausible hypothesis is that the same haplotypes or gene alleles are involved in both AD and these autoimmune diseases but the underlying biological mechanisms are distinct. Taken collectively, these findings illustrate that the genetic relationship between the immune system and AD may not be straightforward; considerable work will be required to carefully characterize the biological mechanisms underlying how each inflammation-associated genetic variant influences AD pathobiology.

We note that in this study the diagnosis of AD was established clinically. It is feasible that the individuals from the IGAP cohort who were clinically diagnosed as having AD may have had concomitant vascular brain disease with underlying inflammation, which may further contribute to their cognitive decline and dementia. As such, an alternative interpretation of our findings is that the susceptibility loci identified in this study may increase brain vulnerability to vascular and/or inflammatory insults, which in turn may exacerbate the clinical consequences of AD pathological changes.

Conclusions

We have identified genetic overlap between AD and immune-mediated diseases, implicating the HLA locus and IPMK in the pathobiology of AD. These findings provide novel insights into the relationship between inflammation and AD. Building on prior genetic and molecular evidence,31 our results are consistent with the hypothesis that rather than representing a downstream effect of neurodegeneration, inflammation influences AD pathogenesis and progression, which may have implications for treatment and prevention strategies in AD.

Back to top
Article Information

Group Information: The Alzheimer’s Disease Neuroimaging Initiative investigators are listed in eAppendix 1 in the Supplement.

Corresponding Author: Jennifer S. Yokoyama, PhD, Department of Neurology, University of California, San Francisco, 675 Nelson Rising Ln, Ste 190, San Francisco, CA 94158 (jennifer.yokoyama@ucsf.edu).

Accepted for Publication: January 12, 2016.

Published Online: April 18, 2016. doi:10.1001/jamaneurol.2016.0150.

Author Contributions: Dr Desikan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Yokoyama, McEvoy, Dillon, Mukherjee, Hess, Z. Miller, Rosen, Andreassen, Dale, Desikan.

Acquisition, analysis, or interpretation of data: Yokoyama, Wang, Schork, Thompson, Karch, Cruchaga, McEvoy, Witoelar, Chen, Holland, Brewer, Franke, Wilson, Z. Miller, Bonham, Shen, Rabinovici, Rosen, B. L. Miller, Hyman, Schellenberg, Karlsen, Andreassen, Dale, Desikan.

Drafting of the manuscript: Yokoyama, Karch, Rosen, Desikan.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Yokoyama, Wang, Schork, Thompson, Karch, Cruchaga, Witoelar, Rosen, Dale, Desikan.

Obtained funding: Rabinovici, Rosen, B. L. Miller, Schellenberg, Andreassen.

Administrative, technical, or material support: Cruchaga, Franke, Dillon, Hess, Z. Miller, Bonham, Rosen, B. L. Miller, Schellenberg, Karlsen, Dale, Desikan.

Study supervision: Yokoyama, Dillon, Wilson, Rosen, B. L. Miller, Schellenberg, Andreassen, Desikan.

Conflict of Interest Disclosures: Dr Brewer reported serving on advisory boards for Elan, Bristol-Myers Squibb, Avanir, Novartis, Genentech, and Eli Lilly and Co and holding stock options in CorTechs Labs, Inc and Human Longevity, Inc. Dr Rabinovici reported receiving research support from Avid Radiopharmaceuticals and speaking honoraria from GE Healthcare, Piramal Imaging, and Medscape. No other disclosures were reported.

