When discussing the use of animal models, their biological similarities are often emphasized, but it is equally true that differences between model organisms and humans can be leveraged for important scientific discoveries. One particular area where that has become increasingly true is regenerative biology. Zebrafish are highly regenerative and one of the very interesting cell types it can regenerate but mammals cannot is the mechanosensory receptors of the inner ear that detect sound and control balance. These receptors, known as “hair cells,” regenerate completely in zebrafish and other non-mammalian vertebrates after being destroyed or damaged, while in mammals, hair cell death results in permanent hearing loss or vestibular dysfunction.
In order to understand hearing regeneration, it is essential to understand how genes respond to injury and how those responses are controlled in the genome. To study this phenomenon closely, we performed a targeted ablation of the mechanosensory receptors in adult zebrafish auditory and vestibular organs and characterized the epigenome and transcriptome at consecutive time-points during the process of regeneration. We were able to identify unique, cell-specific transcription factor (TF) motif patterns in the chromatin that opened specifically during regeneration. We correlated this emergent enhancer activity with differential gene expression to identify key gene regulatory networks driving regeneration. We detected a clear pattern of overlapping Sox- and Six- family transcription factor gene expression and binding motifs, suggesting a combinatorial program of TFs driving regeneration and cell identity. Using pseudo-time analysis of single-cell transcriptomic data, we showed that the support cells within the sensory epithelium changed cell identity to a more pluripotent “progenitor” cell population that could either proceed to differentiate into new hair cells or return to a support cell identity. We showed that sox2 expression became enriched in the progenitor cells and was reduced again when the cells differentiated in either direction. We identified a 2.6 kb DNA sequence element upstream of the sox2 promoter by scATAC-seq that dynamically changed in accessibility during hair cell regeneration. When we deleted it, the upstream regulator of sox2 showed a dominant phenotype that resulted in a hair cell regeneration-specific defect in both the lateral line and adult inner ear of zebrafish. By correlating cell-type, enhancer activation, and TF expression, we are beginning to understand the combinatorial “code” of TFs that initiate regeneration and instruct hair cell differentiation.