Preliminary Study Of Zebrafish Redd1and Redd2Genes In Vivo

Regulated in Development and DNA Damage response (REDD)1and REDD2are members of the DNA Damage Inducible Transcript (DDIT) or Growth ArrestDNA Damage (GADD) protein family. Proteins in this family participate in DNArepair, inflammatory and stress responses, and have anti-tumor actions. In mammals,REDD1is up regulated by hypoxia, oxidative stress, and starvation. The increasedREDD1in turn inhibits mTOR signaling and thereby modulates growth anddevelopment. Up-regulation of REDD1expression is implicated in several humandiseases, such as Parkinson’s disease, tumor growth, and skeletal muscle atrophy. Theclosely related REDD2is functionally complementary with REDD1.The majority of previous studies on REDD1and REDD2have been performedusing various mammalian cell lines in vitro. There is limited information on their invivo expression pattern. Likewise, there is limited knowledge about the in vivofunctions of endogenous REDD1and REDD2. Although the actions of mammalianREDD1and REDD2in regulating mTOR signaling are well documented, it is unclearwhether REDD1and REDD2also influence other signaling pathways.In this study we use zebrafish as a model system to investigate the structure, geneexpression and regulation, as well as the functions of REDD1and REDD2genes.Zebrafish redd1and redd2genes were cloned and characterized, their spatiotemporalexpression pattern and physiological regulation determined. Moreover, their functionsand the underlying molecular mechanisms have been studied by loss-of-function andover-expression studies. Zebrafish redd1is located on chromosome12. It contains3exons and2introns. The full cDNA is1,334bp containing an open reading frame(ORF) of663bp. Zebrafish redd2is located on chromosome14and contains2exonsand1intron. The full cDNA is1,024bp with an ORF of609bp. Sequence comparison, phylogenetic, and synteny analyses indicate that the chromosomallocation, gene, and protein structure are similar to their mammalian orthologs. redd1mRNA was easily detectable in all adult tissues examined. redd2mRNA was mostabundantly expressed in liver, kidney, and ovarian tissues, and were very low in otheradult tissues. During embryogenesis, zebrafish redd1and redd2mRNA levels wereexpressed during the whole embryogenesis but were higher at0,2, and4hours postfertilization periods. In situ hybridization revealed that the redd1transcript waspresent in the fertilized eggs and in all blastodermal cells. At the shield stage, redd1mRNA expression became restricted to the germ ring, where mesoderm precursorsreside. During the segmentation stage and thereafter, redd1mRNA was detectedmainly in the prechordal plate, posterior mesoderm and anterior ectoderm. Comparedwith redd1mRNA, redd2mRNA was more widely spread during early development.During the segmentation stage and thereafter, redd2mRNA signal was found incerebellum and mesencephalon. Among various stressors tested, heat shock andstarvation increased the levels of redd1and redd2mRNA, whereas hypoxia treatmentincreased the levels of redd1mRNA only. Knockdown of redd1or redd2resulted indorsalized phenotypes, such as shortened yolk sac extension, a loss of the caudalventral fin and curved tail. Over-expression of Redd1or Redd2resulted in ventralizedphenotypes, including defective anterior brain and notochord, enlarged tail, andsomites. These morphological changes were confirmed by in situ hybridization resultsusing a number of dorsal-ventral axis marker genes. Knockdown of redd1or redd2expanded the expression of dorsal marker genes chd and gsc while reduced theexpression of ventral marker genes eve1and ved. Over-expression of Redd1or Redd2had the opposite effects. These functional data indicate that Redd1and Redd2regulatedorsoventral pattern formation in zebrafish embryos.Recent studies suggest that another DDIT family member is involved in Wntsignaling and plays a role in dorsoventral pattern formation in Xenopus laevis.Whether REDD1and REDD2regulate embryo development by interacting with Wntsignaling is unknown at present. In this study, we tested whether zebrafish Redd1orRedd2regulates dorsoventral pattern formation through the Wnt signaling pathway. Over-expression of Redd1or Redd2in zebrafish embryos not only inhibited aWnt3a-induced increase in vivo, but also reduced the endogenous canonical Wntsignaling activity. Likewise, over-expression of zebrafish Redd1or Redd2in culturedhuman cells also antagonized Wnt3a-induced canonical Wnt signaling, suggesting thatthis action of Redd1and Redd2is evolutionarily conserved. To further investigatehow Redd1and Redd2interact with Wnt signaling pathway, we co-expressed Redd1or Redd2with β-Catenin N. β-Catenin is an important effector of the canonical Wntpathway and β-Catenin N is a constitutively active mutant form of β-Catenin.Over-expression of β-Catenin N resulted in dorsalized embryos. Co-expression ofeither Redd1or Redd2inhibited β-Catenin N-induced dorsalized phenotype.Similarly, co-expression of zebrafish Redd1or Redd2with β-Catenin N in culturedhuman cells inhibited β-Catenin N-induced Wnt reporter gene activity in adose-dependent manner. These results suggest that Redd1and Redd2negativelyregulate the canonical Wnt signaling activity downstream of the β-Catenin level.The results of this study demonstrate that the structure of Redd1and Redd2ishighly conserved in zebrafish. Zebrafish redd1and redd2genes have overlapping yetdistinct spatial and temporal expression pattern and respond to similar but not distinctstressors. Our study using the zebrafish model reveals that Redd1and Redd2regulatedorsoventral pattern formation by antagonizing the Wnt/β-Catenin signaling pathway.These findings not only deepen our understanding of REDD/Redd, but also unravel anovel mechanism linking the stress-response REDD1and REDD2genes to theWnt/β-Catenin signaling in vertebrate embryo.