UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Functional analysis of MAP kinase-dependent elicitor response signaling in tobacco Hall, Hardy Craig 2005

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_2005-0461.pdf [ 11.85MB ]
JSON: 831-1.0092090.json
JSON-LD: 831-1.0092090-ld.json
RDF/XML (Pretty): 831-1.0092090-rdf.xml
RDF/JSON: 831-1.0092090-rdf.json
Turtle: 831-1.0092090-turtle.txt
N-Triples: 831-1.0092090-rdf-ntriples.txt
Original Record: 831-1.0092090-source.json
Full Text

Full Text

Functional Analysis of MAP Kinase-dependent Elicitor Response Signaling in Tobacco by Hardy Craig Hall B.Sc. University of British Columbia, Canada, 2002  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Botany)  THE UNIVERSITY OF BRITISH COLUMBIA June 2005  © Hardy Craig Hall, 2005  Abstract While the M A P kinase, SIPK, is rapidly activated by stress in tobacco (Nicotiana tabacum), its role in mediating stress response outcomes, particularly in its relation to phytohormones and second messengers is unclear. I investigated the tissue and transcriptional responses of tobacco to bacterial and oomycete elicitors, by comparing SIPK-silenced plant, and cell suspension culture lines to WT.  SIPK-silencing led to increased sensitivity of tobacco leaf tissue to  treatment with bacterial harpin (hrpZ h) {Pseudomonas syringae pv. phaseolicola), indicating Psp  that SIPK plays a negative regulatory role in the hrpZ h-induced hypersensitive response. Psp  HrpZp ph-induced reactive oxygen species (ROS) burst was enhanced in the SIPK-silenced plants S  compared to the WT plants, indicative of a negative regulatory role for SIPK in harpin-induced ROS accumulation. Strong hrpZ h-induced activation of SIPK (and WIPK) in the NahG line Psp  demonstrates that SIPK and WIPK activation are largely SA-independent.  SIPK-silenced  tobacco plants hyper-accumulated SA following harpin treatment, suggesting that SIPK acts upstream as a negative regulator of harpin-induced SA accumulation. To address the functional roles of SIPK in mediating transcriptional responses induced by hrpZpsph  and the oomycetes elicitor P-megaspermin (Phytophthora megasperma H20),  I  monitored the transcriptional responses of WT and SIPK-Ri tobacco cell suspension cultures to each elicitor at 4 and 8-hours post-treatment, using TIGR potato c D N A microarrays (10K element).  SIPK-affected transcriptional responses were analyzed in the context of the WT  responses to these elicitors in order to identify a set of genes associated with each elicitor that respond in a SIPK-dependent manner. SIPK-silencing was found to have a substantial impact on both hrpZpsph- and P-megaspermin-induced transcriptional responses. Hierarchical clustering of genes that responded significantly to either elicitor (but not both) revealed striking functional differences between the two SIPK-dependent responses, suggesting that signal transduction through SIPK plays an important role in conditioning elicitor-specific transcriptional responses. ii  Table of Contents Abstract  "  Table of Contents  »'  List of Tables and Figures  v  Figures  »i "  w  Tables Abbreviations  x  Abbreviations  x  Acknowledgements  1  x  '  Research contributions  xi  Material contributions  xi  Guidance and support  xi  General Introduction  1  1.1  Inducible plant defenses  I  1.2  Elicitors as stress surrogates in triggering defense responses in plants  4  1.2.1  Host vs. non-host resistance  4  1.2.2  Elicitors as stressors  4  1.2.3  Elicitor recognition  7  1.2.4  Signal amplification and second messengers  8  1.2.5  Hormones in defense against pathogens iii  11  1.3  Protein kinases in stress signalling  1.3.1  Plant M A P K s in stress response  16  1.3.2  M A P kinase signalling in tobacco stress response  17  1.4  2  14  Problem Statements and Thesis Objective  20  Salicylate-Induced Protein Kinase (SIPK) Regulates Harpin-induced Stress Response 22  in Tobacco 2.1  Introduction  22  2.2  Methods  24  2.2.1  Plant material and treatment  24  2.2.2  Cell culture and treatment  24  2.2.3  Generation of double mutant NahG/Ri (NGRi) lines  25  2.2.4  Recombinant harpin purification  25  2.2.5  Protein extraction and protein gel blot analysis  25  2.2.6  Immunoprecipitation for in-gel kinase assay  26  2.2.7  In-gel kinase assay  27  2.2.8  In situ staining for hydrogen peroxide  27  2.2.9  S A measurements  28  2.3  Results  29  2.3.1  SIPK-silencing leads to heightened sensitivity of tobacco leaves to hrpZ h  2.3.2  SIPK-silencing leads to increased ROS accumulation following hrpZ h  29  Psp  Psp  treatment 2.3.3  32  Salicylic acid (SA) is a requirement for hrpZ  WIPK) activation  Psph  -induced HR, but not for SIPK (or ••••• 34  iv  2.3.4 2.4  3  SIPK-silencing leads to increased S A levels  36  Discussion  55  2.4.1  SIPK has a negative regulatory role in hrpZp -induced H R  38  2.4.2  SIPK and WIPK interact during the hrpZ -elicited response  39  2.4.3  SEPK negatively regulates hrpZ -induced ROS accumulation  41  2.4.4  SIPK negatively regulates SA accumulation  42  2.4.5  Conclusion  sph  Psph  Psph  .'  43  Microarray Analysis of SIPK-dependent Transcriptional Events in Biotic Stressinduced Tobacco Cell Suspension Cultures  45  3.1  Introduction  45  3.2  Materials and Methods  47  3.2.1  Cell culture and treatment  47  3.2.2  Preparation of megaspermin  47  3.2.3  Medium alkalinization  47  3.2.4  RNAtotai isolation (for expression profiling)  48  3.2.5  R N A labeling, microarray hybridization, and scanning  48  3.2.6  Global expression profiling analysis (microarray)  49  Background correction and signal normalization  49  Detection of significant differential expressions within treatments  49  Statistical tests between gene expression ratios of different treatments  50  Gene annotation  50  Filtering and clustering  51  3.2.7  Reverse transcription and quantitative PCR  v  51  3.3 3.3.1  Results  53  SIPK-silencing has differential effects on hrpZ ph and p-megaspermin-induced Ps  defense responses in tobacco plants and cell suspension cultures  53  3.3.2  SIPK and WIPK transcription and activation in cell suspension culture  54  3.3.3  Confirmation of defense response elicitation in cell suspension culture  57  3.3.4  The effects of silencing SIPK on gene transcription in elicited and non-elicited  tobacco cell suspension cultures 3.3.5  59  SIPK-silencing affects distinct sets of genes in hrpZpsph- and P-megaspermin-  induced responses 3.3.6  64  Functional analysis of SIPK-dependent transcriptional modifications for hrpZPsph-  and P-megaspermin responses 3.4 3.4.1  71  Discussion  79  SIPK has a different role in hrpZPsph- and P-megaspermin-induced defense  responses  79  3.4.2  SIPK-silencing affects the pattern of gene transcription prior to, and during, stress 82  3.4.3  Effects of SIPK-silencing on cell cultures prior to elicitor treatment  3.4.4  A subset of genes affected by silencing SIPK are most likely SIPK-dependent  during elicitor-induced responses  84  85  3.4.5  SIPK modulates transcription of a subset of genes in an elicitor-specific manner .. 88  3.4.6  Conclusion  91  Future Directions  92  Appendix  94  4  4.1  Validation of microarray expression values through qRT-PCR  vi  94  4.2  HrpZp ph and R-megaspermin-induced cell death in WT and SIPK-Ri cell suspension S  culture  97  4.3  Gene ontology summaries for shortlists ofpotato array sequences  4.4  Explanation of correlations performed on short-lists  4.4.1  Conclusion  98 99  104  List of supplemental material (provided on CD)  105  Bibliography  106  vii  List of Tables and Figures Figures Figure 2.1 HR development in WT and SIPK-silenced (SIPK-Ri) leaves upon harpin treatment. 30 Figure 2.2 M A P K activation profiles of WT and SIPK-Ri in hrpZp -induced HR  32  sph  Figure 2.3 ROS and tobacco M A P K interaction during tobacco tissue response to hrpZ  Psph  .... 34  Figure 2.4 SA and SIPK affects on hrpZp h-induced HR  36  Figure 2.5 HrpZp h-induced SA accumulation in WT and SIPK-Ri  37  sp  sp  Figure 3.1 Lesion formation in WT and SIPK-Ri after infiltration with  hrpZp h and Psp  megaspermin  54  Figure 3.2 Relative levels of SIPK transcripts in WT and SIPK-Ri cell cultures following control and elicitor treatments  56  Figure 3.3 SIPK and WIPK activation in wild-type (WT) and SIPK-silenced (SIPK-Ri) tobacco cell suspension cultures  57  Figure 3.4 Cell culture responses to hrpZPsph and P-megaspermin treatments  59  Figure 3.5 Hierarchical average cluster image of 1000 transcripts with significant differences in SIPK-Ri/WT ratios between treatments  61  Figure 3.6 Transcription profile comparison between SEPK-Ri/WT expression ratios for control and treatments  63  Figure 3.7 Correlation between sets of genes that are most significantly affected by (A) SIPKsilencing in the elicitor responses and (B) most significantly induced by elicitor treatment in WT  65  Figure 3.8 Hierarchical linkage clustering of 239 SIPK-dependent genes that respond most exclusively to hrpZ h or P-megaspermin treatment Psp  _  _  72  Appendix  Figure 4.1 Validation of potato microarray expression ratios by qRT-PCR for 12 selected tobacco genes with common Figure 4.2  RNA tai to  samples  HrpZ h and P-megaspermin affects on cell viability Psp  viii  95 97  Figure 4.3 Frequency of occurrence of molecular function gene ontology (GO) terms for considered gene lists (see section 3.3.5) based upon TIGR Solanum tuberosum gene index (StGI)  98  Tables Table 1: 20 genes whose expressions are most likely SIPK-dependent, 4hrs after hrpZp h sp  treatment Table 2:  68  30 genes whose expressions are most likely SIPK-dependent, 4hrs after Pmegaspermin treatment  70  Table 3: Gene ontology summaries of hierarchical clusters of genes uniquely-affected by SIPKsilencing in either of hrpZPsph or P-megaspermin treatment  75  Table 4: Most over-represented 6-mer words in 5' regions (lOOObp) of Arabidopsis gene models associated with co-expressed tobacco genes in Clusters A - D Table 5: Primer sequences for qRT-PCR  78 96  ix  Abbreviations Continued... A G I number  Arabidopsis Genome Initiative locus identifier  ANOVA  analysis of variance  BLASTn  basic local alignment sequence tool (nucleotide)  C(T) CDPKs  DTT  dithiothreitol  ERK  extracellular signal-regulated kinase (human)  GO SLIM  2  PCR  polymerase chain reaction pathogenesis-related Pseudomonas syringae pv. phaseolicola  PT  [mitochondrial] permeability transition  pv.  pathovar  p-value qRT-PCR  false discovery rate R/Avr  'slimmed-down' gene ontologies  superoxide anion programmed cell death  Psph  expressed sequence tags  salicylic hydroxylase  PCD  PR  calcium-dependent protein kinases diaminobenzidine  fdr  0 "  critical threshold  DAB  EST  NahG  r  2  probability value quantitative reverse transcription PCR Resistance/Avirulence Pearson correlation coefficient  H (as in H 4 , H 8 )  40 ug/ml hrpZp phtreatment  HCL  hierarchical cluster linkage  H 0  hydrogen peroxide  RLK  receptor-like kinases  HR  hypersensitive response  ROS  reactive oxygen species  hrp  hypersensitive