Genetic Architecture of Non-Syndromic Hearing Impairment in Ghana

Abstract

Background: Despite the genetic heterogeneity in hearing impairment (HI) aetiology, gap junction protein beta 2 (GJB2) pathogenic variants have been implicated in more than 50% of autosomal recessive hearing impairment that manifests as an isolated disorder worldwide. The reported pathogenic variants associated with HI are extremely population - specific, demonstrated by variable frequencies and allelic heterogeneity reported in different settings. Deleterious GJB2 variants associated with HI are described as stemming from either inheriting germline founder mutations or de novo (hotspots) mutations. The GJB2: c.427C>T p.(Arg143Trp) founder variant is the predominant cause of non-syndromic hearing impairment (NSHI) in Ghana. However, reported isolated incidence of GJB2-p.(Arg143Trp) variant in Asia, Europe, and Middle East though at relatively low frequencies, have raised questions on the variant origin. Therefore, this thesis sought to investigate the provenance and evolutionary history of the GJB2: c.427C>T-p.(Arg143Trp) founder variant systematically reported among Ghanaian HI cohorts, and elucidate the array of genetic marker(s) segregating in multiplex and simplex families negative for GJB2 variants. The hypothesis that the GJB2: c.427C>T p.Arg143Trp founder variant co-evolved as a loss of function or gain of function mutation, positively selected for by malaria endemic in the sub-region was also reviewed. Method: First-degree kindreds from multiplex and simplex families with NSHI were recruited from Mampong, Bechem, and Sekondi-Takoradi Schools for Deaf in Ghana. Genomic deoxyribonucleic acid (gDNA) was isolated from collected peripheral blood samples of recruited participants (18 multiplex and 23 simplex families). In-house developed polymerase chain reaction-restriction fragment length polymorphism (Ncil restriction enzyme) was used to screen recruited participants of the most common genetic marker (GJB2-p.Arg143Trp). Bi directional Sanger sequencing and family segregation analysis were performed to validate probands positive for the GJB2 variant. Selected unrelated GJB2-p.Arg143Trp homozygote participants and hearing controls negative for the founder mutation were whole-exome sequenced to estimate the age and ascertain the origin, and identify the haplotypic backgrounds sustaining the variant segregation in cohorts investigated. Unsolved families negative for the GJB2 variant were also sent for whole-exome sequencing (WES) (newly recruited and archival families). Illumina DRAGEN Germline developed pipeline and genome analysis toolkit (GATK) were used for alignment and calling variants. Identified variants from in silico bioinformatic analysis were validated with bi directional Sanger sequencing using designed sequence specific primers and family segregation analysis. The implicated mutant protein was modelled by homology modelling, using wild type protein sequence as template on web-based tool (SWISS-MODEL) and visualized in PyMOL. Results: At least pedigrees for three generational history and audiological assessments of families investigated showed bilateral severe to profound phenotype (sensorineural isolated HI), with one suspected syndromic HI. The estimated age was consistent between the WES unimputed and imputed computed ancestral informative markers (385 and 380 generations). The GJB2-p.(Arg143Trp)-positive individuals carries a unique haplotype background, specific and clearly distinct from the Ghanaian hearing controls, as well as individuals from Japanese and European populations, and shared no common haplotype. Most of the identified markers in linkage disequilibrium (LD) with the founder allele selected for age estimate were relatively absent or rare in Asia and Europe, demonstrating the independent evolutions and multiple ancestries of the variant. Analysis of WES data identified a novel MARVELD2: c.1058dupT-p.(Val354SerfsTer5) duplication on exon 2 as associated with the bilateral post lingual autosomal recessive non syndromic HI that segregates in a large consanguineous family. In silico variant pathogenicity analysis of the identified thymine (T) duplication at position 1058 in exon 2 predicted a protein with valine substituted by serine at codon 354. This non-synonymous mutation is also projected to change the amino acid reading frame, producing a protein lacking the c-terminal cytoplasmic domain (close to 100 amino acids truncation). The variant was also absent in 46 ethnolinguistic hearing controls and 151 simplex hearing impaired probands screened, as well as hearing loss population database. The observed associated post-lingual severe-profound phenotype in the index case expands the previously reported pre-lingual MARVELD2 variants associated with HI reported in other populations. Equally, the WES analysis also defined another novel variant in GREB1L [c.3041G>A: p.(Gly1014Glu)] as implicated in DFNA80 (deafness autosomal dominant type 80 ) in another family. In silico variant pathogenicity analysis showed a CADD score of 26.5, and variant absent from human genomic databases such as ClinVar, ClinGen, TopMed and gnomAD. Structural modeling and analysis predicted the variant to likely affect the secondary protein structure that may impact the proteins’ function. This was further supported with data showing increased levels of Greb1L expression in mice inner ear during growth and a reduced expression in adulthood, suggesting its importance in early development of the inner ear structures from single cell gene expression resource (gEAR). Conclusion: This thesis has achieved some important findings on the spectrum of gene(s) variant underlying NSHI segregation and highlighted the specific genetic landscape in Ghana. Carriers of the major genetic contributor to NSHI in Ghana are descendants of an indigenous ancestor who lived about ~9625 years ago, illuminating the selection at the locus in Ghana that predates other reports and demonstrated no shared genetic ancestry between homozygotes cohorts and hearing controls, as well as continental populations investigated. This study also shows the high level of allelic heterogeneity implicated in NSHI. GJB2 negative families investigated identified novel variants in known HI genes (GREBIL, MARVELD2, CDH23, among others) to segregate with the phenotype. The study findings have contributed to refining the architecture of NSHI genes in Ghanaian families, and the global knowledge on HI disease – gene pair curation.