NSF Arabidopsis 2010
Expression Profiling of Plant Disease Resistance Pathways Project
Xinnian Dong Lab   Frederick Ausubel Lab   Shauna Somerville Lab
IMDS Database
  User Name

Overview of the 2010 Project

The overall goal of this 2010 project is to elucidate the roles of transcription factors (TFs) in the Arabidopsis defense-response signaling network. The focus will be on TFs because transcriptional profiling is the most powerful genomics approach available for high through-put analysis and because TFs represent key nodes in signaling networks. The first aim is to identify candidate TFs in specific defense response pathways, including SA-dependent but NPR1-independent signaling and the response to pathogen-associated molecular patterns (PAMPs). The second aim is to use forward and reverse genetic approaches to identify key signature TFs in each pathway.  The third aim is to construct GFP or GUS reporters for key signature TFs to identify and sort cells expressing these TFs.  The fourth aim is to carry out transcriptional profiling that will be conducted in sorted cells expressing a particular TF to identify genes whose expression patterns mirror those of the selected TFs (i.e., cohort transcriptomes). TFs will also be fused to a glucocorticoid receptor domain so that nuclear transport of individual TFs can be regulated by DEX with the goal of identifying direct transcriptional targets of individual TFs. Collectively, these experiments will provide spatial, temporal, and hierarchical information, which is currently lacking but needed to build a model of the transcriptional network of interacting defense pathways.

In the first funding period, we performed a large number of expression profiling experiments using Affymetrix GeneChips™ to identify Arabidopsis defense-related genes following infection with a variety of obligate biotrophic pathogens and necrotrophic pathogens. These array data can be viewed at the IMDS (Integrated Microarray Data System) tab on this website.  The array data were used by our collaborators Drs. Fumiaki Katagiri and Jane Glazebrook at the University of Minnesota to construct a small-scale microarray printed with 576 long oligonucleotide probes. The high reproducibility of this pathoarray in measurement of modest numbers of mRNAs in large numbers of samples has recently been reported (Sato et al. 2007. The Plant Journal 49, 565).

To facilitate interactions between laboratories and minimize overlapping research, we are in the process of generating a combined list of TFs that are being studied in our laboratories (see Table).  We are also constructing a list of marker genes for each defense signaling pathway (Table coming soon).  These lists will be updated and publicized through TAIR periodically.

Projects in Xinnian Dong's Lab

*Identification of transcription factors involved in the early steps of SAR  

In order to investigate the expression profile of both basal and R-gene mediated resistance, plants were infected by pressure infiltration with Pseudomonas syringae pv maculicola (Psm) ES4326 and an isogenic strain carrying a plasmid containing the avirulence gene avrRpt2. Samples were collected at five different time points, 4, 8, 16, 24, and 48 hours after infection, and prepared for expression analysis using the Affymetrix ATH1 GeneChip. Therefore, the expression pattern of each gene was examined as the infection and the defense response progressed. In order to enrich the samples for expression changes that are involved in disease resistance, only the uninfected tissue adjacent to the infection site was collected for microarray analysis. There is strong induction of PR genes in this tissue. In addition, control samples infiltrated with 10 mM MgCl2 were analyzed at every time point to identify the genes induced by the infiltration alone. Two biological replicates of the experiment were performed. 

There are 1075 genes that showed a 2-fold change after at least one type of Psm infection when compared to the MgCl2 control at two or more consecutive time points. As observed by others, a similar set of genes showed changes in expression level after infection with both virulent and avirulent bacteria. However, the expression changes were stronger and occurred sooner after avirulent infection when compared to virulent infection. 

We found that over 30 genes predicted to have transcription factor activity are induced by either virulent or avirulent infection. These factors fall into about 6 different transcription factor families. In many cases, two or more members of each class have nearly identical expression patterns. We predict that mutant phenotypes may only appear after knocking-out all of the transcription factors within a family that show similar expression patterns after infection. We currently have homozygous lines for 17 of these transcription factors. We have tested several single, double, triple and quadruple mutants within the WRKY and NAC families.

