Characterization of chromosome 5q35 risk locus in African Ancestry population.

Abstract Background As part of the Research in African American Alzheimer Disease Initiative (REAAADI) and Late‐Onset AD Family Study (LOAD), genotyping array and whole genome sequencing (WGS) data were generated for 51 families (160 affected and 318 unaffected). Multipoint linkage analyses identified a peak LOD score on chr5q35 (HLOD=3.20). Additionally, a suggestive locus flanking the 1‐LOD score was previously identified in an AA GWAS study (p‐value=2.6×10 ‐6 ) (Kunkle et al, 2019). Here, we further characterize this locus and its role in AD. Method We performed region‐wide association analysis in an independent REAAADI dataset of ∼240 cases and ∼650 controls to fine map the signal in the locus. Additionally, we analyzed the WGS data to prioritize variants in the consensus regions based on segregation with disease among affected individuals and rarity (MAF<0.01). All variants were annotated for putative function including protein changes for coding variants and evidence for regulatory activity (ENCODE, RoadMap Epigenome) and chromatin interactions (publicly available HiC and promoter capture C) for noncoding variants. Result Region‐wide association analysis identified two regions of rare (MAF∼0.01) variants downstream (p=3.0×10 ‐8 ) and upstream (p=3.7×10 ‐6 ) of the 1‐LOD linkage region. Segregation analysis using WGS data identified 111 rare variants (MAF<0.01) segregating with disease in the AD individuals of the family with the highest LOD score contributing to the linkage peak. These include a 3’UTR and synonymous variant in INSYN2B/FAM196B as well as a promoter variant in WWC1/KIBRA . Interestingly, four other AA families contributing to the chr5 linkage signal harbor different within‐family shared variants located in INSYN2B ’s promoter or in enhancer regions with evidence for interaction with INSYN2B ’s promoter. WWC1 , expressed in astrocytes, was previously reported in AD context, whereas novel candidate INSYN2B encodes an inhibitory synaptic factor active in oligodendrocytes and neurons. Conclusion Our initial analyses provide evidence for two candidate genes contributing to AD genetics in the AA population. Additional work for functional validation of these candidates for AD is ongoing. This AA population‐specific finding shows the importance of diversifying population‐level genetic data to better understand the genetic determinants of AD on a global scale.