Plant & Animal Genome VI Conference

S8

CLUSTERS OF RESISTANCE GENES

Department of Vegetable Crops, University of California, Davis, CA 95616

Classical genetics has demonstrated that resistance genes tend to be clustered in the genome. This led to the hypothesis that resistance genes are members of multigene families encoding receptors with individual members of these families having diverged to acquire novel recognition specificities. The cloning of resistance genes for a variety of diseases from diverse plant species is providing increasing support for this hypothesis. The similarities between resistance genes has allowed the facile identification of resistance gene candidates (RGCs) using PCR with degenerate oligonucleotide primers to conserved domains. However, there are hundreds of RGCs in each plant species and these tend to be organized in clusters. Therefore it remains difficult to identify sequences that encode an individual recognition specificity. Complementation and mutant analysis are still required. It is currently unknown how many distinct types of resistance genes there are in plants. The most common types are receptor-like sequences containing a nucleotide binding domain and leucine rich region (NBS-LRR). Sequence analysis indicates the presence of several sub-families of varying complexity within the NBS-LRR type. Classical genetics also indicated that clusters of resistance genes were unstable and that unequal crossing over was involved in the generation of new specificities. This led to the idea that clusters of resistance genes are dynamic regions of the genome in which new specificities are frequently being generated. Characterization of resistance gene clusters at the molecular level confirms that unequal crossing-over can occur. However, the rate of change at these loci seems to be slow. There is little evidence for gene conversion homogenizing these multigene families. Clusters of resistance genes seem to be storehouses of variation rather than dynamic, rapidly-evolving pools of sequences. Genome rearrangements may be instrumental in expressing cryptic recognition specificities as well as generating specificities de novo.