Institute of Developmental and Molecular Biology, Texas A&M University, College Station, TX 77843-3155, USA

Background and objectives
Protoplasts from transgenic Nicotiana benthamiana expressing the viral p2 protein were found to support replication of brome mosaic virus (BMV) RNAs -1 and -3. However, when transfected with virion RNA that contains RNA-2 as well as RNAs -1 and -3, no replication of RNA-2 occurred. This is surprising because the three genomic RNA components of BMV are normally maintained at constant ratios. For example, transfection of protoplasts with an inoculum containing a given RNA at an excess ('unbalanced') level does not alter the ratio of progeny RNAs [1]. The results from the transgenic protoplasts indicate interference with mechanisms that control interactions between viral and host proteins and the viral RNAs during replication, and we wish to determine how the transgene interferes with the replication of RNA-2.

In some respects, the silencing of RNA-2 replication is similar to that obtained in previous experiments using deletion derivatives of BMV RNA [2]. These derivatives compete with intact RNA for replication, and were termed parasitic RNAs (pRNAs); their function appears to be analogous to that of defective interfering (DI) RNAs. However, the basis for silencing observed in p2 protoplasts appears to differ from that generally accepted for DI interference, since it does not involve the (-) or (+) strand promoter elements and is therefore unlikely simply to reflect saturation of the replicase by excess (non-productive) template.

We suspect that the high levels of the p2 protein contribute to the specific silencing of RNA-2 replication. Alternatively, this could reflect a novel case of RNA-mediated post-transcriptional gene silencing [3] where the p2 transgene RNA triggers the activation of a host defence system that specifically degrades RNA-2 at the same time as functional levels of p2 protein are produced.

Results and conclusions
Protoplasts from wild-type (wt) and p2 plants were inoculated with RNAs -2 and -3 (which cannot replicate because of the absence of RNA-1) and the degradation of each component was followed over a 24-h period. Similar degradation kinetics were obtained for both RNAs, showing that there was no preferential degradation of RNA-2 in the transgenic (p2) protoplasts. Hence induction of an RNA-2-specific nuclease is unlikely to be involved in the silencing observed. Segments of RNA-2 that collectively cover the entire genome were inserted into RNA-3, but all of these chimaeric molecules were replicated in wt and p2 protoplasts when co-transfected with RNA-1 (and RNA-2 in wt protoplasts). The lack of silencing of the chimaeric RNA-3 molecules strongly suggests that the silencing seen in this system is protein-mediated rather than RNA-mediated. As a model for the interference, we speculate that competitive interaction of excess p2 present in the transgenic protoplasts with the replicase complex displaces nascent RNA-2 (-) strands during replication, leading to their degradation by non-specific cellular RNases. As a result, no de novo RNA-2 (+) strand synthesis occurs.

1. Rao ALN, Hall TC, 1991. Virology 180, 16-22.
2. Marsh LE, Pogue GP, Connell JP, Hall TC, 1991. Journal of General Virology 72, 1787-1792.
3. Baulcombe DC, 1996. Plant Molecular Biology 32, 79-88.