We previously reported that the 5 leader of the mRNA-encoding initiation factor eIF4G in Saccharomyces cerevisiae can function as a translational enhancer and as an internal ribosome entry site (IRES) when tested in cells. However, Verge and colleagues recently suggested that this sequence does not facilitate translation initiation, but inhibits translation in vitro and has promoter activity when tested in cells. We disagree with these conclusions and respond by showing that the data are most consistent with an internal initiation mechanism.Keywords: IRES; TIF4631; eIF4G; translation; mRNA Verge et al. (2004), in the February issue of RNA, reported that contrary to our earlier publication (Zhou et al. 2001), the 5Ј leader of the TIF4631 mRNA does not enhance translation initiation or contain an IRES. The authors assert that they have identified a promoter, located 36-112 nucleotides upstream of the translation initiation site, and concluded in their abstract, "The data show that the IRES activity reported earlier is due to this promoter". However, in our study, the IRES activity was localized primarily to a segment of the 5Ј leader that is completely independent of this putative promoter (nucleotides 250-390, see Fig. 1). Moreover, we showed that a segment of the TIF4631 gene containing the putative promoter had a background level of activity when tested in the intercistronic region of a dicistronic mRNA (construct p150 ). This low level of activity was ∼1% of the activity of the full-length 5Ј leader.The notion that the IRES activities observed in our study were derived from the translation of a shorter monocistronic mRNA is also not consistent with the high levels of activity reported in our paper. By accounting for the differences in luminescence between the Renilla and Photinus luciferases, we estimate that the second cistron expressed ∼40% as much protein as the first cistron in constructs containing the TIF4631 5Ј leader. Inasmuch as the dicistronic mRNA was easily detected in our study, a shorter monocistronic transcript should also have been easily identified; however, we found no evidence for such a transcript, even when our Northern blots were overexposed. Verge et al. (2004) reported promoter activity within the 5Ј leader of the TIF4631 mRNA; however, they did not directly test for this activity, for example, by testing the fragment in question in a promoterless vector or with Northern blots. Instead, indirect methods were used to assess promoter activity. For example, mRNA species were detected by PCR analysis. This is a highly sensitive technique that will detect trace amounts of RNA; however, their experiments were not performed in a quantitative manner, consequently it is not possible to compare mRNA levels or to draw conclusions about the production of a monocistronic mRNA from this data. This limitation is indicated by an internal inconsistency in their data: In transformed cells for which the expression of the dicistronic mRNAs was not induced, the authors obtained a PCR band correspond...