The reverse transcriptase (RT) of human immunodeficiency virus (HIV) catalyzes the critical first step of viral replication, the conversion of the viral RNA genome into a double stranded DNA suitable for incorporation into the genome of the host cell. Several of the anti-HIV drugs currently in use are targeted at RT, with the aim of inhibiting reverse transcription. The substrate for RT is a complex between the viral RNA and a primer tRNA from the host cell. If the structure of this initiation complex were known, it might be possible to design new pharmaceutical agents targeted at the initiation complex, in hopes of disrupting it or rendering it unrecognizable by RT. Since there is no high resolution structure from x-ray crystallography or NMR for the initiation complex, we propose to incorporate data from a variety of in vitro and in vivo experiments into concrete three dimensional models. Wherever warranted by the quantity and quality of data, our models will have atomic detail. In other regions, reduced representa- tions will be used, with a spherical pseudoatom of appropriate diameter representing each RNA nucleotide. The yammp computer program will be used to build and refine the models. For the atomic representations, yammp will be supplemented by MC-SYM, a constraint satisfaction RNA modeling program. These automated methods will be augmented by manual methods using computer molecular graphics - This combined approach will yield a set of models that are compatible with the data, along with quantitative information on the uncertainties in the models and conflicts in the experimental constraints. The models are essential to compiling the data from the chemical and enzymatic probing, NMR and genetic analyses. Further, the models will generate structural hypotheses that will be useful for defining priorities for further experimental work.
