Whole exome sequencing to investigate genetic variants of non-syndromic hearing impairment in a population of African ancestry

Master Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title

University of Cape Town

Introduction: Hearing impairment occurs when a child has hearing loss greater than 30dB in their better hearing ear and an adult cannot detect sound lower than 40dB in the better hearing ear. It is a common sensory disorder that affecting approximately 360 million worldwide, with an incidence of 6 in 1000 live births in developing countries such as those in Sub-Saharan Africa. 50 % of hearing impairment, in developed countries, is due to genetic factors, with 70% of genetic hearing impairment being classified as non-syndromic hearing impairment, which occurs when the hearing impairment presents with no other clinical manifestations. Hearing impairment is associated with over 150 genes, of which two connexin genes, GJB2 and GJB6, are the most prevalent genes associated with hearing impairment in European, Asian and North American of European ancestries populations. These genes have however been shown to be insignificant causes of Hearing Impairment in African populations. Aim: The aim of this study is to determine the rates for putative pathogenic variants in 172 hearing impairment associated genes, among Cameroonian patients affected by hearing impairment, and non-hearing-impaired controls. Methods: Patients and controls Patients were recruited from various schools of the Deaf and Ear, Nose and Throat (ENT) clinics in Cameroon. The patients were examined by qualified medical geneticists and ophthalmologist and detailed family history and medical history was obtained from the patients and their parents. 19 patients, who were negative for GJB2 and GJB6 mutations and presented with putative non-syndromic hearing impairment, were selected from a cohort of 582 patients for the present study. The control population consisted of 130 ethnically matched groups without any personal or familial history of hearing impairment. The controls were recruited from Yaoundé Central Hospital and Laquintinie Hospital in Cameroon. Whole exome sequencing DNA was extracted from whole blood using the salting out procedure and the Puregene Blood kit®. The DNA was subjected to spectrometry and gel electrophoresis to determine the quantity and quality of the DNA samples. The samples were then subjected to whole exome sequencing on the Illumina platform using the Nextera Rapid Capture Exome Kit at an average read depth of 30X, whereby only 18 patients were successfully sequenced. The exomes were then subjected to FastQC and SolexaQC++ for quality control measures and aligned to the hg19 reference genome using GATK and VariantMetaCaller. Bioinformatics analysis Variant annotation was performed using Annovar and the annotated variants were filtered based in rarity and pathogenicity. Tests for genetic differentiation and principle component analysis was performed on the combined patient exomes and combine control exomes. The first principle component analysis included data from African populations from the 1000 Genomes Phase 3 as well as six control samples from the Democratic Republic of Congo; and the second principle component analysis analysed on the Cameroonian patients and control population. Population structure analysis was followed by protein-protein interaction analysis using custom python and R script and pathway enrichment analysis using Enrichr combined with a second custom R script. The proportion of derived and ancestral alleles was computed by downloading the SNP ancestral alleles from Ensembl and verifying the presence of the SNPs in dbSNP database. The combined patient and control exomes were annotated using the VCFtools “fillOaa” script. The ancestral alleles were computed by dividing the number of times the alternative allele matched the ancestral allele with the number of copies of all the alternative alleles across all samples at the particular position. The ancestral alleles were categorised into six bins, based on their minor allele frequency, in the patient and control populations and this was used to contrast their proportions of derived and ancestral alleles. Furthermore, the proportion of ancestral and derived alleles in hearing impairment associated genes was computed at SNP based level for the Cameroonian population and contrasted with population from the Democratic Republic of Congo. Variants validation by Sanger sequencing Primers were designed to amplify the fragment surrounding the purported SNPs in MYO15A, MYO3A, and COL9A3 as well as for the fragments surrounding the population specific SNPs in VTN, RPL3L and DHRS4L2. Polymerase chain reaction was performed for the MYO15A, and MYO3A fragments. This was followed by purification of the PCR products and direct cycle Sanger sequencing of the PCR products. The sequencing products were then purified through ethanol precipitation and the fragments were suspended in HiDi Formamide and run on the capillary electrophoresis. The variants in MYO3A, MYO15A and COL9A3 were viewed in Integrated Genomics Viewer using the