reaction and sensitivity  JA  jasmonic acid or jasmonate  2  2  LAR log LOWESS M (as i n M 4 , M 8 ) M A P kinase  MAPK  Ri  S  R S (as in H 4 ) R S  reverse transcription or reverse transcriptase (as in 'Superscript II RT')  SA  salicylic acid or salicylate  st. dev.  locally weighted linear regression  StGI  9 (ig/ml B-megaspermin treatment mitogen activated protein kinase (any of M A P K K K K s , M A P K K K s , M A P K K s , and MAPKs) M A P kinase phosphorylated by M A P K K s  MES  2-(N-morpholino) ethanesulfonic Acid  NAC  N-acetyl cysteine  standard deviation Solanum tuberosum Gene Index  TAIR  The Arabidopsis Information Resource  TC  tentative consensus sequence (TIGR EST identifier)  TIGR  The Institute for Genomic Research  TMV  tobacco mosaic virus  UV  ultra-violet  var.  variety  VIGS WT  X  responsive to SIPK  RT  local acquired resistance logarithm  R N A interference, small interfering (siRNA)  virus-induced gene silencing wild-type (untransformed)  Acknowledgements Research  contributions  The novelty and significance of this thesis has been greatly enhanced by the contributions of Marcus Samuel, Kinga Drzewiecka, and The Institute for Genomic Research (TIGR). In particular, Marcus Samuel was solely responsible for western blots (Figure 2.2A, Figure 2.3A, and Figure 2.4), the in-gel kinase assay (Figure 2.2B), and P C R verification of the N G R i line (data not shown). Marcus Samuel also played a leading role in designing the microarray experiments involving the SEPK-Ri/WT hybridizations. For quantification of SA, Kinga Drzewiecka performed all H P L C work from tissue samples as well as preliminary data analyses. As part of a collaboration to provide gene expression data for Solanaceae species to the public (Solanum tuberosum Gene Expression Database, SGED), TIGR performed microarray experiments from submitted RNA ai, providing me with raw slide data. These specific research contributions were conducted in the spirit of collaboration with the understanding that such work could potentially appear in my thesis. This work has been, or will be, used exclusively in this thesis, in a recent publication for which I was second author (Samuel et al., 2005), or in possible future publications for which I would be primary author. tot  Material  contributions  I would like to thank J. Lee (Leibniz Institute of Plant Biochemistry, Germany) for providing the recombinant harpin plasmid pT7-7 (with purification protocols) and S. Kauffmann (University of Reims, France) for providing the purified P-megaspermin. Y . Ohashi (National Institute of Agrobiological Resources, Japan) generously donated the anti-WTPK and anti-SIPKantibodies to our lab while J.A. Ryals (J. Ryals, CIBA-Geigy, N C ) provided us with the NahG tobacco seed. In addition to donating microarray slides and associated reagents, TIGR also invested a considerable amount of their time performing the array experiments.  Guidance and support This thesis is the most tangible and perhaps the most complete demonstration to date of my ability to function as a scientist. While I have been acknowledged for this effort with a Master of Science degree, and therefore could consider myself a masterful scientist, there are more than a few extraordinary people that deserve equal share in this glory. M y family, friends and colleagues have shown their true mastery in guiding me through the mine fields of insurmountable biological complexity ever closer to the hallowed ground of good science. I offer my sincerest gratitude to the Botany Department administrators, Veronica Oxtoby, Judy Heyes and Lebby Balakshin who have kept my program on track. I would like to thank Juergen Ehlting, Dana Aeschlimann, and Jochen Brumm for many fascinating discussions on experimental design, statistics, and computing. M y MSc program has been greatly enriched by the guidance and support of Ellis lab group members including Marcus Samuel, Rishi Gill, Godfrey Miles, Greg Lampard, Somrudee Sritubtim, Alana Clegg, Corrine Cluis, Alex Lane, and Jin Suk Lee. I would especially like to thank a top-notch professor, Brian Ellis, for compassionately, often humorously, delivering the frankest judgement my work will likely ever face. Finally, I'd like to thank my family (Noriko, Mom, Dad, Delyth, and Robert) for supporting me through the difficult transition I have made from construction worker to research collaborator. As you can see, I'm running out of space to thank people. I honestly need another 100 pages for that. Instead I'll just say "Thanks!" and "Yay! I'm done!"  xi  1 General Introduction l.l Inducible plant defenses \  Plants exhibit a diversity of constitutive and inducible defense mechanisms that impede their  colonization by disease-causing viruses, prokaryotes, fungi, and oomycetes. As a first line of defense, plants constitutively produce structural barriers such as the cuticle and lignocellulosic cell walls. They also accumulate a diversity of low molecular weight secondary metabolites, socalled phytoanticipins, which often exist in inactive forms that are normally compartmentalized away from their activating enzymes (VanEtten et al., 1994). Considerable variation in the expression of constitutive defenses, between and within taxa, suggests that such metabolic commitments present considerable fitness costs. This may have been compensated for by the evolution of response mechanisms that involve de novo synthesis of antimicrobial compounds, triggered by pathogen invasion, this pattern has been termed the phytoalexin response (Koricheva, 2004). Such inducible defenses would seem to be energetically more efficient than constitutive defenses, since their synthesis is spatially and temporally regulated to match the ever-changing profile of pathogen invasion. While disease symptoms may take days or even months to appear in susceptible plants, inducible defenses are triggered within minutes or hours after resistant plants are first colonized. Plants possess a diverse set of inducible responses to counter the wide range of pathogen invasion strategies.  A common invasion strategy of pathogens is the penetration and  colonization of the apoplast, to which plants respond with a variety of wall-associated defenses such as peroxidative cross-linking of cell wall constituents, cell wall lignification, and papillae formation (through callose, silica, and phenolic deposition) (Ward et al., 1991; Heath, 2000a; Hahlbrock et al., 2003). Pathogenesis-related (PR) proteins are a heterogeneous group of 1  inducible proteins, some of which display antimicrobial activity. They are targeted to both intracellular and extracellular regions and include glucanases, chitinases, chitin-binding proteins and cysteine-rich thionins (Hammond-Kosack and Jones, 1996).  Low-molecular weight  inducible antimicrobial secondary metabolites (phytoalexins) are also frequently produced as a result of biotic stresses. These include flavonoid derivatives such as isoflavonoids and tannins, indole derivatives such as camalexin and many other classes of metabolites (Taiz and Zeiger, 1998; Thomma et al., 1999). I am not aware of evidence which demonstrates that any one component of this suite of constitutive and inducible metabolic responses is capable in itself of providing robust and broadly effective defense against invaders. Instead, it is more likely that such defenses have a cumulative effect in slowing and sometimes stopping pathogen attack. This level of defense is sometimes referred to as horizontal, or non-host, resistance (Holub and Cooper, 2004). In many plant host-pathogen interactions, the initial pathogen attack is followed within hours or days by the formation of local lesions in the challenged tissue, immediately surrounding the point of attempted entry.  This rapid and localized cell death, or hypersensitivity, was first  described in 1915 as part of the response of rust-resistant cereals to infection with avirulent rust strains (Heath, 2000b). Now referred to as the hypersensitive response (HR), this localized cell death pattern is considered to be a key host resistance response that effectively contains the pathogens near the point of entry, and also alerts more distal parts of the plant to possible attack, through a systemic signalling system. HR-related cell death is most likely a form of programmed cell death (PCD), similar to that employed in some aspects of normal plant development and senescence. In HR, cells within the lesion area typically accumulate brown/black pigments and appear aufofluorescent under U V light, presumably due to the build-up and oxidation of phenolic compounds.  However, the  overall cell death process is complex and the associated molecular mechanisms remain obscure 2  (Heath, 2000b). P C D in plant cells is characterized by several visible features that are also detected during programmed cell death (apoptosis) in animal cells, where it is similarly associated with both development and pathogenicity. These features include chromatin aggregation, D N A fragmentation, and cytoplasmic/nuclear condensation. However, not all of these features necessarily accompany H R in plants, nor do they provide a clear distinction between P C D and necrosis resulting from cytotoxic stress (Jabs, 1999).  In contrast to the  largely unregulated necrosis response, H R involves significant gene regulation, processing of external signals, and sustained ATP consumption (Heath, 2000b). H R also occurs over longer periods of time (hours to days), and its early stages are reversible to a certain extent. The role of HR in plant defense is not always clear, since necrotrophic, and some biotrophic, pathogens are unimpeded by lesions in host tissues (Heath, 2000b).  It has therefore been  proposed that such cell death events could also act as cellular and intercellular stress signals, inducing non-fatal defense responses (as described above) in adjacent tissues (Heath, 2000b). In fact, studies have shown that tissues immediately surrounding the biotic stress-induced lesion (within 5-10mm) have increased resistance to other pathogens. This phenomenon is called local acquired resistance (LAR). pathogenesis-related  As in the H R region, cells in the L A R region exhibit elevated  (PR) gene expression along with increased phenylpropanoid and  sesquiterpenoid pathway activity, but without the accompanying cell death (Cordelier et al., 2003). Entire plants often exhibit heightened sensitivity to a broad range of pathogens following a localized H R response to a particular pathogen.  This phenomenon, referred to as systemic  acquired resistance (SAR) (Ross, 1961) has been shown to confer systemic resistance even against pathogens to which the plants are normally susceptible (Ryals et al., 1996). The S A R response exhibits many of the properties of L A R , although the intensity of the responses is typically reduced in more distal parts of the plant (Yalpani et al., 1991). 