FGF-2 mRNA and its antisense message are expressed in a developmentally specific manner in the chick limb bud and...

FGF-2 protein is present in the ectoderm and mesoderm of the developing chick limb bud. Its importance has been shown by the ability of ectopically applied FGF-2 to replace the apical ectodermal ridge, allowing complete outgrowth and subsequent pattern formation of the limb bud. The first goal of this study was to determine whether FGF-2 mRNA was present in the same ectodermal and mesodermal regions of the chick embryo as FGF-2 protein. FGF-2 also has an antisense message that is convergently transcribed from the opposite DNA strand (Kimelman and Kirschner [1989] Cell 59:687-696; Volk et al. [1989] EMBO J. 8:2983-2988). The second goal was to demonstrate the expression and distribution of the antisense message. Using RNAse protection assays we detected a full length protected fragment that corresponds to chick embryo FGF-2 mRNA, and a partially protected fragment that corresponds to the antisense message. We used in situ hybridization to show that FGF-2 mRNA was present in the ectoderm and subjacent mesoderm of the chick wing bud. FGF-2 mRNA was also present in body ectoderm and undifferentiated mesoderm throughout the embryo, and in muscle cells, dorsal neural tube, and mesonephros. In situ hybridization also revealed evidence for the presence of the natural antisense message in the embryo in most, but not all, of the same regions as the FGF-2 mRNA. FGF-2 mRNA and its antisense message colocalized in undifferentiated limb mesoderm; however, antisense message was not detected in differentiated muscle or cartilage. It is important to note that FGF-2 mRNA was always present in the mesonephros but that the antisense message was never observed in the mesonephros, thereby providing an internal control for non-specific signal. Although little is known about its function, Kimelman and Kirschner ([1989] Cell 59: 687-696) proposed that the antisense message may increase turnover of FGF-2 mRNA. When we compared the in situ hybridization data of both mRNAs with levels of FGF-2 protein (Savage et al. [1994] Dev. Dyn. 198:159-170), interesting tissue specific patterns emerged that support this hypothesis.