Enigmas no escuro: revelando RNAs longos não-codificantes nos olhos do tetra-cego Astyanax mexicanus (De Filippi, 1853)

Abstract

Astyanax mexicanus, known as blind tetra and Mexican tetra, is a well-known model for evolutionary studies that have two morphotypes: the surface fish, similar in morphology to other species of Astyanax, and the cavefish, that has numerous adaptations to the cave environment, known as troglomorphic traits. Among the troglomorphic traits of the cavefish, the absence of eyes has been a major target in A. mexicanus studies. It is currently understood that the absence of eyes is the result of a degeneration process during development that, although it is well studied and has numerous genes identified as important candidates, is still not fully understood. One of the greatest knowledge gaps in this process is the role of long non-coding RNAs (lncRNAs), which are known to have a regulatory function in gene expression. Therefore, we sought to identify, describe, annotate and classify lncRNAs expressed in the eyes of cave and surface fish embryos. Additionally, we seek to predict interactions between lncRNAs and protein-coding genes, especially candidate genes for eye development. To achieve this, we used eye tissue RNA-Seq libraries from embryos, available in public databases, to assemble the de novo transcriptome from cave and surface fish. We submitted the transcriptomes to the workflow we developed for the identification of putative lncRNAs. We constructed a database of lncRNAs of A. mexicanus and Danio rerio and annotated the putative lncRNAs against it. Next, we mapped them against their respective genomes and classified and estimated interactions with protein-coding genes. Finally, we searched for interactions between lncRNAs and candidate genes. Consequently, we identified 33,092 putative lncRNAs in the cavefish and 22,453 in the surface fish, of which 21,120 and 13,922 are new lncRNAs, respectively. We successfully mapped most of the lncRNAs, which in cavefish represented 99.93% of the total lncRNAs. Furthermore, we found interactions for about 99% of the lncRNAs mapped, in both morphotypes. In the cavefish, we observed 205 lncRNAs interacting with 57 candidate genes and 143 lncRNAs associated with 72 candidate genes in surface fish. The numerous lncRNAs described in this study and the interactions with candidate genes may be a starting point in the way to explore the possible relationship of lncRNAs in the regulation of these genes and understand the role of lncRNAs in eye development and degeneration. This work can also help direct studies that seek to investigate the relationship of lncRNAs with the rapid evolution of adaptive traits. Therefore, our efforts represent an important step in studies with A. mexicanus, especially in eye development and degeneration, but it also sets a relevant precedent for exploring the role of lncRNAs in different processes.