Funding/Support: Data included in this study were obtained from the National Cell Repository for Alzheimer’s Disease (supported by grant U24 AG21886 from the National Institute on Aging) and the National Alzheimer’s Coordinating Center (supported by grant U01 AG016976 from the National Institute on Aging). This work was supported by grants AG046374 (Dr Andreassen), R01-AG045611 (Dr Rabinovici), P50-AG023501 (Drs B. L. Miller, Rosen, and Rabinovici), RC2DA029475 (Mr Schork), and R01HD061414 (Mr Schork) from the National Institutes of Health; grants 213837, 225989, 223273, and 237250 from the Research Council of Norway, European Union Joint Programme–Neurodegenerative Research (Dr Andreassen); grant 2013-123 from the South-Eastern Norway Health Authority (Dr Andreassen); the Norwegian Health Association (Dr Andreassen); the KG Jebsen Foundation (Dr Andreassen); the John Douglas French Alzheimer’s Foundation (Drs B. L. Miller and Rabinovici); grant K01 AG049152 from the National Institute on Aging (Dr Yokoyama); a Susan Marcus Memorial Fund Clinical Research Grant from the Association for Frontotemporal Degeneration (Dr Yokoyama); grant 2012-A-015-FEL from the Larry L. Hillblom Foundation (Dr Yokoyama); and a graduate fellowship from the Robert J. Glushko and Pamela Samuelson Foundation (Mr Schork). The Alzheimer’s Disease Genetics Consortium supported the collection of samples used in this study through grants U01AG032984 and RC2AG036528 from the National Institute on Aging (Dr Schellenberg). Data for this study were prepared, archived, and distributed by the National Institute on Aging Genetics of Alzheimer’s Disease Data Storage Site at the University of Pennsylvania, supported by grant U24-AG041689-01 from the National Institute on Aging (Dr Schellenberg). The International Genomics of Alzheimer’s Disease Project funding sources are described in eAppendix 2 in the Supplement.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Previous Presentation: This paper was presented at the 68th Annual Meeting of the American Academy of Neurology; April 18, 2016; Vancouver, British Columbia, Canada.

Additional Information: Data used in preparation of this article were obtained from the Alzheimer’s Disease Neuroimaging Initiative (ADNI) database (http://adni.loni.usc.edu). As such, the investigators within the ADNI contributed to the design and implementation of ADNI and/or provided data but did not participate in analysis or writing of this report. A complete listing of ADNI investigators can be found at http://adni.loni.usc.edu/wp-content/uploads/how_to_apply/ADNI_Acknowledgement_List.pdf.

Additional Contributions: Cisca Wijmenga, PhD, Department of Genetics, University of Groningen, Groningen, the Netherlands, and David van Heel, DPhil, Centre for Genomics and Child Health, Blizard Institute, Barts and The London School of Medicine and Dentistry, London, England, provided access to celiac disease data; they received no compensation. The International Genomics of Alzheimer’s Disease Project provided summary results data for these analyses.

