Plant-virus-based vectors for gene transfer may be of considerable use despite a presumed high error frequency during RNA synthesis.

The RNA genomes of some plant viruses have properties which make them suitable for development into vectors for gene transfer. It has been claimed that the low fidelity of RNA synthesizing enzymes may limit the usefulness of such vectors. Arguments are presented here to indicate that the problem may not be serious. Examples are presented of plant RNA replicons which carry relatively stable coding sequences whose products are superfluous to either the replicon or the host. In addition, there are numerous instances in which slowly replicating virus strains are maintained in 'pure' culture in a host and are not overgrown by more rapidly replicating strains. It is difficult to estimate the error frequency of RNA replicating enzymes. Methods which rely on host-shift may be subject to error.