*Salicylic acid inhibits pathogen growth in plants through repression of the auxin signaling pathway 

The phytohormone auxin regulates almost every aspect of plant development.  At the molecular level, auxin induces gene expression through direct physical interaction with the TIR1-like F-box proteins which in turn remove the Aux/IAA family of transcriptional repressors.  A growing body of evidence indicates that many plant pathogens can either produce auxin themselves or manipulate host auxin biosynthesis to interfere with the normal host developmental processes.  In response, plants likely evolved mechanisms to repress auxin signaling during infection as a defense strategy.  Plants over-accumulating the defense signal molecule salicylic acid (SA) frequently display morphological phenotypes that are reminiscent of auxin-deficient or insensitive mutants, indicating that SA may interfere with auxin homeostasis.  Using the Affymetrix ATH1 GeneChip (24,000 genes) for Arabidopsis, we performed a comprehensive study of the effect of SA on auxin signaling.  We found that SA causes a global repression of auxin-related genes, including the receptor TIR1, resulting in stabilization of the Aux/IAA repressor proteins and inhibition of auxin responses.  We demonstrate that this inhibitory effect on auxin signaling is part of the SA-mediated disease resistance mechanism. 

* Transcriptional profiling of Arabidopsis in responses to Hyaloperonospora parasitica infection

 Hyaloperonospora parasitica is a natural biotrophic pathogen of Arabidopsis.  Different isolates of H. parasitica have been found that trigger various responses on Arabidopsis.  Transcriptional profiling for this pathogen infection has not been done thoroughly using the Arabidopsis whole genome chip.  To fill this information gap, we chose to profile gene expression in Arabidopsis Col-0 upon H. parasitica Emwa1 infection. There are two important factors to be considered for the study of H. parasitica infection.  First, H. parasitica infection is asynchronous.  Secondly, the timing of the HR and cessation of pathogen growth is slightly different for each resistance response.  Therefore, the same isogenic host (Col-0 background) and pathogen (Emwa1 isolate) were chosen for this study. The Emwa1 isolate causes the HR on wild type Col-0.  This resistance response is mediated by the RPP4 gene.  We identified the rpp4 knockout mutant in the T-DNA population and showed that it became completely susceptible to Emwa1 infection. Earliest HR symptoms were seen at 12 hpi during incompatible interaction.  By 48 hpi, the wild-type showed induction of PR1 expression and RPP4-mediated HR, which is characteristic of an incompatible interaction, while the rpp4 mutant started to display hyphal growth (Fig. 1).  Therefore, based on gene expression and progression of disease symptoms, microarray experiments were carried out at 0, 12, 48, 96 hpi, and 6 dpi after Emwa1 inoculation. Because both the host and pathogen are isogenic, differences in background transcription are minimized. A mixed-model analysis of variance was performed to identify genes differentially expressed between the various treatments (Levesque et al., 2006).  Eighty genes showed significantly elevated expression (>2-fold, q< 0.01) in the wild-type compared to the rpp4 mutant at 48 hpi (Fig. 2) and were chosen for genetic analysis.  So far, knock out lines for 31 candidate genes have been infected with H. parasitica Ewma1, and 9 of them showed different degrees of susceptibility to this pathogen which is completely avirulent on WT.

*Comprehensive study of Arabidopsis Lypoxygenases

 Lypoxygenases (LOX) are important enzymes involved in plant development and plant defense. LOX are the precursors of oxylipins, which can play a role as signal molecules (e.g. Jasmonates) and as toxic compounds (e.g. aldehydes and divinyl ethers). The Arabidopsis genome encodes six different LOX. We ordered T-DNA insertion lines for each LOX. We created several double mutants to overcome functional redundancy among the 13 LOX family, or to knock out the 9 LOX pathway to study their effects on pathogen resistance and plant development.

 *Genetic interactions of TGA transcription factors in the regulation of pathogenesis-related genes and disease resistance

TGA transcription factors are implicated as regulators of pathogenesis-related (PR) genes because of their physical interaction with the known positive regulator, NPR1. A triple knockout mutant tga2-1 tga5-1 tga6-1 was shown previously to be defective in the induction of PR genes and systemic acquired resistance (SAR), confirming their role in disease resistance. However, the contributions of individual TGA factors have been difficult to discern because of functional redundancy among these factors as well as possible dual functions for some single factors. In the present study, we characterized six TGA factors by reverse genetics.  We show that TGA3 is required for both basal and 2,6-dichloroisonicotinic acid (INA)-induced transcription of PR genes. The tga3-1 mutants were found to be defective in basal pathogen resistance, while the induced resistance was unaffected. TGA1 and TGA4 play partially redundant roles in regulation of basal resistance, having only moderate effects on PR gene expression. Additionally, an activation-tagged mutant of TGA6 was able to increase basal as well as induced expression of PR-1, demonstrating a positive role for TGA6 on PR gene expression. In contrast, TGA2 has a repressor activity on PR gene expression even though it can act as a positive regulator in the tga5-1 tga6-1 null mutant background. Finally we examined the genetic interaction between tga2-2 and the suppressor of npr1 inducible 1 (sni1-1). TGA2’s repressor activity overlaps with SNI1, because the tga2-2 sni1-1 double mutant shows a synergistic effect on PR gene expression.