References
1.
Brookmeyer  R, Johnson  E, Ziegler-Graham  K, Arrighi  HM.  Forecasting the global burden of Alzheimer’s disease.  Alzheimers Dement. 2007;3(3):186-191.PubMedGoogle ScholarCrossref
2.
Eikelenboom  P, Hoozemans  JJ, Veerhuis  R, van Exel  E, Rozemuller  AJ, van Gool  WA.  Whether, when and how chronic inflammation increases the risk of developing late-onset Alzheimer’s disease.  Alzheimers Res Ther. 2012;4(3):15.PubMedGoogle ScholarCrossref
3.
Schmidt  R, Schmidt  H, Curb  JD, Masaki  K, White  LR, Launer  LJ.  Early inflammation and dementia: a 25-year follow-up of the Honolulu-Asia Aging Study.  Ann Neurol. 2002;52(2):168-174.PubMedGoogle ScholarCrossref
4.
Koyama  A, O’Brien  J, Weuve  J, Blacker  D, Metti  AL, Yaffe  K.  The role of peripheral inflammatory markers in dementia and Alzheimer’s disease: a meta-analysis.  J Gerontol A Biol Sci Med Sci. 2013;68(4):433-440.PubMedGoogle ScholarCrossref
5.
Jonsson  T, Stefansson  H, Steinberg  S,  et al.  Variant of TREM2 associated with the risk of Alzheimer’s disease.  N Engl J Med. 2013;368(2):107-116.PubMedGoogle ScholarCrossref
6.
Guerreiro  R, Wojtas  A, Bras  J,  et al; Alzheimer Genetic Analysis Group.  TREM2 variants in Alzheimer’s disease.  N Engl J Med. 2013;368(2):117-127.PubMedGoogle ScholarCrossref
7.
Bradshaw  EM, Chibnik  LB, Keenan  BT,  et al; Alzheimer Disease Neuroimaging Initiative.  CD33 Alzheimer’s disease locus: altered monocyte function and amyloid biology.  Nat Neurosci. 2013;16(7):848-850.PubMedGoogle ScholarCrossref
8.
Desikan  RS, Schork  AJ, Wang  Y,  et al; Inflammation Working Group and International Genomics of Alzheimer’s Disease Project (IGAP) and DemGene Investigators.  Polygenic overlap between C-reactive protein, plasma lipids, and Alzheimer disease.  Circulation. 2015;131(23):2061-2069.PubMedGoogle ScholarCrossref
9.
Karch  CM, Cruchaga  C, Goate  AM.  Alzheimer’s disease genetics: from the bench to the clinic.  Neuron. 2014;83(1):11-26.PubMedGoogle ScholarCrossref
10.
Andreassen  OA, Djurovic  S, Thompson  WK,  et al; International Consortium for Blood Pressure GWAS; Diabetes Genetics Replication and Meta-analysis Consortium; Psychiatric Genomics Consortium Schizophrenia Working Group.  Improved detection of common variants associated with schizophrenia by leveraging pleiotropy with cardiovascular-disease risk factors.  Am J Hum Genet. 2013;92(2):197-209.PubMedGoogle ScholarCrossref
11.
Andreassen  OA, Thompson  WK, Schork  AJ,  et al; Psychiatric Genomics Consortium; Bipolar Disorder and Schizophrenia Working Groups.  Improved detection of common variants associated with schizophrenia and bipolar disorder using pleiotropy-informed conditional false discovery rate.  PLoS Genet. 2013;9(4):e1003455.PubMedGoogle ScholarCrossref
12.
Lambert  JC, Ibrahim-Verbaas  CA, Harold  D,  et al; European Alzheimer’s Disease Initiative; Genetic and Environmental Risk in Alzheimer’s Disease; Alzheimer’s Disease Genetic Consortium; Cohorts for Heart and Aging Research in Genomic Epidemiology.  Meta-analysis of 74,046 individuals identifies 11 new susceptibility loci for Alzheimer’s disease.  Nat Genet. 2013;45(12):1452-1458.PubMedGoogle ScholarCrossref
13.
Franke  A, McGovern  DPB, Barrett  JC,  et al.  Genome-wide meta-analysis increases to 71 the number of confirmed Crohn’s disease susceptibility loci.  Nat Genet. 2010;42(12):1118-1125.PubMedGoogle ScholarCrossref
14.
Stahl  EA, Raychaudhuri  S, Remmers  EF,  et al; BIRAC Consortium; YEAR Consortium.  Genome-wide association study meta-analysis identifies seven new rheumatoid arthritis risk loci.  Nat Genet. 2010;42(6):508-514.PubMedGoogle ScholarCrossref
15.