Bam files as well. Results Putative deleterious variants Single nucleotide polymorphism (SNPs) in MYO3A, MYO15A and COL9A3, were filtered out as putative causative mutations for three, four and two patients respectively. Direct Sanger Sequencing and viewing the patients BAM files did not confirm the presence of any of these putative pathogenic in the patients. Variations in USH2A, HSD17B4 and MYO1A were also filtered out but these variants were not considered disease causing, after a careful genotype to phenotypes correlations. Population genetics variants differentiations At a population level, specific variations were identified in FOXD4L2, DHRS2L6, RPL3L and VTN. Significant genetic differentiation was shown to exist between the control population and the patients’ population with regard to specific variants in VTN and RPL3L; furthermore, it was shown that these variants in VTN and RPL3L interact with other hearing impairment associated proteins with evidences that that VTN is hub protein for a hearing impairment associated pathway along with nine other genes. Conversely, this was not the case for variants described in FOXD4L2 and DHRS2L6. In known hearing Impairment genes, the proportion of ancestral alleles was lowest for the patients’ population for variations with minor allele frequencies between 0.0 and 0.1. The proportion of derived and ancestral alleles was also shown to differ between the Cameroonian and the population from the Democratic Republic of Congo, indication possible regional differences in aetiology of Hearing impairment amongst multiple African populations. Discussion Low putative pathogenic variants in known hearing impairment genes among Africans The low pick up rate for putative pathogenic variants in our patients follows a similar trend observed in the African American populations, with hearing impairment, as well as data from targeted exome sequencing from South African and Nigerian populations. This result is also in agreeance with other studies that interrogated hearing impairment in African populations utilising other means besides next generation sequencing. This result also highlights the importance of validating any results obtained from next generation sequencing through traditional approaches such as Sanger Sequencing or viewing the BAM files on IGV, specifically in African population, poorly represented in Exome databases. Bioinformatics Analysis Exhibited some Specific Variants among Cameroonian Protein-protein interactions and enrichment analysis indicated that VTN and RPL3L, and their interacting proteins, are significantly associated with osteoclast differentiation, which is associated with hearing impairment in osteogenesis imperfecta. VTN was further shown as a hub protein of a protein subnetwork, along ATPB2. The presence of a second protein acting as a hub protein may account for why aberrations in VTN have not been associated with a disease; whereby ATPB2 may ameliorate the pathogenic phenotype that ought to be observed in the presence of null mutations in VTN. Evolutionary adaptation of human hearing Data indicates the patient population carried a higher proportion of derived alleles in known hearing impairment genes, at low minor allele frequencies; possibly indicating, the interactive modifiers capacities of multiple hearing impairment genes, or alternatively, the polygenetic nature of hearing impairment in some patients. The proportion of ancestral and derived alleles was contrasted in the Cameroonian and the population from the Democratic Republic of Congo and it indicated that the variations that may result in hearing impairment in the one population may not be the same variations that result in hearing impairment in the other population Due to this, it is necessary to determine the causative variants resulting in disease in each of these populations independently. Conclusion and perspectives The results support a low pick up rate of putative variants in 172 known genes in groups of Cameroonian patients with HI, underscoring the current Targeted panel sequencing for HI may not be relevant for some African populations. The result also support the need of confirmation of variants found in WES, as well careful genotype to phenotypes correlations, particularly among African, whose sequences exome is relatively low in Exomes databases, and as a result could lead to more false positive results. Population genetic analysis has provided novel insight in the genetic architecture of HI among this group of Africans; particularly, the differential frequencies of ancestral alleles vs derived alleles in HI genes among patients vs controls underline the possibility of multigenic influence on the phenotype of Hearing Impairment that have not been well investigated, and may also signal evolutionary enrichment of some variants of HI genes in the populations as the result of natural selections, that deserve further investigation. The result supports the need of intensive familial studies in multiple African populations in order to unravel the novel genes and those variants that are relevant in clinical practice for people of African ancestry.