3  Plants thus possess layered and integrated inducible defenses designed to impede pathogens. Adding to this complexity is the necessity of mounting an appropriate response to more than one threat simultaneously, and doing so within a highly variable environmental and developmental context. Regulating these processes inevitably requires a sophisticated sensing and transduction network that allows plants to perceive and integrate a wide variety of stimuli, and to translate that information into appropriate responses.  1.2 Elicitors as stress surrogates in triggering defense responses in plants 1.2.1  Host vs. non-host resistance  Plant resistance within a species sometimes appears to be dichotomous, meaning that individual plants in a population are either susceptible or resistant, with no intermediate resistance phenotype occurring within the species. Genetic analysis reveals that such a pattern of strong specific resistance to a particular pathogen is often determined by a single locus in the host genome, a so-called resistance (R) gene (Nimchuk et al., 2003). The role of the R-gene product is apparently to recognize a corresponding avirulence (Avr) gene product whose formation is encoded in the pathogen genome.  However, such R/Avr-defined pathosystems  typically involve only a subset of the vast array of potential plant-microbe interactions. Most plants are resistant to most pathogens (i.e. they display non-host resistance), responding to an attempted invasion with an inducible defense profile that appears very much like the incompatible host response in the R/Avr systems (Heath, 2000a).  1.2.2  Elicitors as stressors  To facilitate understanding of resistance mechanisms, it is experimentally convenient to work with biochemically-defined molecules (elicitors) derived from pathogens, rather than with live pathogens themselves. Elicitors are, by definition, able to trigger plant defenses even in the absence of their associated pathogen (Montesano et al., 2003). 4  Pathogen-derived elicitors  belong to diverse chemical classes including oligosaccharides, lipids, peptides, and proteins; plant cell wall fragments cleaved by pathogens may also act as elicitors (Felix et al., 1999). Depending on the range of plants in which they trigger defense responses, these elicitors have been described as non-specific (affecting a wide range of plant taxa), or specific (affecting only a few genera or species) (Heath, 2000b). Innate defense responses, common to all eukaryotes, are thought to rely primarily upon the detection of non-specific elicitor types.  As with mammalian triggers of innate immune  responses, these non-specific elicitors are often the cell wall constituents of the pathogen. For example, plants exhibit near-universal sensitivity to the bacterial motility protein, flagellin (Felix et al., 1999), and to lipopolysaccharides secreted by Gram-negative bacteria (Nurnberger and Scheel, 2001). Oligosaccharides are common cell wall constituents in plants, oomycetes, and fungi, yet differences in saccharide linkages between these kingdoms provide a basis for plant distinction between endogenous and foreign oligosaccharides (John et al., 1997). This linkage detection ability may have evolved prior to the divergence of vascular plants, since some marine algae also exhibit sensitivity to microbial oligosaccharide elicitors (Potin et al., 1999). Similarly, various fungi and bacteria enzymatically degrade pectin polymers in plant cell walls to yield oligogalacturonide fragments which plants recognize as elicitors (Montesano et al., 2003). Pathogens also secrete diffusible signals that act as non-specific elicitors for particular plant families. Elicitin proteins secreted by oomycetes of Phytophthora spp. and Pythium spp. are pathogenicity factors in disease induction in Nicotiana spp. and Brassica spp., and act as elicitors in these species (Baillieul et al., 2003). Elicitins fall into five structural categories including acidic 10 kDa protein a/p/za-elicitins, basic 10 kDa protein (P)-elicitins, and 14-18 kDa glycoproteins (Wei et al., 1992; Baillieul et al., 2003).  5  Many non-specific proteinaceous diffusible elicitors have been identified from Gramnegative plant pathogenic bacteria on the basis of their ability to elicit H R in both host and nonhost plants. These "hypersensitive reaction and sensitivity proteins" ('hrp' or 'harpins') include hrpN from Erwinia amylovora, hrpNEch from Erwinia chrysanthemi, hrpZ from Pseudomonas syringae pv. syringae, (also tomato, and glycinea pathovars) and popA from Ralstonia solanacearum (Lee et al., 2001b). Harpins are encoded in the hrp gene cluster that appears to be highly conserved amongst pathogenic bacteria (Wei et al., 1992). The majority of these harpins are now known to belong to the type III secretory system (TTSS), an elaborate and evolutionarily conserved device used by bacteria for injecting coercive proteins (effectors) into animal and plant cells (Galan and Collmer, 1999; Tampakaki et al., 2004). Growing evidence supports the concept that TTSS components involved with transporting pathogenesis-associated effectors are recognized by a wide range of plant species, while many of the effector proteins themselves appear more critical in defining pathogenicity within species (Galan and Collmer, 1999; Lee et al., 2001a; Tampakaki et al., 2004). Avr gene products can be considered as very specific elicitors, since they only induce an H R in a subset of a plant species which possess particular R genes (de Jong et al., 2004). The list of known Avr gene products now includes endoxylanases (of Trichoderma viride against tobacco), viral coat proteins (of tobacco mosaic virus [TMV] against tomato and tobacco), protein or peptide toxins (of Helminthosporium victoriae against oat), and syringolides (acyl glycosides) (of P.syringae against soybean) (Montesano et al., 2003). They thus do not appear to possess any common function.  Current models of bacterial Avr gene function propose that they  typically serve a specific function within their bacterial host, while playing additional roles within the host plant cell, where they target specific biological processes or cellular compartments.  Many Avr genes are effector proteins of the type III secretory system,  6  suggesting that R/Avr interactions may reflect the front lines of complex interactions between plant surveillance and pathogen suppression of that surveillance (Nimchuk et al., 2003).  1.2.3  Elicitor recognition  Plants are thought to detect external stimuli through receptors that are located on the plasma membrane, in the cytosol, or on cell organelle membranes (Lee et al. 2001a).  Only a few  putative receptors for non-specific elicitors have been identified. These include receptors for Pelicitins (Wendehenne et al., 1995), harpins (Lee et al., 2001a) and flagellin (Gomez-Gomez and Boiler, 2000). In contrast, ~125 putative R genes have been identified in plants (Nimchuk et al., 2003). R genes were initially thought to act as receptors that directly bind to Avr products (receptor-ligand model), and the vast majority of R genes do contain a leucine-rich repeat (LRR) region, capable of offering such ligand-binding opportunities (Nimchuk et al., 2003). However, very few examples of direct R/Avr interaction have been demonstrated despite intensive effort, suggesting that the receptor-ligand model is likely the exception and not the rule (Tang et al., 1996; Jia et al., 2000; Lahaye, 2004). A more elaborate system of R/Avr interactions has been proposed in the "guard hypothesis", which postulates that Avr products target resistance signalling components, and that R genes monitor, or "guard", such effects by complexing with Avr-affected proteins to trigger defense responses (Van der Biezen and Jones, 1998). Non-specific elicitors provide an opportunity to improve our understanding of innate defense response mechanisms through analysis of related signalling events and defense outcomes. Purified elicitors, even crude extracts, provide researchers with pathogenesis-simulating stimuli that are far less complex than the live pathogen, thus empowering reductionist approaches to understanding the "black box" of defense signalling. For instance, it is possible to easily modify the level of the elicitor being applied, thereby simulating different levels of pathogen assault, and subsequently assess the impact on plant sensitivity and cell fate determination. 7  Inducible defense responses that impede pathogenic invasion are coordinated by early plant responses that involve signal component interactions. While plants are able to rapidly respond to pathogens with an inducible array of defenses, few (if any) of these defense are likely to be directly induced or regulated by an elicitor-receptor complex. Instead, recognition events must trigger a series of biochemical reactions, collectively termed the signal transduction network, that perceive, amplify, and integrate input signals.  Signalling can be expected to be very  complex, and indeed, no complete signal transduction pathway has yet been clearly established that connects an elicitor recognition event in plants with a defense outcome. Instead, many disparate signalling components have been implicated in stress-induced defense responses through genetic or biochemical analysis. These components include second messengers, protein kinase cascades, and hormones, which together are thought to coordinate defense responses in a stress-specific manner.  1.2.4  Signal amplification and second messengers  Pathogen-derived cues are typically subtle, likely as a result of strong selective pressures against them.  Conversely, inducible responses in plants must generally be rapid and severe  (sensitive) i f the attacked plant is to successfully counter the pathogen invasion. Sensitivity to infection is largely achieved through signal amplification which scales up weak signals from rare elicitor-receptor complexes to trigger massive and rapid defense responses. Pathogen attack rapidly induces abrupt and transient increases in the levels of a number of simple, inorganic ions that appear to have little or no direct antimicrobial activity, and are therefore candidates for second messengers.  These second messengers can amplify the signal  in various ways, including activation of protein kinases and phosphatases (Taiz and Zeiger, 1998).  Thus, C a  2+  efflux across the plasma membrane into the apoplast was shown to be 8  required for a host-incompatible H R and a non-host H R (Levine et al., 1996), and to be associated with the activation of pathogen-specific receptors (Blume et al., 2000 in Nurnberger and Scheel, 2001). Moderate receptor-mediated C a  2+  efflux into the cytosol is known to activate  defense genes via calcium-dependent protein kinases (CDPKs) which have mitogen-activated protein (MAP) kinases as likely phosphorylatable targets (Nurnberger and Scheel, 2001). Fluxes of potassium, chloride, and possibly nitrate ions, accompany calcium ion flux and it is likely that they collectively influence early stress response events (Armengaud et al., 2004; Lam, 2004). Lipid derivatives may also function as important second messengers.  