The Genetic Architecture of Noise-Induced Hearing Loss: Evidence for a Gene-by-Environment Interaction

Abstract

The discovery of environmentally specific genetic effects is crucial to the understanding of complex traits, such as susceptibility to noise-induced hearing loss (NIHL). We describe the first genome-wide association study (GWAS) for NIHL in a large and well-characterized population of inbred mouse strains, known as the Hybrid Mouse Diversity Panel (HMDP). We recorded auditory brainstem response (ABR) thresholds both pre and post 2-hr exposure to 10-kHz octave band noise at 108 dB sound pressure level in 5-6-wk-old female mice from the HMDP (4-5 mice/strain). From the observation that NIHL susceptibility varied among the strains, we performed a GWAS with correction for population structure and mapped a locus on chromosome 6 that was statistically significantly associated with two adjacent frequencies. We then used a "genetical genomics" approach that included the analysis of cochlear eQTLs to identify candidate genes within the GWAS QTL. In order to validate the gene-by-environment interaction, we compared the effects of the postnoise exposure locus with that from the same unexposed strains. The most significant SNP at chromosome 6 (rs37517079) was associated with noise susceptibility, but was not significant at the same frequencies in our unexposed study. These findings demonstrate that the genetic architecture of NIHL is distinct from that of unexposed hearing levels and provide strong evidence for gene-by-environment interactions in NIHL.

Keywords: GWAS; HMDP; NIHL; eQTL; gene-by-environment.

Figure 1

Effects of genetic background on postnoise exposure ABR thresholds at 4 kHz (A), 8 kHz (B), 12 kHz (C), 16 kHz (D), 24 kHz (E), and 32 kHz (F) tone burst in 100 HMDP inbred strains. The difference between the strains with the lowest and the highest values were 2.13- (4 kHz), 3.16- (8 kHz), 4.34- (12 kHz), 3.72- (16 kHz), 3.53- (24 kHz), and 3.74-fold (32 kHz). Several strains demonstrated variable susceptibility depending on the frequency tested. The most susceptible strains at 4, 8, 12, 16, 24, and 32 kHz were SEA/GnJ, BXD16/TyJ, NOR/LtJ, BXD32/TyJ, DBA/2J, and AXB15/PgnJ, respectively.

Figure 2

Pairwise correlations between baseline ABR threshold and postnoise exposure threshold for 8, 12, 16, 24, and 32 kHz (A, B, C, D, and E, respectively) after exclusion of baseline hearing impaired strains. These plots were selected as they were representative of the correlations delineated in Table 2.

Figure 3

(A) GWAS results for 32 kHz postnoise exposure thresholds in the HMDP. Manhattan plot showing the association (−log10) P-values (−logP) for 32 kHz in 100 HMDP inbred mouse strains. The analysis was performed using >200,000 SNPs with a minor allele frequency >5%. Each chromosome is plotted on the x-axis in alternating brown and blue colors. SNP on chromosome 6 exceeded the predetermined genome-wide significance threshold (P = 4.1 × 10⁻⁶). (B) GWAS results for 24 kHz postnoise exposure thresholds in the HMDP. Manhattan plot showing the association (−log10) P-values (−logP) for 24 kHz in 100 HMDP inbred mouse strains. The analysis was performed using >200,000 SNPs with a minor allele frequency >5%. Each chromosome is plotted on the x-axis in alternating brown and blue colors.

Figure 4

Regional plot of the 24 kHz and 32 kHz ABR threshold on chromosome 6 association in the HMDP centered on the most significant SNP (rs37517079). The blue diamond represents the most significant SNP (P = 9.8 × 10⁻⁹) at 32 kHz. Phenotype and SNPs are colored based on their linkage disequilibrium (LD) with the most significant SNP being red SNPs in LD at r2 > 0.8, orange SNPs in LD at coefficient of determination (r2) > 0.6, and green SNPs in LD at r2 > 0.4. The positions of all RefSeq genes are plotted using genome locations (NCBI’s Build37 genome assembly). The value on the y-axis represents the −log10 of the P-value and corresponds to the genome-wide significance.

Figure 5

Cochlear eQTL plot. Diagonal line represents cis eQTLs. Dense vertical lines represent trans-eQTL hotspots. x-axis, expression SNP (eSNP) position; y-axis, probe position.