Premium
SORL1 ‐variant carriers in ADES‐ADSP: A higher level of variant pathogenicity associates with earlier age at onset of Alzheimer's disease
Author(s) -
Holstege Henne,
Hulsman Marc,
Charbonnier Camille,
GrenierBoley Benjamin,
Quenez Olivier,
Ahmad Shahzad,
Amin Najaf,
van Rooij Jeroen G.J.,
Grozeva Detelina,
Norsworthy Penny,
Hummerich Holger,
Kawalia Amit,
Mok Kin Y.,
Shoai Maryam,
DolsIcardo Oriol,
van Der Flier Wiesje,
Sims Rebecca,
Amouyel Philippe,
Hardy John,
Clarimon Jordi,
Mead Simon,
van Swieten John C.,
Ramirez Alfredo,
van Duijn Cornelia M.,
Williams Julie,
Nicolas Gaël,
Bellenguez Céline,
Lambert JeanCharles
Publication year - 2020
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1002/alz.044492
Subject(s) - nonsynonymous substitution , genetics , biology , missense mutation , loss function , gene , exome , exome sequencing , disease , allele , 1000 genomes project , single nucleotide polymorphism , genotype , mutation , genome , medicine , phenotype
Background Recent reports indicated that extremely rare protein truncating or missense variants in the SORL1 gene are associated with an increased risk and possibly a causative effect on developing Alzheimer’s Disease (AD). Here, we investigated the relationship between the predicted variant pathogenicity and the age at AD onset (a.a.o.) in the largest whole exome sequencing (WES) dataset of AD cases and controls world‐wide. Method We combined sequencing data from the ADES and ADSP consortia on a single server. After quality control and exclusion of carriers of pathogenic variants in genes associated with Mendelian inheritance of AD, 12,675 AD cases and 8,693 controls were available for analysis. Non‐synonymous variants were scored using the REVEL algorithm, and loss‐of‐function (LOF) variants were predicted using LOFTEE. The variant‐burden was determined across all genes and odds ratios (OR) for AD‐status were calculated using a multinomial logistic test across age groups. Result Of all genes in the genome, rare variants in the SORL1 gene had the strongest association with AD (p = 6.4 × 10 −14 ). We identified 320 rare nonsynonymous missense and loss‐of‐function SORL1 variants (MAF<1 × 10 −4 ). The carrier frequency in AD cases with a.a.o. <65, a.a.o. >65 and cognitively healthy controls was respectively 1.9%, 1.3%, 0.7%. Carrying a SORL1 variant associated with an increased risk of early onset AD (Figure A). Specifically, carriers of LOF variants had a median a.a.o. of 63 years (Figure C), and this associated with a 24.8‐fold increased chance of developing AD before 65 years compared to non‐carriers (Figure B). The median a.a.o. for carriers of nonsynonymous variants with REVEL>75 was 70 years, and such variants associated with estimated 7.6‐fold increased risk of developing AD between 65‐70 years (Figure B). Variants with lower REVEL scores associated with a lower AD‐risk. Conclusion The abundance of predicted pathogenic SORL1 genetic variants and the profound effect on AD risk makes the SORL1 protein product a promising target for the design of selective treatment strategies for affected individuals. Further, our results open the discussion of whether and how AD‐affected SORL1 ‐variant carriers and their (presymptomatic) family‐members should be clinically counselled. For this, pathogenicity screens that distinguish between pathogenic and non‐pathogenic non‐synonymous variants are warranted.