In animal cells, the  phospholipase C pathway mediates calcium channel activity via inositol triphosphate (IP3) which is produced from the membrane phospholipid, phosphatidylinositol, by the action of phospholipase C (Taiz and Zeiger, 1998). Another product of the phospholipase C reaction is diacyl glycerol (DAG).  In animals, D A G activates protein kinase C (PKC), which in turn  activates protein phosphorylation (kinase) cascades.  Recent studies have demonstrated the  presence and functionality of the phosphoinositide system in plant defense signalling, although no recognizable homologue of mammalian P K C has been found. Accumulation of B°3 occurs during the calcium-affected H R of lemon seedlings in response to a fungal pathogen (Ortega and Perez, 2001) while phosphatidic acid, or P A , is produced through phosphorylation of D A G within a few minutes of the Cf-4/AVR4 interaction in the tomato/'Cladosporium fulvum pathosystem (de Jong et al., 2004). In addition, mutations of either EDS1 (enhanced disease susceptibility) or PAD4 (phytoalexin-deficient), genes which encode lipase homologues, affect HR development in Arabidopsis thaliana (Nimchuk et al., 2003; Lam, 2004). A ubiquitous and early feature of plant resistance to biotic stresses is the biphasic accumulation ("oxidative burst") of reactive oxygen species (ROS) such as hydrogen peroxide ( H 0 ) , superoxide anion (0 "), and the hydroxyl radical (OH') (Bestwick et a l , 1997). In fact, 2  2  2  9  many pathogen-induced H R responses are inhibited by treatments with antioxidants or antioxidant enzymes that scavenge ROS (Jabs, 1999). Biotic stresses are known to enhance the accumulation of ROS in the challenged tissue by positively regulating the activities of various ROS-generating enzymes such as N A D P H oxidase, amine oxidases and cell wall-bound peroxidases, while simultaneously suppressing the activities of ROS-scavenging mechanisms (Mittler, 2002). The rapid development of HR at sites of ROS hyper-accumulation suggests that ROS accumulation may be coupled to cell death, yet ROS accumulation is, in itself, considered insufficient to directly kill cells through its oxidative potential (Jabs, 1999).  Instead, ROS  appear to contribute to both defense gene expression and regulation of PCD, suggestive of second messenger roles (Desikan et al., 1998; Jabs, 1999; Mittler et al., 1999). It is unclear what underlying mechanisms link ROS levels to defense responses, but the impact of ROS on the cellular redox status could potentially impact redox-sensitive signalling proteins such as Gproteins, kinases and protein phosphatases (Mittler, 2002). Chloroplasts and mitochondria may play key roles in mediating ROS-dependent defense responses.  A PCD-regulating factor in both animals and plants is the phenomenon of  mitochondrial permeability transition (PT), characterized by the sudden release into the cytosol of ROS, and of PCD-inducing signals such as cytochrome C (Jabs, 1999; Tiwari et al., 2002; Dutilleul et al., 2003). Nitric oxide (NO) may also be important in cell death induction, as demonstrated by an avirulent pathogen-induced rise in NO-synthase (NOS) activity and the cell death-suppressing effects of NOS-inhibitors (Delledonne et al., 1998; Durner et al., 1998).  In stress-induced HR,  N O has been shown to inhibit the cytochrome C release that is characteristic of mitochondrial PT (Zottini et al., 2002). Generally, N O has been implicated in phosphorylation-, nitrosylation-, and calcium-controlled mechanisms linked to inhibition of homeostatic enzymes such as tobacco aconitase, catalase and ascorbate peroxidase (Neill et al., 2002). However, no direct 10  interactions of N O with D N A have been demonstrated, suggesting that N O is strictly a second messenger that indirectly influences downstream metabolic and/or transcriptional changes (Lamattina et al., 2003). Second messengers likely account for the majority of signal amplification that must occur i f plants are to adequately respond to subtle threats, yet secondary messaging may have a more sophisticated/diverse role. The magnitude, timing and spatial characteristics of calcium ion fluxes can vary in response to different stresses (the "calcium signature hypothesis").  This  model postulates that information about the nature of a perceived stress types is encoded through calcium patterning (Scrase-Field and Knight, 2003).  Likewise, variations in the biphasic  modulation and cellular localization of the ROS burst have been observed to differentially affect transcription (Vandenabeele et al., 2003). Second messengers have also been shown to interact with each other in complex ways. N A D P H oxidases, which have been proposed to be responsible for at least part of the extracellular ROS production (02--> H2O2), possess C a has been correlated with C a  2+  2+  binding domains and their activation  ion fluxes (Bolwell et al., 1999) In addition, N O can promote  ROS accumulation through the inhibition of ROS-scavenging enzymes (Neill et al., 2002). However, broad generalizations about these phenomena must be treated with caution, since the extent of these interactions, and the participation of particular second messengers, appears to vary between plant-pathogen pairings.  1.2.5  Hormones in defense against pathogens  Defense responses such as HR and S A R are coordinated tissue responses intercellular communication. This is thought to be based upon the regulated production and perception of a diversity of diffusible chemical messengers. Studies of plant defense responses to abiotic and biotic stress, particularly in Arabidopsis mutants, have pointed to important roles for 11  phytohormones such as salicylate (salicylic acid, or ' S A ' ) , jasmonate (jasmonic acid, or 'JA'), and ethylene as defense response determinants (Nimchuk et al., 2003). SA is a simple phenolic acid that accumulates gradually (over hours to days) both in challenged cells and systemically throughout a plant that is experiencing H R at the infection site (Ward et al., 1991; Yalpani et al., 1991). Studies with transgenic tobacco plants expressing an ectopic SA-degrading enzyme, salicylate hydroxylase (NahG), have established that SA is required for many, but not all, stress-induced H R and SAR responses (Lam, 2004). Consistent with this, exogenous application of S A , or its synthetic analog, B T H (benzothiadiazole), induces PR gene expression and SAR (Shah and Shokat, 2002). Cloning and characterization of genes that affect SA accumulation suggest that SA is synthesized in plants through two alternate branches of the shikimate pathway, involving either isochorismate or phenylalanine as the initial substrate (Shah and Shokat, 2002). SA-mediated PR gene expression and S A R require an ankyrin repeat-containing protein, NPR1 (non-expressor of PR1), which has been shown to migrate to the nucleus following perception of SA (Kinkema et al., 2000). This change in localization has been suggested by Mou et al. (2003) to be mediated through salicylate-induced redox modification of the NPR1 protein. Once in the nucleus, NPR1 facilitates protein-protein interactions with a TGA-family transcription factor that enables the activation of PR gene transcription (Pieterse and Van Loon, 2004). Taken together, current evidence therefore strongly suggests that SA acts as a positive regulator of HR-associated defense responses. Ethylene  (C2H4)  is another important phytohormone that accumulates rapidly upon wounding  and abiotic/biotic stress. Since it also participates in longer-term processes such as meristem development, and senescence, it is understandable that ethylene biosynthesis, perception, and signal transduction are highly regulated. Several potentially rate-limiting steps in ethylene signalling have been identified, including the conversion of methionine to ethylene and the 12  regulation of ethylene sensitivity by receptor abundance (consult Wang et al., 2002 for a review of these). In stress responses, ethylene is recognized for its role in P C D , particularly as a positive regulator of H R lesion development, where it has been associated with an increase in ROS (Overmyer et al., 2003).  Interestingly, unlike SA, ethylene appears to inhibit the cell  death induced by some necrotrophic pathogens, while enhancing the cell death induced through perception of other host- and non-host-specific stresses (Wang et al., 2002). Jasmonic acid (JA), and its more volatile ester derivative, methyl-JA (MeJA), are important fatty acid derivatives which accumulate within hours in plant tissues as a response to wounding and various biotic stresses. Exogenous applications (sprayed on leaves) of JA, MeJA, and their precursors induce anti-herbivory proteinase inhibitors, known to impede insect digestive processes (Farmer and Ryan, 1990, 1992).  While JA is best known for its role in mediating  herbivore-targeted defenses, it has also been implicated in regulating defense responses to pathogens. For instance, J A has been shown to mediate the expression of a class of PR genes encoding cysteine-rich proteins (thionins) in Arabidopsis (Epple et al., 1995 in HammondKosack and Jones, 1996).  JA is also thought to help regulate P C D , since it inhibits the  propagation of cell death, as demonstrated by the increased sensitivity of JA-insensitive and JA biosynthesis-defective Arabidopsis mutants to oxidative stress (ozone) (Overmyer et al., 2003), and the decreased lesion formation observed in tobacco plants treated with M J before exposing them to cytotoxic levels of ozone (Orvar et al, 1997).  J A intermediates such as 12-  oxophytodienoic acid (OPDA) are also known to act as signals, and it has been suggested that complex regulation of the J A biosynthetic pathway may contribute to a stress-specific fatty acidderived signal ensemble called the 'oxylipin signature' (Overmyer et al., 2003)  13  1.3 Protein kinases in stress signalling Protein phosphorylation/desphosphorylation is a universal intracellular signalling process in both eukaryotic and prokaryotic cells. This mechanism of reversible protein modification is particularly well-suited to rapid amplification of input signals, and has been found to be involved in all aspects of early stress responses, from ligand perception through to modification of ultimate targets, including enzymes, transporters, and transcription factors. Several different classes of enzymes are involved in mediating this interconversion.  Plants appear to have  retained the basic two-component phosphorylation regulatory system that is also found in bacteria.  