John lnnes Centre, Norwich Research Park, Colney, Norwich NR4 7UH, UK

Background and objectives
The most obvious feature of the compatible interaction between a virus and its host is the appearance of symptoms. These symptoms are the end result of molecular responses at the cellular and tissue level that affect host gene expression, physiology, growth and development. It can be predicted that for each infection two main factors contribute to these effects: (i) the action of a specific virus product(s) (protein or nucleic acid); and/or (ii) changes associated more generally with virus replication as a drain on cellular resources. Despite the obvious importance of these factors in disease, our understanding of the basic mechanisms is very limited. It is true that there are many examples of the identification of the virus determinants of symptoms, even down to single nucleotides, but these have rarely been followed by in-depth analyses of how these determinants influence the host to achieve symptom expression. Similarly, it is only in recent years that attempts have been made to understand the basis of changes in host physiology associated with virus multiplication. The speakers in this Symposium have been selected as some of those who have taken on the challenge of the 'how?' and 'why?' questions in the subtle but complex interaction that leads to symptom expression.

The metabolic diversion associated with virus replication can be enhanced through the down-regulation of cellular genes, an effect first identified for animal virus infections and called host gene 'shut-off' [1]. We have been studying plant-virus infections to see if similar effects occur and what the long-term influence of those effects might be.

Materials and methods
The nature of progressive infections in plant tissues means that we had first to identify the precise sites of virus replication and to use techniques that allowed us to identify transient events associated with replication cycles that may last no more than a few hours. In the absence of appropriate cell culture systems from plants, we have studied intact infected tissues using cell biological techniques that provide 'snap-shots' in time of the dynamic interaction between host and virus. We have used virus (predominantly pea seedborne mosaic potyvirus; PSBMV) infection of pea tissue and applied in situ hybridization to the front of virus invasion to record changes in host gene expression.

Results and conclusions
In pea, PSBMV replication is restricted to a narow zone of cells at the infection front. Following the onset of virus replication, host gene shut-off is seen as a loss of most mRNAs. The shut-off is transient, after which host gene expression returns to a near-normal state, although the cells still contain large amounts of virus and virus-specific products [2]. The diversity of transcripts affected points to a mechanism of mRNA degradation. Exceptions to this down-regulation have been identified in the coordinate induction of two heat-inducible genes (HSP70 and polyubiquitin). This induced state is more transient than shut-off and is visible in only a few cells behind the infection front [3]. Associated with both of these changes is a loss of endoplasmic reticulum. This does not, however, reflect a loss of translatability since both viral RNA and the induced mRNAs are translated in newly infected cells. In more recent experiments, we have tested the scope of these responses for different viruses and different plant tissues, and tested whether the induction reflects a non-specific (heat-shock-related) stress response. The results of these experiments will be discussed together with the potential of these transient but dramatic changes to have a longer term influence in the tissue that could lead to symptom expression.

1. Aranda MA, Maule AJ, 1998. Virology (in press).
2. Wang D, Maule AJ, 1995. Science 267, 229-231.
3. Aranda MA, Escaler M, Wang D, Maule AJ, 1996. Proceedings of the National Academy of Sciences, USA, 93, 15289-15293.