Gene-mapping study of extremes of cerebral small vessel disease reveals TRIM47 as a strong candidate

Abstract

Cerebral small vessel disease is a leading cause of stroke and a major contributor to cognitive decline and dementia, but our understanding of specific genes underlying the cause of sporadic cerebral small vessel disease is limited. We report a genome-wide association study and a whole-exome association study on a composite extreme phenotype of cerebral small vessel disease derived from its most common MRI features: white matter hyperintensities and lacunes. Seventeen population-based cohorts of older persons with MRI measurements and genome-wide genotyping (n = 41 326), whole-exome sequencing (n = 15 965), or exome chip (n = 5249) data contributed 13 776 and 7079 extreme small vessel disease samples for the genome-wide association study and whole-exome association study, respectively. The genome-wide association study identified significant association of common variants in 11 loci with extreme small vessel disease, of which the chr12q24.11 locus was not previously reported to be associated with any MRI marker of cerebral small vessel disease. The whole-exome association study identified significant associations of extreme small vessel disease with common variants in the 5' UTR region of EFEMP1 (chr2p16.1) and one probably damaging common missense variant in TRIM47 (chr17q25.1). Mendelian randomization supports the causal association of extensive small vessel disease severity with increased risk of stroke and Alzheimer's disease. Combined evidence from summary-based Mendelian randomization studies and profiling of human loss-of-function allele carriers showed an inverse relation between TRIM47 expression in the brain and blood vessels and extensive small vessel disease severity. We observed significant enrichment of Trim47 in isolated brain vessel preparations compared to total brain fraction in mice, in line with the literature showing Trim47 enrichment in brain endothelial cells at single cell level. Functional evaluation of TRIM47 by small interfering RNAs-mediated knockdown in human brain endothelial cells showed increased endothelial permeability, an important hallmark of cerebral small vessel disease pathology. Overall, our comprehensive gene-mapping study and preliminary functional evaluation suggests a putative role of TRIM47 in the pathophysiology of cerebral small vessel disease, making it an important candidate for extensive in vivo explorations and future translational work.

Keywords: GWAS; TRIM47; cerebral small vessel disease; endothelial cells; whole-exome association study.

Figure 1

Contributing studies and defined number of extreme-SVD cases and controls for GWAS and WEAS.

Figure 2

Miami plot of GWAS and WEAS of extreme-SVD. Top: The GWAS plot reports chromosomal location and nearest gene to the top genome-wide significant variant. Bottom: The WES plot reports whole exome-wide significant associations with the most likely functional variants and affected gene.

Figure 3

SMR associations at chr17q25.1 (A) SMR associations in brain and blood vessels; (B) SMR effect size plot of TRIM47 eQTLs in brain; and (C) SMR effect size plot of TRIM47 eQTLs in blood vessels.

Figure 4

TRIM47 depletion in human brain endothelial cells increases endothelial permeability. (A) Efficient reduction of TRIM47 expression in endothelial cells was achieved by siTRIM47 #1 and 2 versus siControl. Ve-cadherin expression, a marker of endothelial cell junction was followed and α tubulin was used as a loading control. (B) Endothelial transwell permeability assay. Quantification of dextran leakage expressed as fold change compared to HBMEC and (C) hCMEC/D3 (D) cells treated with control siRNA TRIM47 #1 and #2 versus siControl. Fluorescent intensity was normalized to siControl knock down. Two-way ANOVA with a Dunnett’s multiple comparisons test was performed. Data are presented as mean ± SD. **P-value < 0.01, ***P < 0.0001, ***P > 0.00001.