This consists of a sensor protein whose kinase domain phosphorylates a response  regulator protein, which then transmits a downstream signal to eventually enact transcriptional and/or physiological change. In addition to this type of signal transduction,, all eukaryotes possess a more elaborate protein kinase network that transmits signals in a similar one-way fashion through a diverse set of receptor kinases, tyrosine kinases, calcium-affected kinases, and serine/threonine kinases, including mitogen-activated (MAP) kinases (Taiz and Zeiger, 1998). In plants, serine/threonine-specific receptor-like kinases (RLKs) appear to substitute for the well-known animal receptor tyrosine kinases (RTKs), acting as plant-specific transmembrane receptors (Stone and Walker, 1995). Most R L K s possess an extracellular leucine-rich repeat (LRR) and a cytoplasmic serine/threonine kinase domain. Arabidopsis, for instance, has >600 R L K s , a family that has increased significantly in size from the >440 estimated to have been present in the ancestral angiosperms (Shiu et al., 2004). The R L K s include a growing number of R genes, such as Arabidopsis FLS2 and rice Xa21 (Di Gaspero and Cipriani, 2003), but other serine/threonine kinase R proteins have been identified which lack the L R R domain, such as tomato Pto (Nimchuk et al., 2003).  14  Similarly unique to plants are calcium-dependent protein kinases (CDPKs), which possess a calmodulin-like domain. The CDPKs in plants appear to play the role played in animals by calcium-phospholipid-dependent (PKA/PKG) and calmodulin-dependent (CaMK) kinases in sensing calcium fluxes.  The plant C D P K group is less extensive than the R L K family,  comprising -34 family members (Hrabak et al., 2003), which has led to the suggestion that CDPKs are more likely to act as generic signal enhancers than as stimuli-specific switches. The evidence accumulated to date makes it clear that calcium-dependent signalling events and R L K mediated phosphorylation events are not co-requisites for all stress responses. By contrast, downstream [stress, cell division and development] signalling events are commonly regulated by a highly conserved cytoplasmic M A P kinase network.  Since M A P  kinases typically have steep stimulus-response curves initiated minutes after stimulus perception, they may act as "molecular switches" to turn on specific inducible events related to development or to stress response (Zhang and Klessig, 1997). Their relevance to defense signalling has been demonstrated by the ability of general M A P kinase inhibitors to block defense gene transcription (Zhang e t a l , 1998). Three major M A P kinase subfamilies have been identified that operate in a three-tier phosphorylation hierarchy (Ichimura, 2002; Jonak and Hirt, 2002). M A P kinase kinase kinases ( M A P K K K s ) can be activated through a diversity of inputs such as M A P K K K clustering, physical interaction/phosphorylation by second messengers, or direct binding with receptors (Jonak et al., 2002).  Once activated, M A P K K K s enable the phosphorylation of two serine  and/or threonine sites of specific M A P kinase kinases (MAPKKs). These M A P K K s , in turn, catalyze the dual phosphorylation of a diversity of M A P K s at both threonine and tyrosine residues of a specific T - X - Y motif found in the activation loop of the M A P K protein. Output from this signalling cascade is achieved through phosphorylation by activated M A P K s of  15  enzymes, cytoskeletal elements, other protein kinases (non-MAP kinases), and transcription factors. While the mechanisms of signal transduction through M A P kinase cascades have been extensively studied in metazoans, most of the definition of plant M A P kinase network function has been accomplished in Arabidopsis  thaliana.  Sequencing of the Arabidopsis genome has  revealed >60 putative M A P K K K s , but only 10 M A P K K s and 20 M A P K s (Ichimura, 2002). Phylogenetic analysis of Arabidopsis M A P kinases has further revealed that these kinase types resolve into subgroups according to sequence similarities (Ichimura, 2002).  Only a few  Arabidopsis M A P kinases have been assigned to putative functional modules on the basis of protein-protein interactions (Ichimura et al., 1998) and in vivo (protoplast) sequential kinase activations (Asai et al., 2002). Since cellular functions have been tentatively assigned to only a small number of the kinases, it is not clear whether the structural subgroups reflect similarities in their biological roles.  1.3.1  Plant MAPKs in stress response  One group of M A P K s with homologues identified from several plant species have been found to be responsive to many stresses, and thus appear to be common convergence points in stress signalling. In Arabidopsis, AtMPK3 (a 43-44kDa protein) and A t M P K 6 (47kDa) are strongly activated minutes after application of various abiotic stresses (Kovtun et al., 2000; Ahlfors et al., 2004). A t M P K 4 (45kDa) activation has also been induced by various abiotic stresses (Ichimura et al., 2000).  In general, biotic stresses appear to result in activation of  AtMPK4 and A t M P K 6 (Desikan et al., 2001b; Nuhse et a l , 2000), while evidence for AtMPK3 activation by biotic stresses is relatively limited. In tobacco, salicylic acid-induced protein kinase (SIPK) is the putative ortholog of AtMPK6. SIPK displays rapid activation in response to exogenous SA application (Zhang and Klessig, 1997), exogenous N O (Kumar and Klessig, 2000), wounding (Zhang and Klessig, 1998a), 16  fungal elicitors (Zhang et al., 1998), Cf-9/Avr9 interaction (Romeis et al., 1999), harpin (hrpZp ph) (Lee et al., 2001a), osmotic stress (Hoyos and Zhang, 2000), and U V - B (Miles et al., S  2002) . Wound-induced protein kinase (WIPK) is the putative tobacco ortholog of AtMPK3 and is known to be activated by H2O2 (Kovtun et al., 2000), ozone (Samuel and Ellis, 2002), and bacterial flagellin (Asai et al., 2002).  Interestingly, stress leads to the rapid induction of  transcription of the gene encoding WIPK, and to de novo synthesis of WIPK, whereas SIPK is constitutively expressed. This indicates that these closely-related M A P K s are governed at the transcriptional level by different regulatory mechanisms during stress response (Zhang and Klessig, 1998b). A multiple stress'study on the Peruvian tomato (Lycopersicon peruvianum) revealed that the activation of L e M P K l (putative ortholog to SIPK and A1MPK6), was similarly induced by systemin (herbivory cue), four oligosaccharide elicitors and U V - B , whereas activation of LeMPK3 (putative ortholog to WIPK and AtMPK3), was induced only by U V - B (Holley et al., 2003) . Treatment  of  alfalfa  (Medicago  sativd) with different  elicitors  (chitin  and N-  acetylglucosamine oligomers, b-glucan, and ergosterol) resulted in differential activation patterns of SIMK (ortholog to SIPK/A1MPK6), S A M K (ortholog to WIPK/AtMPK3), and two other kinases, M M K 2 and M M K 3 . This demonstrates the potential for plants to sense different cues from the same intruder (yeast) through multiple M A P K activation patterns (Cardinale et al., 2000). From these studies, it is evident that, for many angiosperm species, M A P K homologs of AfMPK6 are particularly relevant to responses to diverse biotic stimuli.  1.3.2  MAP kinase signalling in tobacco stress response  While M A P K K K s and M A P K K s likely play critical roles in mediating stress responses, it is believed that it is primarily the terminal M A P K s that phosphorylate downstream signalling targets, thereby creating the necessary output to the regulatory machinery of the cell. As part of 17  that forward signalling, eukaryotic M A P K s  have also been shown to directly interact with  transcription factors, placing them close, in terms of causation, to the changes in gene expression that help to re-program the cell's physiology and metabolism in response to the perceived stress (Asai et al., 2002). SIPK and WIPK are well-characterized tobacco M A P K s that have been implicated in stress responses to many divergent stimuli, and thus appear to act as convergence points in defense signalling.  The observation that SEPK and WIPK exhibited  increased localization in the nucleus following harpin (hrpZp h) challenge of tobacco cell sp  suspension culture (Clegg, 2004) is consistent with a recent report that A t M P K 6 and AtMPK3 migrate to the nucleus from the cytoplasm upon stress activation in Arabidopsis (Ahlfors et al., 2004). The ease of transforming tobacco using Agrobacterium tumefaciens-based vectors has enabled transgenic gain- or loss-of-function genetic studies on tobacco M A P kinases. The correlation of rapid and sustained SIPK activation with stress-induced H R prompted various studies on the effects of transgenic manipulations of SIPK, or its related kinases, on defense responses. Ectopic expression of a constitutively active form of N t M E K 2 (endogenous activator of SIPK) in tobacco leaves resulted in elevated activation of both SEPK and WEPK, leading to HR in the absence of stress (Yang et al., 2001), while transient over-expression of SIPK itself led to HR-like cell death in A. tumefaciens-'mf\\trated tobacco leaves (Zhang and Liu, 2001). These observations imply that SIPK activation plays a definitive role in cell fate determination. SIPK activation is concomitant with both the ROS burst and S A accumulation in many stress-induced responses. Exploration of the relationship between SIPK and these other signalling components has thus far been addressed in the context of the oxidative stresses, ozone (Samuel et al., 2000; Samuel and Ellis, 2002) and U V - C (Miles et al., 2002). Through its ability to spontaneously produce an oxidative burst, ozone challenge mimics the ROS accumulation which accompanies many HRs induced by biotic stresses (Orvar et al., 1997). ROS-generating 18  ozone has been shown to be sufficient for rapid SIPK activation (10 minutes post-infiltration) (Samuel et al., 2000), and this effect appears to be contingent upon the involvement of membrane-associated receptors (Miles et al., 2002).  This evidence supports a role for ROS in  mediating at least some early M A P kinase-related signalling events. However, contrary to expectations, suppression of SIPK in transgenic tobacco resulted in hyper-accumulation of ROS (8 hours post-infiltration), and ultimately rendered these transgenic plants more prone to ozone-induced HR-like lesion formation (Samuel and Ellis, 2002). Thus, SIPK and ROS appear to interact with each other, precluding a simple, linear relationship between the two in HR-related signal transduction.  Equally enigmatic is the relationship  between SIPK and its putative inducer SA (Zhang and Klessig, 1997). SIPK activation has been shown to be induced in the NahG line upon cryptogein treatment, implying that the activation operates in an SA-independent manner (Lebrun-Garcia et al., 2002).  The role of SIPK in SA  accumulation, or on plant sensitivity to SA, has not yet been addressed. Interestingly, stable transgenic (RNAi) silencing of SIPK also resulted in elevated and prolonged WIPK activation levels, whereas ectopic over-expression of SEPK led to decreased WEPK activation (Samuel and Ellis, 2002). Thus SEPK appears to affect both defense outcomes and WEPK activation. WIPK silencing, either transient or stable, has not yet been achieved in tobacco, thus preventing observation of possible reciprocal effects of WIPK on SEPK activity. However, the patterns of expression and activation of SEPK and WEPK in response to different stresses make it likely that SEPK (perhaps via WEPK) plays a critical role in defining defense outcomes (Samuel and Ellis, 2002; Liu et al.,2003).  