We have shown in previous studies that a correlation exists between the presence of hypomodified tRNAs (i.e., tRNAs lacking a specific highly modified base) in infected cells and the utilization of these tRNAs in ribosomal frameshifting in the corresponding retrovirus. It has been shown in this and other laboratories that a mutation at the frameshift site in MMTV results in a dramatic reduction in frameshifting even though the same tRNA species decodes both the wild type and mutant signals. It is also known that Asp-tRNA without Q base preferentially decodes AAC over AAU codons. These studies suggest that a hypomodified tRNA is required for ribosomal frameshifting and in the event such a tRNA is required, it provides an avenue for inhibition of retroviral expression. Using Xenopus oocytes, we previously developed an in vitro assay for monitoring MMTV ribosomal frameshifting. RNA generated from the MMTV frameshift site, which had been cloned into an appropriate expression vector, was microinjected into oocytes and the retroviral translation product isolated by a specific immunoassay. The level of frameshifting was monitored following microinjection of Q base or asparagine tRNA with or without Q base. These studies have given inconclusive results where, on the one hand, some of the results suggested that hypomodified tRNA is required, but on the other, hypomodified tRNA is not required. Clearly, more studies are needed to resolve the important issue of whether a hypomodified tRNA is essential for frameshifting. In a collaborative study with Dr. S. Wilson, University of Texas, Galveston, TX, we have prepared a synthetic gene encoding the HIV primer, tRNA[Lys-3], where large quantities of tRNA[Lys-3] can be generated. These studies have provided strong evidence that the nucleic acid binding domain of tRNA[Lys- 3] in HIV reverse transcriptase is located between residues 273-302.
