Internal ribosome entry site-mediated translation of Smad5 in vivo: requirement for a nuclear event.

Smad5 is thought to relay signals of the bone morphogenetic protein pathway. The 5' untranslated region (5'UTR) of human Smad5 mRNA is long, has the potential to form secondary structures and contains five AUG codons. Here we show that the 5'UTR of Smad5 contains an internal ribosome entry site (IRES) located within 100 nt of the 3' end of the 5'UTR. The Smad5 IRES was 4-8-fold more active than the poliovirus IRES in C2C12 cells, which have osteoblastic differentiation ability, but was 5-10-fold less active than the poliovirus IRES in 293T cells. When an in vitro transcript of a dicistronic Smad5 IRES construct was transfected into C2C12 cells, the Smad5 IRES was not able to stimulate the translation of the downstream cistron, although the cap-dependent translation of the upstream cistron was efficient. In contrast, the poliovirus IRES in a dicistronic in vitro transcript was able to stimulate the translation of the downstream cistron to a similar extent as in the case of transfection of the corresponding dicistronic DNA construct. These results suggest that Smad5 IRES activity displays cell specificity and that some as yet unidentified nuclear event may be required for efficient Smad5 IRES-driven translation initiation.

Lower concentration of La protein required for internal ribosome entry on hepatitis C virus RNA than on poliovirus RNA.

Translation initiation of poliovirus and hepatitis C virus (HCV) RNA occurs by entry of ribosomes to the internal RNA sequence, called the internal ribosomal entry site (IRES). Both IRES bind to the La protein and are thought to require the protein for their translation initiation activity, although they are greatly different in both the primary and predicted secondary structures. To compare the La protein requirement for these IRES, we took advantage of I-RNA from the yeast Saccharomyces cerevisiae, which has been reported to bind to La protein and block poliovirus IRES-mediated translation initiation. In a cell-free translation system prepared from HeLa cells, yeast I-RNA inhibited translation initiation on poliovirus RNA as expected, but did not significantly inhibit translation initiation on HCV RNA. However, the translation initiation directed by either IRES was apparently inhibited by I-RNA in rabbit reticulocyte lysates, in which La protein is limiting. I-RNA-mediated inhibition of HCV IRES-dependent translation in rabbit reticulocyte lysates was reversed by exogenous addition of purified recombinant La protein of smaller amounts than necessary to reverse poliovirus IRES-dependent translation. These results suggest that HCV IRES requires lower concentrations of La protein for its function than does poliovirus IRES. Immunofluorescence studies showed that HCV infection appeared not to affect the subcellular localization of La protein, which exists mainly in the nucleus, although La protein redistributed to the cytoplasm after poliovirus infection. The data are compatible with the low requirement of La protein for HCV IRES activity.