19  1.4 Problem Statements and Thesis Objective Problem statement Pathogens and/or elicitors appear to commonly activate a subset of highly conserved M A P K s , suggesting that these kinases may be convergence points in stress signalling. These M A P K s are likely to function in a network with other signalling components such as second messengers and hormones, but the functional relationship of the stress-responsive M A P K s to other signal components is largely undefined. A major limitation in assigning functional roles to M A P K s has been the inability to selectively inactivate particular kinases through inhibition or loss-offunction (knockout), but the availability of SIPK-silenced tobacco lines has made it possible to explore the role of SIPK in a specific stress response pathway.  20  Thesis objectives 1. to investigate the impacts of silencing of SIPK on the ability of tobacco plants to mount HR when challenged by a bacterial elicitor, harpin (hrpZpsph). 2.  to observe the impact of SIPK-silencing on SA and ROS accumulation in the context of hrpZpsph-induced HR.  3. to investigate the extent of SIPK involvement in mediating gene transcription during tobacco response to challenge from a bacterial elicitor  (hrpZp h), using sp  potato cDNA  microarrays. 4.  to compare the SIPK-dependent transcriptional responses to  hrpZp h with the responses sp  to an oomycete (Phytophthora megasperma) elicitor, P-megaspermin.  21  2 Salicylate-Induced Protein Kinase (SIPK) Regulates Harpininduced Stress Response in Tobacco 2.1 Introduction Time- and concentration-specific application of various purified harpins has been used in numerous experiments to examine associated early signalling events.  In tobacco cell  suspensions, hrpN (E. amylovord) has been shown to induce a biphasic ROS burst and medium alkalinization (K+/H+ exchange), two early (minutes-hours) indicators of pathogen-induced cell death and likely H R signalling events (Baker et al., 1993). The importance of ROS in enabling harpin-induced H R in Arabidopsis cell suspension cultures has been shown through reduction of hrpZ-induced H R by pharmacological inhibition of two potential ROS generator targets, N A D P H oxidase (0 ->0 ") and superoxide dismutase (0 "->H 0 ) (Desikan et al., 1996). 2  2  2  2  2  Harpins have been observed to induce several PCD-related processes, including inhibition of ATP synthesis (related to mitochondrial PT) (Xie and Chen, 2000), and expression of the cell death-related gene, HSR203J (Pontier et al., 1998). While harpin-induced responses have not been extensively studied in the context of hormones, Arabidopsis NahG transgenic plants failed to produce lesions and SAR upon harpin treatment, consistent with a general SA-dependence of elicitor-induced H R and SAR (Dong et al., 1999). Activation of protein kinases, and of M A P K s in particular, has been implicated as one of the early responses to harpin treatment.  Protein kinase inhibitors were shown to block hrpN-  induced medium alkalinization (Popham, 1995), cell death and defense gene activation (Desikan et al., 1999), indicating that protein kinase activity is required for activation of these processes. In addition, hrpPss-induced cell death and defense gene expression were inhibited by a M A P kinase-specific inhibitor, thus defining an essential role for one or more M A P kinase cascades in  22  hrpPss-induced H R (Desikan et al., 1999). Subsequent studies have allowed this definition to be refined. Harpin challenge has since been shown to lead to activation of specific plant M A P kinases. Harpin (hrpPss) treatment found to selectively activate Arabidopsis A t M P K 4 and AtMPK6 (Desikan et al., 2001b), while two tobacco M A P kinases, SIPK (orthologue to AtMPK6) and WfPK  (orthologue  to  AfMPO),  Pseudomonas .syringae-derived  were  simultaneously  activated  harpins (Zhang and Klessig, 2000).  by  treatment with  Taken together, these  observations suggest key roles for these M A P kinase species, particularly SIPK/MPK6, in mediating signalling events related to harpin-induced HR. To gain further insight into the role of SIPK in controlling the response to elicitation, we chose to focus on the interaction between a specific biotic stress, a hrpZ gene product hrpZp h sp  (from P.syringae pv. phaseolicola) and tobacco. The impact of this particular harpin on tobacco cells has been studied previously, where it was shown to form ion-conducting pores in cultured tobacco cells in vitro, perhaps reflective of its proposed biological role in pathogenicity (Lee et al., 2001b). In addition, a putative hrpZp h receptor has been identified on the plasmalemma of sp  tobacco plants (Lee et al., 2001a). Comparison of harpin-elicited N.tabacum var. Xanthi wild-type (WT) and SIPK-silenced lines (SIPK-Ri) demonstrated that SIPK-silencing renders tobacco plants more prone to hrpZp ph-induced H R (i.e. sensitive). Furthermore, my results indicate that SIPK normally acts S  as an upstream negative regulator of ROS accumulation, while both SIPK activation and ROS accumulation appear to operate upstream, or independent, of SA signalling. SIPK-silencing also resulted in hyper-accumulation of SA in hrpZp h-elicited tobacco cells, which suggests that sp  SIPK might normally play a negative regulatory role in the process of S A accumulation.  23  2.2 Methods 2.2.1  Plant material and treatment  Tobacco (Nicotiana tabacum cv. Xanthi-nc) plants of all genotypes were grown for six weeks in soil-less flood bench mix (3 peat : 1 perlite, v/v], watered twice daily with fertilized water (Scotts 15-5-15 Cal-Mag, stock #: 91940, total nitrogen = 120ppm), and provided with climatecontrolled greenhouse conditions (25/20°C, 16h light (supplemented seasonally)/8h dark). Six week-old plants were used for the treatments. The 3rd and 4th fully developed leaves were infiltrated with different concentrations of harpin (suspended in 5 m M MES) using a hypodermic syringe without needle. The infiltrated zones were either observed for lesion development, or harvested using a scalpel (for H2O2 assay), and stored at -80° C for further analysis. 2.2.2  Cell culture and treatment  To obtain cell cultures from WT and SIPK-Ri, and SEPK-OX plants, 1 cm leaf disks derived from 4 week-old plants were rinsed in EtOH for 30 seconds, soaked in 5.25% sodium hypochlorite (aq.) for 15 minutes then plated on medium containing 4.3g/l Murashige and Skoog (MS) salt mixture (Gibco, cat#l 1117-017), containing lOuM N A A , l p M kinetin, 30 g/1 sucrose, lg/1 M S vitamin solution (Sigma, cat#M3900), 1 g/1 Phytagel, and 3 g/1 agar. Calli were then transferred to the same genotype-specific solutions, lacking agar and Phytagel, and cultured in darkness at constant temperature (25°C) on a reciprocating shaker (50 oscillations per minute) to achieve cell suspension cultures. Seven- to ten-day cell cultures were routinely transferred to fresh medium. To test the effects of general ROS inhibition, suspension-cultured cells were pre-treated with N-acetyl cysteine (10 mM) (Sigma, cat#A9165) for 45 minutes followed by treatment with  24  harpin (2 ug/ml). The cells were harvested by vacuum filtration, frozen in liquid nitrogen and stored at -80°C for further analysis.  2.2.3  Generation of double mutant NahG/Ri (NGRi) lines  N G R i lines were created through fertilizing pistils of emasculated SIPK-Ri plants with NahG pollen. A n initial screening for the presence of the SIPK-Ri insertion containing the kanamycin resistance gene was achieved by plating F l seeds on M S agar with 50 p.g/ml kanamycin. Confirmation that F l lines contained both constitutively-active SIPK-Ri and NahG insertions was achieved through polymerase chain reaction (PCR), using a cauliflower-mosaic virus (CaMV) 35S-specific forward primer in combination with either SIPK-specific reverse primer (Samuel and Ellis, 2002) or NahG-specific reverse primer (5' GTC G C G C A A CTC G T A T A A CTC 3') (Gaffney et al., 1993). A l l experiments were performed on double-mutant F l tissues which were confirmed by PCR (as above). 2.2.4  Recombinant harpin purification  Recombinant harpin from E.coli BL-21 cells harboring the pT7-7 plasmid containing the D N A fragment encoding hrpZp h (Lee et al., 2001b) was purified essentially according to Lee et sp  al. (2001b) except 40% saturation of ammonium sulfate was used for the precipitation, and the following desalting and concentration was achieved through dialysis (14 -18 kD cut-off) in 5 m M MES. The protein content was quantified using the Bradford dye-binding assay 2.2.5  Protein extraction and protein gel blot analysis  For each sample tissue, 40 to 80 ug of frozen tissue was ground in liquid nitrogen and mixed with two volumes of extraction buffer [50mM HEPES pH7.5, 5 m M E D T A , 5 m M E G T A , lOmM DTT, I m M sodium orthovanadate (Na3V04), lOmM sodium fluoride (NaF), I m M phenylmethanesulfonyl fluoride (PMSF), 2u.g/ml antipain, 2ug/ml leupeptin, 10 ug/ml aprotinin, 5 ug/ml pepstatin, 10% glycerol, 7.5% polyvinylpolypyrrolidone (MW=111.14) (PVPP), and 25  one mini protease inhibitor tablet (Roche, cat# 11836170001) per 10 ml extraction buffer]. The mixture was incubated for 10 minutes at 4°C on a reciprocating shaker (100 oscillations per min), followed by centrifugation at 15,500 g. The supernatant was assayed directly or flash-frozen and stored at -80°C. The protein content was quantified using the Bradford dye-binding assay (Samuel et al., 2000). Proteins were boiled for 5 minutes with 5 X loading buffer (0.625 M Tris-HCl, pH 6.8, 5% SDS, 40% glycerol, 0.125% bromophenol blue, 40% v/v 0mercaptoethanol), fractionated by 10% SDS-PAGE, and transferred to P V D F (polyvinyl difluoride) membranes (Millipore, cat# IpvhOOOlO). A primary antibody dilution of 1:1000 was used for polyclonal phosphospecific anti-pERK (New England Biolabs), and a dilution of 1:5000 was used for both anti-SIPK and anti-WEPK (Seo et al., 1999; Y . Ohashi, personal communication). After blocking for 2 hours in 4% BSA, peroxidase-conjugated goat anti-rabbit (Dako, cat#AD0106) was used as the secondary antibody (1:5000). M A P K s were visualized using an enhanced chemiluminescence protocol (ECL-plus detection reagent, Amersham, cat# RPN2132) according to the manufacturer's directions.  2.2.6  Immunoprecipitation for in-gel kinase assay  Leaves on six week-old WT and SIPK-Ri plants were infiltrated with hrpZp h, harvested, and sp  tissue was ground in liquid nitrogen. Ground tissue (250 ug) was then extracted with 500 pi of extraction buffer (without PVPP) and centrifuged. combined with  either  immunoprecipitation.  