Exome sequencing implicates ancestry-related Mendelian variation at SYNE1 in childhood-onset essential hypertension

Abstract

Childhood-onset essential hypertension (COEH) is an uncommon form of hypertension that manifests in childhood or adolescence and, in the United States, disproportionately affects children of African ancestry. The etiology of COEH is unknown, but its childhood onset, low prevalence, high heritability, and skewed ancestral demography suggest the potential to identify rare genetic variation segregating in a Mendelian manner among affected individuals and thereby implicate genes important to disease pathogenesis. However, no COEH genes have been reported to date. Here, we identify recessive segregation of rare and putatively damaging missense variation in the spectrin domain of spectrin repeat containing nuclear envelope protein 1 (SYNE1), a cardiovascular candidate gene, in 3 of 16 families with early-onset COEH without an antecedent family history. By leveraging exome sequence data from an additional 48 COEH families, 1,700 in-house trios, and publicly available data sets, we demonstrate that compound heterozygous SYNE1 variation in these COEH individuals occurred more often than expected by chance and that this class of biallelic rare variation was significantly enriched among individuals of African genetic ancestry. Using in vitro shRNA knockdown of SYNE1, we show that reduced SYNE1 expression resulted in a substantial decrease in the elasticity of smooth muscle vascular cells that could be rescued by pharmacological inhibition of the downstream RhoA/Rho-associated protein kinase pathway. These results provide insights into the molecular genetics and underlying pathophysiology of COEH and suggest a role for precision therapeutics in the future.

Keywords: Genetic variation; Genetics; Hypertension; Population genetics.

Figure 1. Rare, predicted damaging SYNE1 variants in COEH probands.

(A) Pedigrees and variant segregation in affected families. n, multiple individuals of unknown or mixed sexes. (B) Variant position, annotation, in silico predictions, and frequencies as well as main clinical features of probands. Chromosome position (Chr) is relative to the hg19 human genome reference. Dam., damaging prediction proportion; CADD, Combined Annotation-Dependent Depletion; Cons., conservation prediction proportion; Hom., number of homozygotes reported; HTN, hypertension; BMI, body mass index; LDL, low-density lipoprotein; LVH, left ventricular hypertension; Med, medications; Aml, amlodipine; Lis, lisinopril. (C) Locations of impacted SYNE1 (Nesprin-1) amino acid residues. KASH, Klarsicht, ANC-1, Syne homology.

Figure 2. Comparison of biallelic SYNE1 missense proportions.

Statistical comparisons were performed using a binomial test. **P < 0.01; ***P < 0.001. Group1, group 1 of COEH cohort (n = 16 probands); COEH, full COEH cohort (n = 64 probands); BHCMG, Baylor-Hopkins Center for Mendelian Genomics; AFR, gnomAD self-reported African samples; GEN, gnomAD combined samples; NFE, gnomAD self-reported non-Finnish European.

Figure 3. Population ancestry studies.

(A) Principal component analysis (PCA) of ES-derived genetic ancestry for “super-populations” (super_pop) from the 1000 Genomes Project (AFR, African; EAS, East Asian; SAS, South Asian; EUR, European; AMR, Latino/Admixed American) and COEH. n = 64 (probands with SYNE1 variants are labeled). (B) Mean minor allele frequencies (MAFs) among gnomAD populations of 6 putative disease variants identified in COEH. *P < 0.05. (C) MAFs of all SYNE1 rare and damaging missense variants in gnomAD. ***P < 0.001; **P < 0.01. Ethnicity abbreviations are per gnomAD: AFR, African; AMR, Latino/Admixed American; ASJ, Ashkenazi Jewish; EAS, East Asian; FIN, Finnish European; NFE, non-Finnish European; SAS, South Asian. Statistical significance in B and C was determined by 1-way ANOVA.

Figure 4. SYNE1 functional studies.

Low- (A) and higher (B) magnification scans to 130 and 30 μm (x-y), respectively, showing detailed topography of control and SYNE1-KD cells from atomic force microscopy imaging. (C) Height images (60 μm² scans) were used to evaluate roughness within a 20 μm (x-y) measuring box to define a specific area of analysis (400 nm²) in each cell. (D) Roughness estimates between SYNE1-KD cells and control (wild-type) cells (n = 2 each), quantifying presumed actin filament formation. (E) Comparison of stiffness (measured as Young’s modulus) in control (wild-type) cells, SYNE1-KD cells, and both wild-type and SYNE1-KD cells treated with fasudil. Box plots show the interquartile range, median (line), and minimum and maximum (whiskers). Statistical comparisons were performed using 1-way ANOVA. ***P < 0.001.