Abstract Body

Genome-wide association studies (GWAS) have identified more than 40 loci associated with Alzheimer’s disease (AD), but the causal variants, regulatory elements, genes and pathways remain largely unknown, impeding a mechanistic understanding of AD pathogenesis. Previously, we showed that AD risk alleles are enriched in myeloid-specific epigenomic annotations. Here, we show that they are specifically enriched in active enhancers of monocytes, macrophages and microglia. We integrated AD GWAS with myeloid epigenomic and transcriptomic datasets using novel analytical approaches to link myeloid enhancer activity to target gene expression regulation and AD risk modification. We identify AD risk enhancers and nominate candidate causal genes among their likely targets (including AP4E1, AP4M1, APBB3, BIN1, MS4A4A, MS4A6A, PILRA, RABEP1, SPI1, TP53INP1, and ZYX) in twenty loci. Fine-mapping of these enhancers nominates candidate functional variants that likely modify AD risk by regulating gene expression in myeloid cells. In the MS4A locus we identified a single candidate functional variant and validated it in human induced pluripotent stem cell (hiPSC)-derived microglia and brain. Taken together, this study integrates AD GWAS with multiple myeloid genomic datasets to investigate the mechanisms of AD risk alleles and nominates candidate functional variants, regulatory elements and genes that likely modulate disease susceptibility.

Abstract Body

Decoding the genetic mechanisms of late-onset Alzheimer’s disease (LOAD) is a major challenge in the post-GWAS era, since the majority of the LOAD associated SNPs are in noncoding regions. Noncoding disease-associated loci have been shown to be enriched for regulatory elements in tissues and cells relevant to the disease. Thus, the advancement of genetic discoveries in LOAD requires an in-depth characterization of cell-type-specific DNA regulatory elements and gene expression patterns. Towards this goal we generated multi-omics profiles using NeuN sorted- and single- nuclei obtained from LOAD brain tissues compared to healthy controls. The ATAC-seq differential analyses identified LOAD-associated chromatin accessibility sites in neuronal-nuclei and in female non-neuronal cells. Furthermore, the LOAD-associated cell-type-specific changes in chromatin accessibility overlapped with ~25% of the LOAD-GWAS regions. The differential expression analyses using single-nuclei (sn)RNA-seq demonstrated LOAD-associated cell-type specific changes including for genes within LOAD-GWAS regions. We further functionally validated the ATAC-seq findings using parallel snRNA-seq datasets and demonstrated that LOAD-associated changes in chromatin accessibility in specific cell-types can result in gene dysregulation. Moreover, integrative genomic analysis identified overlaps between LOAD-specific chromatin accessibility signals, transcriptome profiles, and risk genetic variants. Thus, suggesting a noncoding regulatory mechanism by which several LOAD-GWAS loci may exert their pathogenic effect. Collectively, the outcomes unveiled how cell-type-specific and sex-dependent alterations in chromatin structure and transcription programs are associated with LOAD. These findings enhance the interpretation of LOAD-GWAS discoveries towards translating association to causation and provide potential pathomechanisms. Furthermore, our results convey mechanistic insights into sex differences in LOAD-risk and clinicopathology.

Aims

The Familial Alzheimer Sequencing (FASe) project aims to identify rare and high penetrant variants that have strong effect in the etiology of Alzheimer Disease (AD) by using sequencing data from families densely affected by late onset AD (fLOAD).

Methods

We have generated whole genome sequence (WGS) data for 952 samples (758 cases, 194 controls) from the Knight-ADRC at Washington University (WASHU), the NIALOAD and NCRAD repositories. These samples are being added to our current dataset of whole exome (WES) and WGS from 1,235 non-hispanic white participants (824 cases, 411 controls) across 285 fLOAD families. These samples have no or minimum overlap with the families sequenced by the ADSP consortia which will also be incorporated to our dataset; a total of 440 families and 3,187 samples (average of 5 cases and 2 controls per family) will be analyzed. We are processing all the data using the same bioinformatics pipeline. Briefly, sequence reads are aligned against reference build GRCh38 using BWA; variant calling is restricted to exonic regions following GATK v4.1.2 best practices. Data analysis includes single variant association, segregation, gene-based and pathway analysis.

Results

We have detected a genetic cross-over between AD, Frontotemporal Dementia and Parkinson disease, and we also identified rare variants in novel candidate genes for AD (PLD3, UNC5C, CPAMD8) highlighting the power of our dataset and the feasibility of our approach.

Conclusions

We hope to identify novel variants and pathways implicated on AD, which will be followed-up in the case-control ADSP.

