OBJECTIVE 1: EXPRESSION ANALYSIS OF ARABIDOPSIS WRKY GENES
A. Expression profile of the Arabidopsis WRKY gene superfamily during plant defense response (Dong et al, 2003, Plant Mol. Biol. 51:21-37).
In this comprehensive expression analysis, we used northern blotting to examine transcript levels of 72 Arabidopsis WRKY genes in plants after infection with an avirulent strain of the bacterial pathogen Pseudomonas syringae or treatment with salicylic acid (SA). Expression of these 72 WRKY genes in roots, young seedlings and mature plants were also examined. In addition, we have analyzed expression of the pathogen- or SA-induced WRKY genes in Arabidopsis mutants defective in SA accumulation or signaling. Major findings from this comprehensive expression analysis include:
1. 49 of the 72 WRKY genes were differentially regulated in the plants infected by the avirulent strain of P. syringae or treated by SA. Induction of many of these WRKY genes was rapid, strong and transient. Generally speaking, pathogen infection induced higher levels of the WRKY genes than SA treatment. A number of WRKY genes were induced only by pathogen infection but not by SA while several other WRKY genes (e.g. WRKY38 and WRKY62) were induced by SA but not by pathogen infection.
2. Pathogen-induced expression of many of the WRKY genes examined occurred normally in the npr1 mutant defective in SA signaling. A majority of pathogen-regulated WRKY genes were also induced by P. syringae in the transgenic nahG plants that do not accumulate SA. These results indicated that pathogen-induced expression of many of the Arabidopsis WRKY genes is SA-independent. By contrast, SA-induced expression of many of the WRKY genes was either abolished or reduced in the npr1 mutant plants. These results indicated that pathogen- and SA-induced expression of many of the Arabidopsis WRKY genes is mediated through different signaling pathways.
3. A number of WRKY genes (e.g. W14, W35, W36, W69 and W72) were expressed almost exclusively in roots. In addition, many of pathogen- and SA-induced WRKY genes also exhibited developmentally regulated expression patterns in healthy, uninfected plants with a majority of them exhibiting low basal levels expression in young seedlings but higher basal expression levels in mature plants.
B. Expression of pathogen-induced WRKY genes in eds and pad mutants (Figure 1 PDF).
In Arabidopsis, a number of defense-defective eds and pad mutants have been isolated. Although some of these mutant genes (e.g. eds1, eds5 and pad4) have been cloned, a majority of them are not well characterized and the molecular basis for their defective defense mechanisms is unclear. Likewise, although many pathogen-induced WRKY genes have been identified, their specific roles in plant defense responses are yet to be demonstrated. To provide new insights into these defense-defective mutants and defense-regulated WRKY genes, we have analyzed expression of 29 pathogen-induced Arabidopsis WRKY genes using quantitative RT-PCR in the Arabidopsis eds and pad mutants after infection with a virulent strain of P. syringae. Major findings from this expression analysis include:
1. In a majority of the eds and pad mutants, there were diverse expression patterns of the pathogen-induced WRKY genes. Many of these mutants had decreased expression for some WRKY genes, but increased expression for other WRKY genes. In addition, many mutants had altered WRKY gene expression patterns in a time-dependent manner with enhanced expression of some WRKY genes at one time point but decreased expression at another time point after the pathogen infection. Some mutants including eds1 and eds4, however, had substantially reduced expression for a majority of the 29 tested WRKY genes at the three time points tested (Figure 1 PDF).
2. Expression patterns of the 29 pathogen-induced WRKY genes in the eds and pad mutants were also quite diverse but could be classified into two loosely connected groups. The first group (located in the upper part of Figure 1 PDF) had both reduced basal levels of expression and reduced induction of expression at 2 and 8 hours post inoculation in a majority of the eds and pad mutants. The second group of WRKY genes (located in the lower section of Figure 1 PDF) displayed an overall increase in their transcript levels in a majority of the eds and pad mutants at one or more tested time points. WRKY23, WRKY26 and WRKY48 exhibited particularly strong patterns of enhanced expression in these mutants at the three tested time points. Since higher levels of transcripts for these genes were observed even in uninfected healthy eds and pad mutant plants (Figure 1 PDF), it is unlikely that the increased expression of these three WRKY genes was resulted from a more severe disease state in the mutants.
