Deletion of Clusterin Protects Cochlear Hair Cells against Hair Cell Aging and Ototoxicity

Abstract

Hearing loss is a debilitating disease that affects 10% of adults worldwide. Most sensorineural hearing loss is caused by the loss of mechanosensitive hair cells in the cochlea, often due to aging, noise, and ototoxic drugs. The identification of genes that can be targeted to slow aging and reduce the vulnerability of hair cells to insults is critical for the prevention of sensorineural hearing loss. Our previous cell-specific transcriptome analysis of adult cochlear hair cells and supporting cells showed that Clu, encoding a secreted chaperone that is involved in several basic biological events, such as cell death, tumor progression, and neurodegenerative disorders, is expressed in hair cells and supporting cells. We generated Clu-null mice (C57BL/6) to investigate its role in the organ of Corti, the sensory epithelium responsible for hearing in the mammalian cochlea. We showed that the deletion of Clu did not affect the development of hair cells and supporting cells; hair cells and supporting cells appeared normal at 1 month of age. Auditory function tests showed that Clu-null mice had hearing thresholds comparable to those of wild-type littermates before 3 months of age. Interestingly, Clu-null mice displayed less hair cell and hearing loss compared to their wildtype littermates after 3 months. Furthermore, the deletion of Clu is protected against aminoglycoside-induced hair cell loss in both in vivo and in vitro models. Our findings suggested that the inhibition of Clu expression could represent a potential therapeutic strategy for the alleviation of age-related and ototoxic drug-induced hearing loss.

The hair cell analysis toolbox is a precise and fully automated pipeline for whole cochlea hair cell quantification

“…We found no bias in accuracy with respect to estimated best frequency. To assess HCAT performance on a diverse set of cochlear micrographs, we sampled 88 images from 15 publications [ 16 – 30 ] that represent a wide variety of experimental conditions, including ototoxic treatment using aminoglycosides, genetic manipulations that could affect the hair cell anatomy, noise exposure, blast trauma, and age-related hearing loss ( Table 2 ). We performed a manual quantification and automated detection analysis of these images after they were histogram-adjusted and scaled via the HCAT GUI for optimal accuracy.…”

“… HCAT detection performance was assessed by running a cell detection analysis in the GUI on 88 confocal images of cochlear hair cells sampled from published figures across 15 different original studies [ 16 – 30 ]. None of the images from this analysis were used to train the model.…”

“…Comparing HCAT to manually annotated cochlear coils (not used to train the model), we report a 98.6 ± 0.005% true positive accuracy for cell identification and a <0.01% classification error (8 cochlear coils, 4,428 IHCs and 15,754 OHCs; S1 Fig) . We found no bias in accuracy with respect to estimated best frequency. To assess HCAT performance on a diverse set of cochlear micrographs, we sampled 88 images from 15 publications [16][17][18][19][20][21][22][23][24][25][26][27][28][29][30] that represent a wide variety of experimental conditions, including ototoxic treatment using aminoglycosides, genetic manipulations that could affect the hair cell anatomy, noise exposure, blast trauma, and age-related hearing loss (Table 2). We performed a manual quantification and automated detection analysis of these images after they were histogram-adjusted and scaled via the HCAT GUI for optimal accuracy.…”

The Hair Cell Analysis Toolbox: A machine learning-based whole cochlea analysis pipeline

“…We found no bias in accuracy with respect to estimated best frequency. To assess HCAT performance on a diverse set of cochlear micrographs, we sampled 88 images from 15 publications 15–29 that represent a wide variety of experimental conditions, including ototoxic treatment using aminoglycosides, genetic manipulations that could affect the hair cell anatomy, noise exposure, blast trauma and age-related hearing loss ( Table 2 ). We performed a manual quantification and automated detection analysis of these images after they were histogram-adjusted and scaled via the HCAT GUI for optimal accuracy.…”

“…HCAT detection performance was assessed by running a cell detection analysis in the GUI on 88 confocal images of cochlear hair cells sampled from published figures across 15 different original studies 15–29 . None of the images from this analysis were used to train the model.…”