Aims

The ancestral genetic heterogeneity of Caribbean Hispanics makes studies of this population critical to the discovery of ancestry-specific genetic factors in Alzheimer disease(AD). In this study, we performed whole-genome sequencing(WGS) in PR multiplex families and analyzed existing WGS in an independent PR case-control dataset(PR10/66), to identify rare causal variants influencing AD through linkage and segregation-based approaches.

Methods

WGS was performed in 100 individuals(61 affected) from 23 multiplex PR families. Parametric linkage analysis was performed and known AD genes were screened. To identify novel loci within the linkage region, we filtered variants using minor allele frequency(MAF), function potential CADD score, and segregation with AD within the families. Findings from the PR families were further evaluated using the PR10/66 dataset.

Results

A genome-wide significant linkage peak was found on 9p21 with HLOD score of 5.1, supported by 9 families with the strongest signal in Fam#87663(HLOD=1.8). 9p21 overlaps with the region previously reported in two linkage studies. The region harbors C9orf72 gene, but no expanded repeats in C9orf72 gene were observed in the PRADI families. We prioritized 7 variants that were observed in Fam#87663, were rare(MAF<0.01), have high functional potential(CADD>10) and showed evidence for association(P<0.05). Remarkably, a missense variant (rs35199210;CADD=20.3) in gene UNC13B segregated in two families, and another rare missense variant (rs41276043;CADD=25.0) in UNC13B gene segregated in a third family.

Conclusions

WGS in PR multiplex families suggested several novel candidates for AD, demonstrating the importance of family-based studies in discovering rare variants. Analysis is underway to assess the functional relevance of these candidates in AD.

Aims

In a cohort of Flanders Belgian Alzheimer’s disease (AD) patients, we identified 11 unrelated index patients carrying the presenilin 1 (PSEN1) missense mutation, p.Cys263Phe, plus three affected relatives of family DR1633 (n=14). We aimed to delineate a clinicopathological phenotype, and next compared this to genotype–phenotype data of AD patients carrying other causal gene mutations i.e. PSEN1 (n=25), PSEN2 (n=1), and APP (n=5).

Methods

Reviewing medical records of mutation carriers to obtain clinicopathological data for defining genotype-phenotype data.

Results

Mean onset age of the PSEN1 p.Cys263Phe carriers was 62.3±4.5 years (range 53-69), with a disease duration of 9.0±4.0 years (range 4-13). We observed a positive familial history in 90% of carriers and in family DR1633 co-segregation of AD was found with an autosomal dominant inheritance pattern. Amnestic presentation was present in all carriers, however, three patients also showed important frontal symptoms. Neuroimaging (n=10) displayed diffuse (sub)cortical atrophy, with evident hippocampal atrophy in three carriers. We observed severe signs of small vessel disease in four patients. Cerebrospinal fluid AD biomarkers were characteristic of AD in all. Neuropathology in two patients demonstrated severe levels of AD hallmarks plus cerebral amyloid angiopathy (CAA).

Carriers of PSEN1 p.Cys263Phe mutation had a later age at onset (62.3 years) than other PSEN1 carriers (50.8 years) or other causal gene mutation carriers (51.1 years).

Conclusions

PSEN1 p.Cys263Phe carriers presented with early-onset AD. Severe levels of AD neuropathology were seen with high levels of CAA. Disease onset of PSEN1 p.Cys263Phe carriers was later than other causal gene mutation carriers.

Aims

Strong efforts are still needed to characterize the genetic architecture of Alzheimer’s disease (AD). We thus conducted a complementary genome-wide association study (GWAS) with increased sample size and improved imputation quality of low frequency variants by applying the new TOPMed imputation panel.

Methods

The GWAS was performed in the European Alzheimer Disease Biobank (EADB) dataset. It groups together the main European AD GWAS consortia and a new dataset of 20,464 AD cases and 22,244 controls of European ancestry. Imputation was performed with the TOPMed reference panel or with the Haplotype Reference Consortium panel. The EADB results were meta-analysed with a proxy-AD GWAS performed in the UK Biobank, leading to a total Stage 1 sample size of 39,106 clinically diagnosed AD cases, 46,828 proxy-AD cases and 401,577 controls. The best hits from Stage 1 were finally tested in a large set of independent samples from ADGC and CHARGE.

Results

We identified 65 loci with a genome-wide significant signal of association (P<5x10^-8), including 31 new AD loci. The most significant gene sets identified by a pathway analysis relate to amyloid-beta and tau, while many of the other most significant sets relate to lipids and immunity.

Conclusions

The EADB project allowed us to identify several new associated loci for AD, including several candidate genes linked to amyloid precursor protein metabolism or that are likely to be involved in AD-related microglia dysfunctions, and loci already associated with the risk of developing other neurodegenerative diseases. Additional insights into the genetics of AD are expected from other ongoing analyses.

Abstract Body