1.1.4S
NEW APPROACHES TO FINDING BACTERIAL AVR PROTEINS AND ASSESSING THEIR ROLE IN PROMOTING PARASITISM

A COLLMER1, JR ALFANO2, M BALLANTYNE1, AO CHARKOWSKI1, W-L DENG1, DE FOUTS1 and JH HAM1

1Department of Plant Pathology, Cornell University, lthaca, NY 14853-4203, USA; 2Department of Biology, University of Nevada, Las Vegas, NV 89154-4004, USA

Background and objectives
The hrp genes of Pseudomonas syringae and other common bacterial pathogens of plants encode a type III protein secretion pathway that can deliver Avr (avirulence) proteins into plant cells [1]. Although named for their ability to cause virulent pathogens to become avirulent on host plants carrying a cognate resistance (R) gene, some avr genes contribute measurably to bacterial virulence (in hosts lacking a cognate R gene), and collectively, avr genes appear to be required for the parasitic and pathogenic abilities of these bacteria. This is based on the observation that hrp mutants, which cannot transfer Avr proteins, typically fail to multiply in planta, to elicit the defence-associated hypersensitive response (HR) in non-hosts, or to produce disease in hosts [1]. Thus finding avr genes and determining the function of their products in planta is key to understanding bacterial parasitism. A priori, these functions are likely to involve nutrient release to the apoplast and suppression of host defences. Our objectives are to develop methods to: (i) find genes encoding candidate Avr-like effector proteins without the need for an avirulence test (which is dependent on the unpredictable presence of cognate R genes in test plants) or a virulence test (which may fail to detect a phenotype with redundant effectors); (ii) assay rapidly for a parasitic growth benefit of candidate effector proteins expressed in planta, and (iii) assay for suppression of defences by these proteins expressed in individual plant cells. Our efforts are focused on two well-studied strains, P. syringae pv. tomato (Pst) DC3000 (a pathogen of tomato and Arabidopsis) and P. syringae pv. syringae (Pss) B728a (a model epiphyte and pathogen of bean).

Results and conclusions
Several candidate effector protein genes were found by sequencing the DNA regions flanking the hrp clusters of Pst DC3000 and Pss B728a. Unlike the hrp genes, the genes in the region beyond hrp1 were found to differ strikingly, even among strains of Pss with virtually identical hrp sequences. A system for secreting P. syringae Avr proteins was developed. Cosmid pCPP2156, which encodes a cluster of Erwinia chrysanthemi hrp genes, was found to enable Escherichia coli to elicit an HR in Nicotiana clevelandii and soybean cultivar Norchief (but not Acme), but only if avrB was also expressed by the bacterium. E. coli (pCPP2156), but not an hrp mutant derivative, efficiently and selectively secreted AvrB and AvrPto (tagged with the FLAG epitope [2]) to the culture medium, whereas neither of these proteins was secreted by E. coli (pHIR11), which carries a functional cluster of Pss 61 hrp genes. This system can be used to identify candidate effector protein genes for subsequent in planta expression studies. avr genes can be individually and transiently expressed within plant cells (and at levels that can be regulated to better study virulence effects) following delivery by Agrobacterium tumefaciens. We have determined that A. tumefaciens does not grow within Arabidopsis leaves unless co-inoculated with a virulent pathogen, such as Pst DC3000. Thus, A. tumefaciens cells delivering candidate Pst effector protein genes (or mixed cultures delivering combinations thereof) can be assayed directly for general bacterial growth stimulation. The Arabidopsis ELI3 gene is rapidly induced by bacteria expressing avrB and provides a marker for AvrB-dependent defence elicitation. To monitor defence elicitation in individual plant cells, we have constructed transgenic Arabidopsis ColO plants in which the green fluorescent protein gfp gene is expressed from the EL13 promoter. Effector proteins now can be assayed for their ability to suppress AvrB-dependent GFP expression in individual plant cells. These approaches are directed at finding those effector proteins most important for parasitism.

References
1. Alfano JR, Collmer A, 1997. Journal of Bacteriology 179, 5655-5662.
2. Gopalan S, Bauer DW, Alfano JR et al., 1996. Plant Cell 8, 1095-1105.