In addition, we have analyzed a number of WRKY genes for pathogen-induced expression in the jasmonic acid (JA)-insensitive coi1 mutant and ethylene (ET)-insensitive ein2 mutant. We have also analyzed induced expression of a number of Arabidopsis WRKY genes in responses to SA, JA and ET. These studies revealed that expression patterns of different WRKY genes are quite diverse and may be regulated through a number of distinct pathways of plant defense responses.
OBJECTIVE 2: FUNCTIONAL ANALYSIS OF ARABIDOPSIS WRKY GENES USING BOTH LOSS-OF-FUNCTION AND GAIN-OF-FUNCTION APPROACHES.
A. Overexpression of Arabidopsis WRKY genes in transgenic plants (Table 2 PDF).
We have generated constitutive or inducible overexpression constructs for more than 40 expressed Arabidopsis WRKY genes using the CaMV 35S promoter or the steroid-inducible Gal4 promoter. We are generating data from overexpression lines in the following two areas. First, many of overexpression lines exhibited altered growth, development and morphology. For example, overexpression lines for WRKY18, WRKY40 and WRKY48 have significantly stunted plants and more serrated leaves. These plants also displayed delayed flowering. On the other hand, overexpression lines for other WRKY genes such as WRKY7, WRKY25 and WRKY26 flowered earlier than wild type plants. It is unclear whether these WRKY genes affect plant growth and development directly or indirectly through affecting expression of genes involved in plant stress responses -- A number of studies have recently shown that plant stress can usually promote plant reproductive growth such as flowering.
Second, we have analyzed transgenic plants that overexpress Arabidopsis WRKY genes for response to the hemibiotrophic bacterial pathogen P. syringae and the necrotrophic fungal pathogen B. cinerea. These studies have revealed that overexpression of some of the WRKY genes tested altered plant responses to these pathogens. For example, overexpression of WRKY18 leads to enhanced resistance to P. syringae while overexpression of several other WRKY genes caused enhanced susceptibility to the bacterial pathogen. Interestingly, overexpression of some of the Arabidopsis WRKY genes has opposite effects on these two pathogens, indicating that plant defense mechanisms against different types of pathogens may have antagonistic relationship.
B. Loss-of-function mutants for Arabidopsis WRKY genes (Table 3 PDF)
We have isolated T-DNA insertion or transposon-tagged mutants for about ~60 Arabidopsis WRKY genes. A majority of these mutants were identified from the Salk T-DNA insertion populations. Other mutants were identified from the Syngenta T-DNA population, as well as from other knockout populations in Germany, France and Japan.
We have screened knockout mutants for ~50 WRKY genes for altered responses to two pathogens, the bacterial pathogen P. syringae and the fungal pathogen B. cinerea. Our screens of the loss-of-function mutants for response to P. syringae have focused on symptom development, pathogen-induced PR gene expression and growth of virulent strains of the bacterial pathogen. These analyses have identified a number of WRKY genes with roles in plant responses to the bacterial pathogens. Our screens of the loss-of-function mutants for responses to the fungal pathogen B. cinerea have so far identified a number of WRKY genes that may play roles in responses to the fungal pathogen. Like some of the overexpression lines, mutations of some Arabidopsis WRKY genes had opposite effects on P. syringae and B. cinerea. These results support the existence of multiple competing signaling pathways in plant defenses. It appears that action of several WRKY genes promote activation of defense pathways effective to one type of pathogens but antagonize defense against another type of microbial pathogens.
Because the number of loss-of-function mutants for WRKY genes with detectable phenotypes is still very low, there might be extensive functional redundancy among structurally related Arabidopsis WRKY proteins. To overcome the functional redundancy, we are in the process of generating double and triple mutants for genes encoding structurally related WRKY proteins. Once the double and triple knockout mutants are generated, they will be examined for altered growth, development and disease resistance to both bacterial and fungal pathogens.