* WRKY transcription factors differentially regulate NPR1-mediated plant defense responses

In Arabidopsis, the onset of systemic acquired resistance (SAR) is preceded by accumulation of the signaling molecule salicylic acid (SA).  SA induces nuclear translocation of the transcription cofactor NPR1 to activate a myriad of genes required for resistance.  NPR1 also feedback regulates SA synthesis to mitigate its cytotoxic effect.  It is known that NPR1 controls expression of antimicrobial pathogenesis-related (PR) genes by interacting with TGA transcription factors. Even though WRKY transcription factors have also been implicated to play a role in SAR-related transcription, a direct genetic link between a WRKY factor and SAR has been lacking because of the difficulties involved in studying this large gene family (74 genes).  To dissect the transcription cascade leading to SAR, we performed a microarray experiment using the Affymetrix ATH1 GeneChips (24,000 genes) to identify direct transcriptional targets of NPR1.  By using a transgenic line expressing NPR1-glucocorticoid receptor (GR) fusion (in npr1-3), whose nuclear translocation is under the control of dexamethasone, we were able to specifically activate NPR1-dependent transcription and, at the same time, inhibit the translation of nascent mRNA by addition of cycloheximide.  Among the direct targets of NPR1, we identified eight WRKY (WRKY 18, 38, 53, 54, 58, 59, 66, and 70) transcription factor genes whose expression was reproducibly induced by NPR1. Single insertion mutants were identified in these WRKY genes and double mutants.  To these mutants, we tested various phenotypes associated npr1, namely: a lack of SA-induced and pathogen-induced resistance, enhanced disease symptoms (EDS), and a lack of feedback regulation of SA biosynthesis.  We found that WRKY18 plays a positive role in SAR as the wrky18 deficient mutant showed impaired SAR and enhanced disease susceptibility to a bacterial pathogen.  In contrast, WRKY58 responds to suboptimal levels of SA to prevent spurious activation of SAR.  NPR1’s ability to negatively regulate the levels of SA is carried out by WRKY70 and WRKY54; knocking out both genes caused SA to accumulate to abnormally high levels. WRKY70 is also required for SA-mediated PR gene expression and resistance. Even with elevated levels of SA, wrky70, in combination with wrky53, shows enhanced susceptibility to infection. This genomics-directed genetic study allowed us to place specific WRKY factors in the intricate transcriptional regulatory network underlying the onset of SAR.

 *Characterization of NAC transcription factors induced during SAR

 In our SAR transcription profiling experiment we found over 30 transcription factor genes induced in systemic tissue by either virulent or avirulent infection.  Besides the WRKY TF described above, NAC transcription factors are another major family of TFs in this group.  We are characterizing T-DNA insertion mutants in these TF genes.

* Identification of TL1 element-binding TF(s) involved in early plant defense

Accumulation of unfolded proteins in ER causes stress and activates a signaling network called the unfolded protein response (UPR), a ubiquitous mechanism observed in all eukaryotic organisms from humans to yeast to plants, which relays signals from the ER lumen to activate target genes in the nucleus. Thanks to support of the NSF 2010 project, we have previously carried out microarray analyses of NPR1-dependent genes and identified a cis-regulatory element designated TL1, which coordinates the expression of ER-resident genes involved in early plant defense responses in an NPR1-dependent manner (Wang et al. Science 2005). We found that mutations in these genes, e.g. BiP2 (At5g42020) causes dramatic collapse of proper protein folding in the ER and leads to a plant-specific form of the UPR in the form of cell death (Wang et al. Science 2005). The identification of TL1-mediated induction of the plant secretory pathway defined a novel pathway in plant immunity; however, the TL1-binding TF(s) is yet to be identified. We are currently pursuing two complementary approaches in order to reveal the identity of TL1-binding factor(s): yeast one-hybrid screen using the TL1 element-enriched BiP2 promoter fragment as bait against an Arabidopsis cDNA library derived from healthy and infected leaf tissue. (b) candidate gene approach.

