Accurate, TATA box-dependent polymerase III transcription from promoters of the c-myc gene in injected Xenopus oocytes.

We have investigated the factors that permit a gene normally transcribed by RNA polymerase II to be transcribed by RNA polymerase III. It was shown previously that the human c-myc gene could be transcribed in vitro and in Xenopus oocytes by both alpha-amanitin-sensitive and alpha-amanitin-resistant polymerases, probably corresponding to polymerase II and polymerase III. We confirmed this observation in microinjected oocytes and showed that the alpha-amanitin-resistant transcription of c-myc was competed by known polymerase III genes. Polymerase III transcription of c-myc was very inefficient compared to other polymerase III genes, however, and was observed only when large amounts of template DNA were injected. At lower DNA concentrations the gene was transcribed, exclusively by polymerase II. In contrast, the adenovirus major late promoter was not transcribed by polymerase III. The 5' ends of polymerase III RNAs were almost indistinguishable from those of polymerase II RNAs initiating at the P1 and P2 promoters of the human and mouse c-myc genes. Furthermore, point mutations in the TATA box of the human P2 promoter greatly reduced polymerase III activity. At this promoter, therefore, polymerase II and polymerase III recognize a common element, the TATA box, which probably plays an important role in specifying the start site of transcription for both polymerases. We suggest that the highly accurate though inefficient mimicry of polymerase II by polymerase III at the c-myc promoters reflects the common evolutionary origin of these two enzymes.