Morphogenetic signaling by retinoic acid and its regulation during inner ear development
Final Report Abstract
We identified the zebrafish aldh1a3 gene, which encodes a retinoic acid (RA) synthesizing enzyme and determined its embryonic expression pattern. We found a very dynamic pattern of expression in the developing inner ear and, through a comparative genomics approach, established that Aldh1a3 and Aldh1a2 are the only RA synthesizing enzymes in teleosts. We successfully established a reporter line, Tg(aldh1a2:aldh1a2-gfp), by BAC recombineering, creating a gene fusion of gfp to the 3'-end of the aldh1a2 gene that is under the control of endogenous promoter/enhancer elements. The line faithfully recapitulates embryonic aldh1a2 expression patterns from late stages of gastrulation onwards. The transgene rescues homozygous aldh1a2 mutants into adult stages, confirming that all gene regulatory sequences that control aldh1a2 are present on the insertion. Importantly, our reporter line identifies novel sources of RA synthesis in postembryonic developmental stages, such as in mantle cells of the lateral line organ and cells associated with the branchial arch skeleton. In support of a previous study that predicted the existence of a source of RA that promotes vertebral ossification, we show aldh1a2:gfp expression in cells of the vertebral column. Finally, we report aldh1a2:gfp expression in the corpuscules of stannius, an organ derived from the pronephric kidney, which suggests a role for RA signaling in controlling calcium homeostasis. We showed that a Gal4-RARα fusion construct is responsive to RA signaling in Tg(UAS:GFP) reporter fish, i.e. it reports the presence of RA in tissues known to express activated RARs and regulates according to the presence and depletion of RA. Unfortunately, a stable transgene utilizing a feedbackactivated variant of this construct (FIND-RAR) turned out to be constitutively active and not regulated by RA. In the course of an unanticipated setback in the phenotypic analysis of RA function in the inner ear, we ran out of time to bring this project to publication. However, we continue to develop the Gal4-RAR; UAS:GFP system to a stage where stable transgenes can be used to detect the presence of bioactive RA. The construct will also be adapted to work in the cephalochordate Amphioxus, in which no methods currently exist to detect RA activity in vivo. Our original hypothesis, that RA controls the development of the epithelial projections that form the semicircular canals (SCCs), had to be dropped after we confirmed that pharmacological inhibition of Aldh1a-activity, as well as knockdown of aldh1a3 transcripts with two independent morpholinos, delayed rather than inhibited SCC formation and had no discernable effects on sensory epithelia. Additional support for the notion that aldh1a3 is not required for inner ear morphogenesis comes from a viable loss-of-function mutant in aldh1a3 (sa0118, obtained through TILLING), that so far has not revealed developmental defects in the inner ear. Because medaka fish have lost aldh1a3 in evolution, but unlike zebrafish express aldh1a2 in the endolymphatic duct, we continue to test the idea that this could be a functional compensation that ensures RA synthesis in this organ. We are examining aldh1a3sa0118 mutants for defects in the endolymphatic duct, adult vestibular system and for its ability to regenerate hair cells in the inner ear.
Publications
- 2008. Expression of zebrafish aldh1a3 (raldh3) and absence of aldh1a1 in teleosts. Gene Expression Patterns 8:141–147
Pittlik, S., Domingues, S., Meyer, A., Begemann, G.