 * Transcriptome analyses on the cells specifically responsive to pathogen infection

Microarrays are being used to comprehensively study gene expression networks during the plant defense response that is triggered when a plant cell encounters a pathogen. However, conventional transcriptional profiling typically lacks precise temporal and spatial information as tissues collected after infection often contain cells at various infection stages along with naïve cells, making it difficult to distinguish cascades of signaling events. Transcriptional profiling on cells specifically responsive to a pathogen will lead to the identification of genes with similar expression patterns (‘cohort transcriptomes’ or regulons). Here, we are making use of an innovative strategy to capture the pathogen-responsive cells from infected transgenic plants expressing a fluorescence marker under the control of a pathogen inducible promoter, using flow cytometry/fluorescence-activated cell (nuclei) sorting (FACS). We have generated six reporter lines carrying transcriptional fusions of promoters of pathogen-responsive genes We will use selected transgenic lines for microarray experiments to identify groups of genes with similar expression patterns (‘cohort transcriptomes’ or regulons). Subsequently, novel cis-regulatory elements of plant immunity can be deduced by promoter analysis of the newly identified ‘cohort’ member genes, and the TFs binding to these elements can be discovered, following the successful strategy of Wang et al. (2005).

Projects in Frederick Ausubel's Lab

Defining common and unique components of responses to different PAMP (MAMP) elicitors

Expression profiling of 10-day old Arabidopsis seedlings treated with oligogalacturonides (OGs) or the 22 amino acid flagellin peptide Flg22 for 1 or 3 hours revealed a high degree of similarity in the early response (at 1 hour) to these two disparate molecules, with significant divergence in the later response (3 hours). We are defining the common and unique aspects of these induced responses. Both elicitors trigger comprehensive transcriptional reprogramming of defense-associated genes at the early time point: key regulatory genes such as EDS1, PAD4, NPR1, and RIN4 are induced, as are LOX4, ACS7, EDS5, and AtPCb, involved in generating, respectively, the signaling molecules jasmonic acid (JA), ethylene (Et), salicylic acid (SA), and H2O2. Of these, the most highly induced is LOX4, which is expressed at more than 100-fold higher levels after elicitor treatment. Consistent with early activation of the JA signaling pathway, several WRKY transcription factors, which have been reported by other groups to positively regulate JA-dependent gene expression, are also induced after 1 hour of elicitor treatment. At later times (3 hours), both elicitors induce enzymes involved in indole secondary metabolism, including genes with a role in the biosynthesis of indole glucosinolates and camalexin, and phenylpropanoid metabolism. Other groups have reported that enzymes for the production of indole glucosinolates are up-regulated by methyl jasmonate, pointing to a role for jasmonates in the early activation of responses that are common to both elicitors. Both elicitors also induce higher level expression of the JA/Et pathway signaling molecule, ERF1, which positively regulates expression of the JA- and Et-dependent defense-related gene PDF1.2. However, PDF1.2 is not strongly up-regulated by either elicitor.

Interestingly, functional classification of genes that are induced predominantly at the 3 hour time point by Flg22 indicates that Flg22 but not OGs activates secretory processes and a senescence program. Induction of the secretory pathway during systemic acquired resistance is dependent on SA activation of the signaling molecule NPR1. Of 18 NPR1-dependent secretory pathway genes, 9 are induced more strongly or exclusively by Flg22. Induction by Flg22 but not OGs has been confirmed by RT-qPCR for the most highly regulated of these secretory pathway genes. Additionally, Flg22 but not OGs induces high-level expression of the SA response regulator WRKY70 and the SA-dependent pathogenesis-related protein PR1. Together, these data indicate that Flg22 but not OGs activates SA-dependent defense responses. Down-regulation of photosynthesis-associated genes by Flg22 suggests activation of a senescence program specifically by Flg22 but not OGs. Both SA and Et have been implicated in the regulation of senescence. An isoform of 1-aminocyclopropane-1-carboxylate synthase, the rate-limiting step in ethylene biosynthesis, is induced strongly by Flg22 but not by OGs. This isoform is strongly anti-correlated with genes encoding chloroplast-associated proteins, suggesting that it mediates ethylene production during defense-related senescence. Finally, we have found that Flg22 induction of PR1 is dependent not only on SA synthesized via isochorimate synthase, but also on ethylene signaling, as induction is severely diminished in ics1 or ein2 mutant plants.