Barrett  JC, Clayton  DG, Concannon  P,  et al; Type 1 Diabetes Genetics Consortium.  Genome-wide association study and meta-analysis find that over 40 loci affect risk of type 1 diabetes.  Nat Genet. 2009;41(6):703-707.PubMedGoogle ScholarCrossref
16.
Dubois  PCA, Trynka  G, Franke  L,  et al.  Multiple common variants for celiac disease influencing immune gene expression.  Nat Genet. 2010;42(4):295-302.PubMedGoogle ScholarCrossref
17.
Ellinghaus  D, Ellinghaus  E, Nair  RP,  et al.  Combined analysis of genome-wide association studies for Crohn disease and psoriasis identifies seven shared susceptibility loci.  Am J Hum Genet. 2012;90(4):636-647.PubMedGoogle ScholarCrossref
18.
Anderson  CA, Boucher  G, Lees  CW,  et al.  Meta-analysis identifies 29 additional ulcerative colitis risk loci, increasing the number of confirmed associations to 47.  Nat Genet. 2011;43(3):246-252.PubMedGoogle ScholarCrossref
19.
Zhang  B, Gaiteri  C, Bodea  L-G,  et al.  Integrated systems approach identifies genetic nodes and networks in late-onset Alzheimer’s disease.  Cell. 2013;153(3):707-720.PubMedGoogle ScholarCrossref
20.
Beecham  GW, Hamilton  K, Naj  AC,  et al; Alzheimer’s Disease Genetics Consortium.  Genome-wide association meta-analysis of neuropathologic features of Alzheimer’s disease and related dementias.  PLoS Genet. 2014;10(9):e1004606.PubMedGoogle ScholarCrossref
21.
Andreassen  OA, Zuber  V, Thompson  WK,  et al; PRACTICAL Consortium; CRUK GWAS.  Shared common variants in prostate cancer and blood lipids.  Int J Epidemiol. 2014;43(4):1205-1214.PubMedGoogle ScholarCrossref
22.
Andreassen  OA, McEvoy  LK, Thompson  WK,  et al; International Consortium for Blood Pressure Genome-Wide Association Studies, Genetic Factors for Osteoporosis Consortium.  Identifying common genetic variants in blood pressure due to polygenic pleiotropy with associated phenotypes.  Hypertension. 2014;63(4):819-826.PubMedGoogle ScholarCrossref
23.
Andreassen  OA, Desikan  RS, Wang  Y,  et al.  Abundant genetic overlap between blood lipids and immune-mediated diseases indicates shared molecular genetic mechanisms.  PLoS One. 2015;10(4):e0123057.PubMedGoogle ScholarCrossref
24.
Nichols  T, Brett  M, Andersson  J, Wager  T, Poline  JB.  Valid conjunction inference with the minimum statistic.  Neuroimage. 2005;25(3):653-660.PubMedGoogle ScholarCrossref
25.
Kim  S, Kim  SF, Maag  D,  et al.  Amino acid signaling to mTOR mediated by inositol polyphosphate multikinase.  Cell Metab. 2011;13(2):215-221.PubMedGoogle ScholarCrossref
26.
Xu  R, Paul  BD, Smith  DR,  et al.  Inositol polyphosphate multikinase is a transcriptional coactivator required for immediate early gene induction.  Proc Natl Acad Sci U S A. 2013;110(40):16181-16186.PubMedGoogle ScholarCrossref
27.
Xu  R, Sen  N, Paul  BD,  et al.  Inositol polyphosphate multikinase is a coactivator of p53-mediated transcription and cell death.  Sci Signal. 2013;6(269):ra22.PubMedGoogle Scholar
28.
Wickramasinghe  VO, Savill  JM, Chavali  S,  et al.  Human inositol polyphosphate multikinase regulates transcript-selective nuclear mRNA export to preserve genome integrity.  Mol Cell. 2013;51(6):737-750.PubMedGoogle ScholarCrossref
29.
Ahmed  I, Sbodio  JI, Harraz  MM,  et al.  Huntington’s disease: neural dysfunction linked to inositol polyphosphate multikinase.  Proc Natl Acad Sci U S A. 2015;112(31):9751-9756.PubMedGoogle ScholarCrossref
30.
Jolles  J, Bothmer  J, Markerink  M, Ravid  R.  Phosphatidylinositol kinase is reduced in Alzheimer’s disease.  J Neurochem. 1992;58(6):2326-2329.PubMedGoogle ScholarCrossref
31.
Heppner  FL, Ransohoff  RM, Becher  B.  Immune attack: the role of inflammation in Alzheimer disease.  Nat Rev Neurosci. 2015;16(6):358-372.PubMedGoogle ScholarCrossref
×