Background and objectives
We have been utilizing two systems to study the mechanism of post-transcriptional gene silencing (PTGS). Both systems have characteristics that have become defined as PTGS. These include a reduction in mRNA levels while maintaining relatively high levels of transcriptional activity, and resetting of the mechanism at each generation. However, the predictability of silencing in the two systems differs. The first system is comprised of tobacco lines containing transgenes that encode tobacco etch virus (TEV) coat protein (CP) mRNA. Silencing of the TEV CP transgene is correlated with virus resistance. Two phenotypes for virus resistance in these lines have been described: immune (no virus infection, high transgene copy number), and recovery (initial systemic symptoms followed by gradual recovery over several weeks, low transgene copy number). We have shown that at early times in development, immune lines exhibit 'meiotic resetting' as is seen in the recovery lines, providing molecular evidence for a common mechanism of gene silencing and virus resistance in both cases. We also investigated the characteristics of two sets of low molecular weight RNAs that appear only in silenced tissue. The results obtained from these analyses support a hypothesis for transgene/viral RNA degradation in which RNA degradation involves an endoribonuclease and occurs in the cytoplasm while in association with polyribosomes.

The second system is comprised of Arabidopsis thaliana lines containing transgenes encoding phosphoribosylanthranilate transferase (PAT1, a member of the tryptophan biosynthetic pathway). These lines also contain an endogenous copy of the PAT1 gene. Silencing of PAT1 mRNA results in a co-suppression phenotype and is similar to PAT1 mutant phenotypes (i.e. blue fluorescence under UV light). However, the pattern of the silenced phenotype is variable among several independent transformants and the number of plants showing the blue fluorescent phenotype varies within homozygous populations. While segregation analysis indicates that the T-DNA insertions segregate as single loci, Southern analysis shows that all lines which silence PAT1 contain multiple T-DNAs. Based on the data from these two systems and from a number of other laboratories, we propose a hypothesis for TEV CP mRNA and PAT1 mRNA silencing in which the initiation of silencing in these two systems may differ while the maintenance mechanism is shared. The implications of this hypothesis in the analysis and development of transgenic plants which stably silence the transgenes will be discussed.