1.1.11
SEPARATING SPECIFICITIES AT THE RPP1 LOCUS IN ARABIDOPSIS THALIANA USING GENETIC RECOMBINATION, ARTIFICIAL MUTATION AND GENE ISOLATION

MPC PINEL, A GORDON, P BITTNER-EDDY, M REDMOND, JL BEYNON and EB HOLUB

Horticulture Research International, Wellesbourne, Warwick CV35 9EF, UK

Background and objectives
There is increasing evidence that disease resistance genes occur as discrete clusters in plants. One of the best examples is the rp1 complex in maize, in which numerous resistance specificities have been characterized genetically in a region of less than 1 cM [1]. Clustering of genes for resistance to Peronospora parasitica (RPP genes) has been shown in Arabidopsis thaliana, but at a larger scale (15 cM). Three isolates of P. parasitica, Emoy2, Hiks1 and Waco5, are recognised by a gene or genes at the RPP1 locus in the accession Niederzenz (Nd-1); incompatibility is characterized by spreading necrosis referred to as pitting [2]. The objective of this study is to determine whether different isolate specificities can be separated in the RPP1 locus using genetic recombination, artificial mutation and gene isolation.

Materials and methods
Three approaches were used. Firstly, phenotypic markers that flank RPP1 (a glabrous mutation, gl1, and the RPP13 locus discriminated by isolate Maks9) were used to select recombination events in the RPP1 region among recombinant inbred (RI) lines from the cross Col-5 x Nd-1. Secondly, Nd-1 seed was subjected to either fast neutron radiation or ethyl methane sulphonate (EMS) mutagenesis. M2 seedlings were screened for susceptibility to isolate Emoy2, and progeny from selected plants were tested for response to the three isolates diagnostic for RPP1. The third approach involved screening an Nd-1 cDNA library for candidate RPP1 genes using sequences designed from conserved regions of known NBS-LRR type resistance genes. Candidate cDNA clones and hybridizing genomic clones are being transformed into susceptible host plants to assign isolate specificities to particular gene sequences.

Results and conclusions
140 recombinant inbred (F9) lines from a population of ca 3000 F2s were selected for crossovers between GL and RPP13. One line shows susceptibility to Emoy2 and resistance to Waco5 and Hiks1. This demonstrates the existence of alleles at at least two loci within RPP1. Of 44,000 Nd-1 mutagen-treated M2 seedlings (derived from 4780 M1 plants) screened with Emoy2, 43 susceptible individuals were selected. These were screened at M3 with all the isolates recognised by the RPP1 locus. Only four of the putative mutants potentially distinguished the three isolate specificities. Candidate cDNAs were shown to hybridize to a multigene family that mapped to the RPP1 locus. Two full-length cDNAs have been sequenced and found to belong to the NBS-LRR class of resistance gene. Analysis of transgenic plants carrying one of these cDNAs under the control of a 35S promoter has been completed, and no functional resistance to any of the three isolates was observed. Generation of additional transgenic plants carrying the other gene is in progress. To complement this approach, one of the cDNAs was used to identify homologous sequences by probing a genomic lambda library. 60 plaques were taken and grouped into two classes. Several of these have been sequenced and cloned prior to being used in transformation experiments.

References
1. Hulbert SH, 1997. Annual Review of Phytopathology 35, 293-310.
2. Holub EB, Beynon JL, Crute IR, 1994. Molecular Plant-Microbe Interactions 7, 223-239.