Department of Molecular Biology, Massachusetts General Hospital and Department of Genetics, Harvard Medical School, Boston, Massachusetts, USA

Background and objectives
To identify the signal transduction pathways leading to the induction of defence genes in compatible pathogen interactions, Arabidopsis thaliana is being used as a model host. Infection of Arabidopsis with a virulent bacterial pathogen leads to the induction of a variety of defence genes, including the pathogenesis-related genes PR1, BGL2 (PR2), and PR5. One of the major signalling pathways leading to the induction of these genes involves the secondary signal molecule salicylic acid (SA); exogenous SA can induce PR1, BGL2 and PR5 and systemic acquired resistance (SAR). The NPR1 gene in Arabidopsis was identified as downstream of the SA signal and npr1 mutants exhibit significant reduction in the induction of PR1, BGL2 and PR5 by exogenous SA, impaired SAR after exposure to avirulent pathogen or exogenous SA, and increased susceptibility to virulent pathogens [1]. However, while npr1 mutants are blocked in SA induction of PR1, BGL2 and PR5, the virulent pathogen Pseudomonas syringae pv. maculicola ES4326 elicits a different response. PsmES4326-inoculated npr1 plants show approximately 10-fold less induction of PR1 compared to wild-type plants, but wild-type levels of BGL2 and PR5 induction [2]. This indicates that there is one or more npr1 and perhaps SA-independent signalling pathway(s) leading to the induction of BGL2 and PR5. To identify components of this pathway, a screen is being conducted for mutants that lose the ability to induce BGL2 in response to virulent P. syringae infection in an npr1 mutant background using a BGL2-GUS (beta-glucuronidase) fusion transgene as a reporter [1].

Results and conclusions
3500 EMS mutagenized M2 plants (parent - npr1 BGL2-GUS) have been screened for reduced GUS activity after infiltration with the virulent pathogen P. syringae pv. tomato DC3000. 173 (5%) putative mutants were selected and re-tested for GUS activity with PstDC3000. 65 (2%) exhibited reduced GUS activity and four (0.1%) of these showed no detectable GUS activity. The four mutants which showed no GUS activity and three mutants which exhibited the lowest GUS activity were tested by Northern blot for induction of endogenous BGL2 in response to PstDC3000 exposure, to determine whether the reduced GUS activity was due to a block in the BGL2 signalling pathway or due to a mutation in the reporter fusion. One (P8) of the four mutants showing no GUS activity showed no induction of endogenous BGL2 after pathogen exposure. This mutant (P8) appears to be blocked specifically in the BGL2 signalling pathway, since PR1 and PR5 showed normal induction after PstDC3000 exposure. A second mutant (P124), which shows no GUS activity, exhibits increased expression of endogenous BGL2 compared to the parental line. Preliminary results with P124 indicate that this expression of the endogenous BGL2 is not constitutive, but induced by pathogen. Genetic and phenotypic characterization of P8, P124 and the other putative mutants is ongoing.

1. Cao H, Bowling SA, Gordon AS, Dong X, 1994. Plant Cell 6, 1583-1592.
2. Glazebrook J, Rogers E, Ausubel FM, 1996. Genetics 143, 973-982.