Together, these data suggest that both OGs and Flg22 induce an early response that involves jasmonic acid and ethylene, but that only Flg22 triggers a later response that is orchestrated by Et and SA. This work is currently (as of July 2007) being prepared for submission to Plant Physiology (C. Denoux, S. Ferrari, S. Gopalan, R. Galletti, G. De Lorenzo, F. M. Ausubel, J. Dewdney. OGs and Flg22 elicitors differentially activate defense-signaling pathways.)

Characterization of functional role of Arabidopsis oxidative burst peroxidases

An early response in plants upon recognition of potential pathogens is the generation of reactive oxygen species. In the past year, we have, in collaboration with Paul Bolwell and his group at Royal Holloway, University of London, published the results of our joint studies that demonstrate a role for Arabidopsis apoplastic peroxidases in both the production of ROS following elicitor or pathogen challenge and in limiting growth of virulent and avirulent bacterial pathogens and virulent fungal pathogens (L.V. Bindschedler, J. Dewdney, K.A. Blee, J.M. Stone, T. Asai, J. Plotnikov, C. Denoux, T. Hayes, C. Gerrish, D.R. Davies, F.M. Ausubel, G.P. Bolwell 2006 Peroxidase-Dependent Apoplastic Oxidative Burst in Arabidopsis Required for Pathogen Resistance. The Plant Journal 47:851-863). Based on expression profiling of the heterologous antisense line H4, we also identified the Arabidopsis candidates, peroxidases AtPCa (At3g49110) and AtPCb (At3g49120), most likely to fulfill this function (Bindschedler et al., 2006).

We are currently extending our characterization of the role of these two apoplastic peroxidases. Preliminary data on susceptibility of H4 plants to a non-host pathogen, Pseudomonas syringae pv phaseolicola strain NPS3121 (Psp NPS3121), supports the model that AtPCa and/or AtPCb are required for innate immune responses. In addition, expression analysis of selected genes following inoculation of H4 plants with Pseudomonas syringae pv tomato DC3000/avrRPM1 (Pst DC3000/avrRpm1) indicates that the expression of a number of pathogen-inducible genes is partially dependent on AtPCa and/or AtPCb.  To elucidate the role of each of these two peroxidase isoforms, insertion mutants that specifically disable either AtPCa or AtPCb have been obtained and homozygous lines have been isolated. These mutants are currently the subject of active investigation.

Characterization of MAMP response signal transduction pathway

From expression profiling of seedlings treated with OGs or Flg22 elicitors, we identified approximately 80 transcription factors that are induced after 1 hour with a probability of 0.01 or better and a change of 2-fold or more. Subsequent studies have focused on ten of those transcription factors, including several members of the WRKY and MYB transcription factor families. Induction of these transcription factors by Flg22 and OGs has been confirmed by Northern blot analysis, which also indicated that a majority of these genes are up-regulated within 10 to 20 minutes of exposure to OGs. Mutant (insertion) lines for these ten rapidly-induced genes have been obtained and homozygous insertion lines isolated. Of the eight lines that have been tested, five show reduced expression of the mutated gene relative to wild-type, and three have enhanced transcript levels. In addition, overexpression lines have been generated for nine transcription factor genes, and reporter lines carrying promoter-GUS fusions have been created for eight genes.

As we reported previously, the relevance of several of these transcription factors to defense has been demonstrated by disease susceptibility assays.  Results from several groups have shown that treatment with Flg22 or OGs prior to  subsequent pathogen challenge can enhance resistance to infection by P. syringae or Botrytis cinerea, respectively. Interestingly, lines with insertions in particular WRKY TFs showed enhanced protection against a virulent bacterial pathogen relative to wild-type plants when inoculation was preceded by an 8-hour incubation with OGs. Similarly, a line with an insertion in a Myb TF exhibited enhanced Flg22-induced resistance. Conversely, insertions in other TF genes result in a deficient Flg22-induced resistance response. These results suggest that WRKY and other TFs function as either positive or negative regulators of PAMP-induced resistance.

Projects in Shauna Somerville's Lab

[Coming soon. In the meantime, here is her lab website:]


Last updated on 2/17/09.
Department of Molecular Biology • Massachusetts General Hospital • Simches Research Center • 185 Cambridge Street CPZN7250 • Boston, MA 02114-2790
Phone: (617)726-5968 • FAX: (617)726-5949

Contact lim @molbio.mgh.harvard.edu
@2005 Massachusetts General Hospital. All rights reserved.