C. Functional analysis of genes regulated by plant WRKY transcription factors.
In Arabidopsis, there is a family of receptor-like protein kinases (RLKs) containing novel cysteine-rich repeats in their extracellular domains. Genes encoding many of these cysteine-rich RLKs (CRKs) are induced by pathogen infection, suggesting a possible role in plant defense responses. We have shown that many of these pathogen-induced CRK genes contain a large number of TTGACC/T W-box sequences in their promoters that are recognized by plant WRKY proteins. Mutations of the W boxes abolished inducible expression of the CRK genes, indicated that that these pathogen-induced CRK genes are target genes regulated by WRKY transcription factors. To study the roles of pathogen-induced CRKs in plant defense responses, we have generated Arabidopsis plants expressing four pathogen-regulated CRK genes (CRK5, 6, 10 and 11) under control of a steroid-inducible promoter and found that induced expression of CRK5, but not the other three CRK genes, triggered hypersensitive response-like cell death in transgenic plants. We have further analyzed the structural relationship of the CRK family and identified three CRKs (CRK4, 19 and 20) that are structurally closely related to CRK5. Genes encoding these three CRKs are all induced by SA and pathogen infection. Furthermore, induced expression of CRK4, 19 and 20 all activates rapid cell death in transgenic plants. Thus, the activity of inducing rapid cell death is shared by these structurally closely related CRKs. We have also performed yeast two-hybrid screens and identified proteins that interact with the kinase domains of CRKs. One of the identified CRK-interacting proteins is the kinase-associated type 2C protein phospohatase known to interact with a number of other RLKs through its kinase-interacting FHA domain. Other CRK-interacting proteins include a second protein with a FHA domain and another type 2C protein phosphatase. Interactions of CRKs with these three proteins in vivo were demonstrated through co-immunoprecipitation. These CRK-interacting proteins may play roles in the regulation and signaling of CRKs.
OBJECTIVE 3. IDENTIFICATION OF WRKY-INTERACTING PROTEINS
We have been using the yeast two-hybrid screens to identify proteins that interact with Arabidopsis WRKY proteins. We initially used the well-established Gal4 system. As a first step, we generated the fusion constructs for about 45 Arabidopsis WRKY proteins with the DNA-binding domain of the Gal4 transcription factor. The fusions were then tested in yeast for Gal4-regulated expression of a reporter gene used in the two-hybrid screens. These assays revealed that a majority of Arabidopsis WRKY proteins could activate transcription in yeast cells. This property makes the Gal4 yeast two-hybrid system unsuitable for identifying WRKY-interacting proteins. For a few WRKY proteins with no transcriptional activation activity in yeast, we have used the system and identified a number of proteins that interact with Arabidopsis WRKY proteins. For example, we have found that Arabidopsis WRKY18 is able to interact with itself and forms homo-complex.
Because of the problem associated with the Gal4 two-hybrid system, we have decided to use the recently developed Cytotrap yeast two-hybrid screens for identification of proteins that interact with Arabidopsis WRKY proteins. As a first step, we have generated an Arabidopsis Cytotrap library using RNA isolated from salicylate-treated Arabidopsis leaves. Genes encoding WRKY proteins with transcriptional activation activity in yeast have also been cloned into the bait plasmid of the Cytotrap system and screened for positive clones. These screens have identified interacting proteins for a number of Arabidopsis WRKY proteins. The identified interacting proteins include transcription factors, putative chromatin-remolding factors and other regulatory proteins. Confirmation of the interactions for some of these identified positive candidates has also been performed using in vitro pull down assays and in vivo immunoprecipitation. During the remaining funding period of the project, a major effort will be taken to identify interacting proteins for additional WRKY proteins from Arabidopsis. A selected number of these interacting proteins are also being analyzed for their roles in WRKY-mediated biological processes.