The genome of RNA tumor viruses: a functional requirement for a polyploid structure?

This paper points out certain theoretical problems in DNA synthesis associated with antiprimer transcription and with circularization that could oblige RNA tumor viruses to rely on a polyploid genome. It is suggested that each completed act of reverse transcription may be coupled with an act of genetic recombination aimed at recovering the antiprimer information from an adjacent genome subunit in a polyploid train. A partially double-stranded DNA transcript could then be formed with sufficient terminal redundancy to permit circularization. The model provides satisfactory explanations for observed genetic interactions (particularly recombination and heterozygote formation), for inactivation data and for selective subunit transcription.