Transcriptomic Profiling of Zebrafish Hair Cells Using RiboTag

Maggie S. Matern, Alisha Beirl, Yoko Ogawa, Yang Song, Nikhil Paladugu, Katie S. Kindt, Ronna Hertzano

The zebrafish inner ear organs and lateral line neuromasts are comprised of a variety of cell types, including mechanosensitive hair cells. Zebrafish hair cells are evolutionarily homologous to mammalian hair cells, and have been particularly useful for studying normal hair cell development and function. However, the relative scarcity of hair cells within these complex organs, as well as the difficulty of fine dissection at early developmental time points, makes hair cell-specific gene expression profiling technically challenging. Cell sorting methods, as well as single-cell RNA-Seq, have proved to be very informative in studying hair cell-specific gene expression. However, these methods require that tissues are dissociated, the processing for which can lead to changes in gene expression prior to RNA extraction. To bypass this problem, we have developed a transgenic zebrafish model to evaluate the translatome of the inner ear and lateral line hair cells in their native tissue environment; the Tg(myo6b:RiboTag) zebrafish. This model expresses both GFP and a hemagglutinin (HA) tagged rpl10a gene under control of the myo6b promoter (myo6b:GFP-2A-rpl10a-3xHA), resulting in HA-tagged ribosomes expressed specifically in hair cells. Consequently, intact zebrafish larvae can be used to enrich for actively translated hair cell mRNA via an immunoprecipitation protocol using an antibody for the HA-tag (similar to the RiboTag mice). We demonstrate that this model can be used to reliably enrich for actively translated zebrafish hair cell mRNA. Additionally, we perform a global hair cell translatome analysis using RNA-Seq and show enrichment of known hair cell expressed transcripts and depletion of non-hair cell expressed transcripts in the immunoprecipitated material compared with mRNA extracted from whole fish (input). Our results show that our model can identify novel hair cell expressed genes in intact zebrafish, without inducing changes to gene expression that result from tissue dissociation and delays during cell sorting. Overall, we believe that this model will be highly useful for studying changes in zebrafish hair cell-specific gene expression in response to developmental progression, mutations, as well as hair cell damage by noise or ototoxic drug exposure.