The functions of Myf5 and Myod are well known in trunk myogenesis. However, the roles that Myf5 and Myod play during craniofacial myogenesis are far from well known. We observed that zebrafish myf5 was detected in the primordia of the obliques, lateral rectus, sternohyoideus, and pharyngeal mesoderm cores. In contrast, myod transcripts were expressed in all head muscle precursors at later stages. Knockdown of myf5 revealed that Myf5 was required for the development of the obliques, lateral rectus, sternohyoideus, and all pharyngeal muscles, whereas knockdown of myod proved that Myod was required for the development of superior rectus, medial rectus, inferior rectus, lateral rectus, and the ventral pharyngeal muscles. myod mRNA did not rescue the loss of the cranial muscle caused by injecting myf5-morpholino, or vice versa, suggesting that the functions of Myf5 and Myod were not redundant in head paraxial mesoderm, a finding different from their functions in trunk myogenesis. Myf5, but not Myod, was required for the forward migration of myf5-positive oblique precursors. All evidences reveal that Myf5 and Myod function independently during cranial myogenesis. On the basis of the expression patterns of myf5 and myod, we propose a model to present how Myf5 and Myod are involved in head myogenesis of zebrafish.

No abstract

Upstream open reading frame (uORF)-mediated translational inhibition is important in controlling key regulatory genes expression. However, understanding the underlying molecular mechanism of such uORF-mediated control system in vivo is challenging in the absence of an animal model. Therefore, we generated a zebrafish transgenic line, termed huORFZ, harboring a construct in which the uORF sequence from human CCAAT/enhancer-binding protein homologous protein gene (huORFchop) is added to the leader of GFP and is driven by a cytomegalovirus promoter. The translation of transgenic huORFchop-gfp mRNA was absolutely inhibited by the huORFchop cassette in huORFZ embryos during normal conditions, but the downstream GFP was only apparent when the huORFZ embryos were treated with endoplasmic reticulum (ER) stresses. Interestingly, the number and location of GFP-responsive embryonic cells were dependent on the developmental stage and type of ER stresses encountered. These results indicate that the translation of the huORFchop-tag downstream reporter gene is controlled in the huORFZ line. Moreover, using cell sorting and microarray analysis of huORFZ embryos, we identified such putative factors as Nrg/ErbB, PI3K and hsp90, which are involved in huORFchop-mediated translational control under heat-shock stress. Therefore, using the huORFZ embryos allows us to study the regulatory network involved in human uORFchop-mediated translational inhibition.

Myf5, one of the basic helix-loop-helix transcription factors, controls muscle differentiation and is expressed in somites during early embryogenesis. However, the transcription factors bound to the cis-elements of myf5 are poorly understood. In this study, we used the yeast one-hybrid assay and found that Forkhead box d3 (Foxd3) interacted specifically with the -82/-62 cassette, a key element directing somite-specific expression of myf5. The dual-luciferase assay revealed that the expression of Foxd3 potently transactivated the myf5 promoter. Knocking down foxd3 with morpholino oligonucleotide (MO) resulted in a dramatic down-regulation of myf5 in somites and adaxial cells but not in the presomitic mesoderm. On the other hand, myod expression remained unchanged in foxd3 morphants. Foxd3 mediation of myf5 expression is stage-dependent, maintaining myf5 expression in the somites and adaxial cells during the 7- to 18-somite stage. Furthermore, in the pax3 morphant, the expression of foxd3 was down-regulated greatly and the expression of myf5 was similar to that of the foxd3 morphant. Co-injection of foxd3 mRNA and pax3-MO1 greatly restored the expression of myf5 in the somites and adaxial cells, suggesting that pax3 induces foxd3 expression, which then induces the expression of myf5. This report is the first study to show that Foxd3, a well-known regulator in neural crest development, is also involved in myf5 regulation.