The supernatant from this extract was  5 pg anti-SIPK antibody or 5 ug anti-WIPK  antibody for  After incubation overnight at 4°C, the immunoprecipitates were  recovered by incubation with 15 pi Protein-A-Sepharose suspension for 2 h, followed by centrifugation (15,500 g) for 1 minute at 4°C. The pellet was washed thrice with extraction buffer without PVPP, combined with 16 pi extraction buffer (without PVPP), mixed with 4 pi of  26  5x loading buffer (as described in 2.2.5), and then boiled for 5 minutes. The released proteins were used for in-gel kinase assays. 2.2.7  In-gel kinase assay  Immunoprecipitated proteins (80 ug per sample) were electrophoresed on 10% SDSpolyacrylamide gels embedded with 0.1 mg/ml myelin basic protein (MBP) in the separating gel as substrate for the kinase. After electrophoresis, SDS was removed by washing the gel with washing buffer (25 m M Tris, pH 7.5, 0.5 m M DTT, 0.,1 mm N a V 0 , 5 m M NaF, 0.5 mg/ml 3  4  BSA, and 0.1% Triton X-100[v/v]) three times for 30 minutes each at room temperature. The kinases were allowed to renature in 25 m M Tris, pH 7.5, 1 m M DTT, 0.1 m M Na3V04, 5 m M NaF at 4°C overnight with three changes of buffer.  The gel was then incubated at room  temperature in 30 ml reaction buffer (25 m M Tris, pH 7.5, 2 m M E G T A , 12 m M M g C l , I m M 2  DTT, and 0.1 m M N a V 0 ) with 200 n M ATP plus 50 uCi [y- P]-ATP (3000 Ci/mmol) for 60 32  3  4  minutes. The reaction was stopped by placing the gel in 5% trichloroacetic acid (w/v) / 1% NaPPi (w/v). The unincorporated [y- P]-ATP was removed by washing in the same solution 32  for at least 6 hours with five changes. The gel was then dried onto Whatman 3 M M paper and exposed to Kodak XAR-05 film. Pre-stained size markers (Bio-Rad) were used to calculate the size of the kinases. 2.2.8  In situ staining for hydrogen peroxide  Hydrogen peroxide was visualized  in Situ  by 3,3-diaminobenzidine (DAB) staining  performed essentially according to Torres et al. (2002). Leaf halves were collected after 4 hours exposure to 0.5 pg/ul hrpZp h and vacuum infiltrated with the lpg/ul D A B solution. Infiltrated sp  leaves were placed under high humidity and fluorescent light until brown precipitates were observed (5 to 6 hours) and then fixed with a solution of ethanol:lactic acid:glycerol (3:1:1, v/v) for 2 days, followed by further clearing in 95% methanol. 27  2.2.9  S A measurements  For each time point (0, 24, 48 hours), two harpin-infiltrated zones of tissue from 3 and 4' rd  leaves of 6 week old plants were pooled together to comprise a biological replicate. Two biological replicates for each genotype at each time point (approximately 0.5-1 g of tissue per sample) were then used for measuring free and total SA as described previously (Li et al., 1999). Unless otherwise indicated, all experiments were repeated, with consistent results.  28  2.3 Results 2.3.1 SIPK-silencing leads to heightened sensitivity of tobacco leaves to hrpZ  Psph  Purified hrpZpsph (40 (ig/ml) has previously been shown to elicit a H R in N. tabacum cv. Samsun N N two days after infiltration (Lee et al., 2001a). We confirmed this response, finding that treatment of N. tabacum var. Xanthi leaves with 0.5 mg/ml hrpZp hin resulted in the sp  induction of the H R response (Figure 2.IB).  Thin, translucent regions appeared within the  infiltrated zone 8-12 hours after harpin treatment (0.5 mg/ml) (Figure 2.1 A), suggestive of the cellular collapse associated with bacterial and fungal HRs (Wei et al., 1992)(Heath, 2000b). To investigate the potential influence of SIPK on this cell death outcome, leaves on SIPK-Ri plants were infiltrated with 0.5mg/ml harpin, and the response compared to that of similarly infiltrated WT leaves. In the SIPK-Ri line, tissue translucence occurred in a greater proportion of the infiltrated zone than in WT (12 hours post-infiltration), while the lesion area was also more extensive in SIPK-Ri than in WT leaves by 3 days post-infiltration. No lesions formed as a result of control (5mM MES) injections (Figure 2.1 A).  This result suggests that SIPK may  normally play a negative regulatory role in hrpZp h-induced HR. sp  29  5mM  MES  0.5ug/ml hrpZ  Figure 2.1 HR development in WT and SIPK-silenced (SIPK-Ri) leaves upon harpin treatment. (A) Cell collapse in WT and SIPK-silenced (SIPK-Ri) tobacco leaves 12 hours after infiltration with 5 m M MES or 0.5 ug/ul hrpZpsph. (B) Lesion development in tobacco leaves 3 days after infiltration. Black lines in ' A ' mark the infiltration boundary. Injection sites (red) ~ 4 mm. Since absence of SEPK appears to affect the sensitivity of tobacco plants to  hrpZ h, Psp  we  wanted observe the pattern of activation of both SEPK and WEPK in WT and SIPK-Ri tissue during the harpin response.  A 48 kDa protein has previously been shown to be transiently  activated (5-30minutes) after  hrpZ h treatment of tobacco Psp  cells (using anti-pERK antibodies),  and its identity as SIPK has been verified through immunoprecipitation (Lee et al., 2001a) using the SEPK-specific antibody Abp48N (Zhang et al., 1998). In the present experiments, I observed rapid and prolonged activation of two M A P K s in harpin-challenged WT plant tissue, as determined by western blot using anti-pERK antibodies (Figure 2.2A). These M A P K s were subsequently identified as SEPK and WIPK through immunoprecipitation using SIPK- and 30  WrPK-specific antibodies (Seo et al., 1999) and subsequent in-gel kinase assay of the 2 hour leaf tissues (Figure 2.2B). In the SIPK-Ri line, only a low level of activated SIPK could be detected by anti-pERK western blot, following similar treatment and extraction as for WT tissue (Figure 2.2A). However, SIPK activity was not detectable in the SIPK-Ri tissue by in-gel kinase assay, in contrast to the strong SIPK kinase activity observed in the WT protein extract (Figure 2.2B). More prolonged WIPK activation was also observed in the SIPK-Ri tissue in comparison to WT, where WIPK activation declined sharply 8 hours after harpin challenge.  31  UT  WT SIPK-Ri 1h 2h 4h 8h 24h~ UT 1h 2h 4h  8h 24h |  Figure 2.2 M A P K activation profiles of W T and S I P K - R i in hrpZ  Psph  — SIPK WIPK  -induced H R .  (A) Anti-pERK antibody was used to probe proteins extractions from WT and SIPK-Ri untreated plant tissue (UT) as well as 1,2,4,8, and 24-hours after cell culture treatment with 0.5 ug/ul hrpZp ph. (B) Immunoprecipitation of 2-hour plant protein extractions using anti-SIPK and anti-WIPK was followed by in-gel kinase to confirm SIPK and WIPK activities. S  2.3.2  SIPK-silencing leads to increased ROS accumulation following hrpZ treatment  Psph  Intracellular and extracellular ROS accumulation can occur in plant tissues within minutes after treatment with various harpins.  Sustained ROS elevation has been shown to be a  requirement for many pathogen-induced HR responses through ROS scavenging experiments (Chapter 1).  Since exogenously-applied ROS have been shown to induce the activation of  SIPK in tobacco cell suspension culture (Samuel et al., 2000), and protein kinase inhibitors block hrpN-induced H R (Desikan et al., 1999), ROS appear likely to act as up-stream activators in SIPK-mediated cell death. To test the ROS-dependency of harpin-induced SIPK activation, WT (and SIPK-silenced) cell suspension cultures were treated with lOmM N-acetyl cysteine (NAC, a general antioxidant) prior to treatment with 2 ug/ml  hrpZ  Psp  h  (Figure 2.3A). Pre-treatment with N A C did not fully  suppress activation of SIPK by harpin in WT cultures, but activation of WIPK was suppressed in 32  both the WT and SIPK-silenced lines. One interpretation of this result is that at least one alternative harpin-induced SIPK activation pathway/mechanism may exist in tobacco cells. On the other hand, activation of WIPK by harpin treatment appears to be fully dependent on ROS intermediaries. Given that ROS accumulation is likely accentuated through positive feedback loops involving other HR-related signals (Jabs, 1999), we assessed the potential for SIPK activation to affect ROS accumulation in response to harpin-treatment.  WT and SIPK-Ri leaves were infiltrated  with 0.5 ug/ul hrpZpsph, then histochemically stained for H2O2 accumulation 4-hours after hrpZpsph  treatment (Figure 2.3B). SIPK-silenced leaves exhibited more extensive staining within  the infiltration zone than did WT leaves, indicating that the suppression of SIPK protein and associated SIPK activation levels in the SIPK-Ri line resulted in greater H2O2 accumulation within apoplastic and/or symplastic spaces. The local regions of darkest staining observed within the WT infiltration zone are artefacts of tissue bruising incurred during the staining process.  33  A  WT  SIPK-Ri + NAC 20 mM  +  — SIPK — WIPK  WT  SIPK-Ri  B  Figure 2.3 R O S and tobacco M A P K interaction during tobacco tissue response to h r p Z p . sph  (A) SIPK and WIPK activation 30 minutes after treatment of cell suspension cultures with 2 pg/ml hrpZp ph in the presence (+) or absence (-) of ROS scavenger N A C . (B) In situ detection of hydrogen peroxide accumulation, using D A B staining (browning), in WT and SEPK-Ri leaves 4-hours after treatment with 0.5 pg/ul hrpZ h . Images are representative of genotypic responses. Injection sites (red) ~ 4 mm. S  Psp  2.3.3 Salicylic acid (SA) is a requirement for hrpZ -induced HR, but not for SIPK (or WIPK) activation Psph  SA has been shown to be required for harpin-induced H R in Arabidopsis (Dong et al., 1999). While the tobacco M A P K , SIPK, has been named according to its SA-induced activation (Zhang and Klessig, 1997), SIPK activation has, in fact, been shown to occur in an SA-independent manner following treatment of tobacco cells with cryptogein (Lebrun-Garcia et al., 2002) or tobacco leaves with ozone (Orvar et al., 1997).  We compared the response to harpin of a  transgenic tobacco (var. Xanthi) line expressing NahG with the response of WT plants, to confirm the requirement for S A signalling in the hrpZ h-induced HR. In contrast to either the Psp  34  WT or SIPK-Ri lines, the infiltration of NahG tissues with harpin (0.5 mg/ml) resulted in no lesion formation over a 4 day period of observation (Figure 2.4A). The levels of SIPK and WIPK activation induced in the NahG lines 30 minutes after infiltration with 0.5 mg/ml hrpZp  sph  appeared similar to those observed in WT tissue (see Figure 2.IB), indicating that SA is not required for hrpZp h-induced SIPK activation (Figure 2.4B). sp  To determine i f SIPK-silencing could possibly compromise the relative insensitivity of NahG plants to  hrpZp h, sp  a cross was generated between SIPK-Ri (maternal line) and NahG plants.  The resulting progeny, which possessed both the NahG trait and the SIPK-silencing R N A i construct, were designated N G R i . Lesion development in leaves of the N G R i line following harpin challenge most closely phenocopied that of NahG line (Figure 2.4A). Likewise, SEPK and WIPK activation profiles of N G R i plants phenocopied that of SIPK-Ri plants at the 24-hour time point (Figure 2.4C). Together, these data confirm that SA is necessary for the  hrpZ  Psp  h-  induced H R of N.tabacum var. Xanthi, and that this effect acts downstream, or independent, of hrpZp h-induced SIPK activation. sp  35  o  SIPK-Ri  WT  NatiG  J  B NahG  Harpin  UT  c  • SIPK I— WIPK  24h  Psph  y  48h  24h  48h  | — SIPK - WIPK  NGRI  Figure 2.4 S A and S I P K affects on hrpZ  NGRI Harpin  MES  MES  A  f  ip  n\  -induced H R .  (A) Lesion formation in WT, SIPK-Ri, NahG, and NahG x SIPK-Ri (NGRi) leaves 4 days after infiltration with 0.5 ug/ml hrpZp h. (B) SIPK and WIPK activation profiles in NahG 30 minutes after 5 m M M E S or 0.5 ug/ml hrpZp h treatments. (C) SIPK and WIPK activation profiles in N G R i plants for untreated (UT), 5 m M M E S (24 and 48 hours), and 0.5 ug/ml hrpZp (24 and 48 hours). Dotted lines indicate injection site (circles ~ 4 mm), and/or infiltrated zone boundaries. Images are representative of genotypic responses. sp  sp  sph  2.3.4  SIPK-silencing leads to increased SA levels  SIPK appears to play a negative regulatory role in hrpZp h-induced HR, as evidenced by the sp  hypersensitivity of SIPK-Ri plants to harpin treatment. Since SIPK also appears to act up-stream, or independently, of SA in hrpZ h-induced HR, it seemed possible that SIPK might negatively Psp  regulate SA accumulation, and thereby influence H R induction. To explore this possibility, we measured total SA levels in untreated WT and SIPK-Ri tissues, and compared these to SA levels in tissues collected for each genotype 24 and 48 hours after being treated with 0.3 ug/ul hrpZp h . Harpin treatment led to marked increases in SA sp  levels in both genotypes (Figure 2.5), but at both the 24- and 48-hour time points, SA levels in the SIPK-Ri line were approximately two-fold higher than those measured in WT tissues. As expected, given the presence of the NahG enzyme in their tissues, neither NahG nor N G R i plants accumulated substantial amounts of SA over the treatment course (data not shown).  36  \  o SIPK-Ri  WT  NGRI MES  B NahG  MES  Harpin  UT  c  | — SIPK WIPK  24h  Psph  48h SIPK WIPK  NGRI  Figure 2 . 4 S A and S I P K affects on hrpZ  Harpin  48h 24h  -induced H R .  (A) Lesion formation in WT, SEPK-Ri, NahG, and NahG x SIPK-Ri (NGRi) leaves 4 days after infiltration with 0.5 ug/ml hrpZ h. (B) SIPK and WIPK activation profiles in NahG 30 minutes after 5 m M MES or 0.5 ug/ml hrpZp p treatments. (C) SIPK and WIPK activation profiles in N G R i plants for untreated (UT), 5 m M MES (24 and 48 hours), and 0.5 ug/ml h r p Z (24 and 48 hours). Dotted lines indicate injection site (circles ~ 4 mm), and/or infiltrated zone boundaries. Images are representative of genotypic responses. Psp  S  h  Psph  2.3.4  SIPK-silencing leads to increased SA levels  SEPK appears to play a negative regulatory role in hrpZ h-induced HR, as evidenced by the Psp  hypersensitivity of SIPK-Ri plants to harpin treatment. Since SEPK also appears to act up-stream, or independently, of SA in hrpZ  Psprr  induced HR, it seemed possible that SIPK might negatively  regulate SA accumulation, and thereby influence HR induction. To explore this possibility, we measured total S A levels in untreated W T and SEPK-Ri tissues, and compared these to SA levels in tissues collected for each genotype 24 and 48 hours after being treated with 0.3 pg/ul hrpZ  Pspri  . Harpin treatment led to marked increases in SA  levels in both genotypes (Figure 2.5), but at both the 24- and 48-hour time points, SA levels in the SEPK-Ri line were approximately two-fold higher than those measured in WT tissues. As expected, given the presence of the NahG enzyme in their tissues, neither NahG nor N G R i plants accumulated substantial amounts of SA over the treatment course (data not shown).  36  3000  2500 0) 3  o TO <  •*-> o  2000 • WT  1500  SRI  1000  500  UT  24 hrs  48 hrs  0.3 ug/ul h r p Z  Psph  Treatment Figure 2.5 HrpZp h-induced SA accumulation in WT and SIPK-Ri leaves. sp  Total S A was measured in leaf extracts taken from WT and SIPK-Ri tissues that were treated with 5mM M E S or with 0.3 ug/ml hrpZ h and harvested at 24 or 48 hours after infdtration. Error bars represent range, n = 2 (see page 28 for details). Psp  37  2.4 Discussion 2.4.1  SIPK has a negative regulatory role in hrpZp -induced HR sph  SIPK clearly plays an important role in mediating defense outcomes in tobacco, particularly HR (Zhang and Liu, 2001; Samuel, 2002), and the present study extends our knowledge of SIPK involvement in H R by correlating SIPK activation with changes in two key HR-related signals, ROS and SA, in the context of responses to hrpZp h. sp  Since stable silencing of SIPK renders tobacco plants more prone to hrpZp h-induced lesion sp  formation, I conclude that SIPK activity normally plays a role in reducing the extent of H R proliferation in elicitor-challenged tissues.  In support of this idea, lesion development was  found to be less severe than WT in a transgenic line over-expressing SIPK following harpin treatment (data not shown). In contrast to these findings, suppression of SIPK activation in tobacco through P V X virusinduced gene silencing (VIGS) technology resulted in increased susceptibility of the plants to T M V infection (Jin et al., 2003).  This implies that SIPK influences H R in a completely  different manner during the response to the bacterial elicitin (hrpZp ) and to a viral challenge. spn  In a similar vein, it has been observed that tobacco lines over-expressing SIPK are more sensitive to inoculation with P.syringae  pv phaseolicola strains than are WT plants, independent  of the presence or absence of the hrpZp h gene in the Pseudomonas sp  strain (Samuel et al., 2005).  This indicates that SIPK activation may affect the plant responses to both hrpZ h-dependent Psp  and -independent signals from a single pathogen. While no group besides ours has yet reported stable SIPK-silencing in tobacco, stable transformants have been obtained for a constitutivelyactivated upstream M A P K K (NtMEK2) that is capable of activating SIPK (Liu et al., 2003). Treatment of plants carrying this steroid-inducible construct with dexamethasone  led to  formation of spontaneous lesions in the absence of any applied stress, a response that would be 38  consistent with the hypothesis that activated SIPK is a positive regulator of cell death. This is, of course, contradictory to the earlier conclusion that SEPK normally suppresses cell death during the elicitor-induced HR, although it is unclear whether over-expression of Ca-NtMEK2 might also be influencing downstream targets outside of SIPK. Thus, while SIPK appears to play a significant role in mediating biotic stress-induced HR, the available data still do not support a single model for the mode of action of this regulation. It seems likely that the frequently conflicting outcomes observed by different research groups reflect both the complexity of SIPK's function(s) and the diversity of those functions across different experimental systems.  2.4.2  SIPK and WIPK interact during the hrpZ -elicited response Psph  Our study supports previous findings that SIPK transcription is largely unresponsive to stress, whereas the SIPK gene product is rapidly activated minutes after stress treatment (Zhang et al., 1998; Samuel and Ellis, 2002; Takahashi et al., 2003). Even in apparently unstressed tissues, SIPK exhibits a basal level of activation (Liu et al., 2003), yet our studies show that SIPKsilencing nearly abolishes this basal SIPK activation with no apparent phenotypic differences between unstressed WT and SIPK-Ri plants. However, cell suspension cultures derived from these WT and SIPK-Ri plants do exhibit differences in the (apparently) unstressed state, suggesting that this basal SIPK activity has a role in cellular homeostasis (see Chapter 3 discussion). Our findings indicate that activation of a second, lower molecular-weight M A P K , (identified as WIPK), occurs as early as one hour post-infiltration with harpin.  B y contrast, WIPK  activation has been reported to occur nearly two hours after treatment with oomycete cell wallderived elicitors, and this activation was shown to coincide with H R induction (Zhang et al., 2000).  In our hands, WIPK activation arising from either hrpZ h (this study) or ozone Psp  treatment (Samuel and Ellis, 2002) occurs prior to development of any visible H R symptoms. It 39  has also been proposed by Zhang and colleagues that VIGS silencing of SIPK leads to reduced levels of WEPK protein accumulation during T M V infection, a result they interpreted as indicating that SEPK positively regulates WIPK transcription and/or translation (Jin et al., 2003). While the two-hour delay in activation observed by Jin et al. (2003) might be consistent with such a model, we have observed strong WIPK activation occurring as early as ten minutes after elicitor infdtration of leaf tissue (data not shown), or 30 minutes after elicitor treatment of cell suspension cultures (Figure 3.3).  In addition, we found that silencing of SEPK did not impede  hrpZp ph-induced activation of WEPK and, instead, appears to result in prolonged WEPK S  activation (Figure 2.2).  This implies that SEPK may negatively regulate sustained WEPK  activation. If the hypothesis that WEPK is a positive regulator of cell death is correct, elevated and/or prolonged WIPK activation in the SIPK-Ri line could account for the heightened sensitivity of SEPK-Ri lines to hrpZp h infdtration (Figure 2.1). sp  In support of a negative  regulatory role for SIPK in hrpZp h-induced HR, over-expression of SIPK results in a dramatic sp  reduction of WEPK activation levels and increased resistance to hrpZp h treatment (data not sp  shown). While both our present study and that of Liu et al. (2003) suggest that WEPK has a positive regulatory role in HR, it is important to emphasize that this association remains entirely correlative. It is also worth noting that our findings do not preclude direct de novo WEPK activation occurring as a result of elicitor treatment. However, it seems unlikely that WEPK activation would be delayed until the onset of H R two hours post-infiltration or that SEPK activation is a requirement for WEPK activation in all experimental contexts as Jin et al. (2003) suggest. Clarification of this issue will require effective silencing or inactivation of WIPK which would make it possible to determine i f WEPK is essential for the induction of SEPKmediated HR.  40  Under each experimental condition, it is clear that SIPK and WIPK activation are tightly regulated over time. However, their activation levels are ultimately a reflection of the signalling mechanisms that SIPK and WIPK mediate. The fact that the pattern of activation of four alfalfa M A P K s appears specific to the type of elicitor treatment (Cardinale et al., 2000) suggests that temporal and/or spatial patterning of M A P kinase protein and/or activation levels may, themselves, be important signalling mechanisms, similar to roles hypothesized to be played by calcium and oxylipin signatures in conditioning elicitor-specific responses.  2.4.3  SIPK negatively regulates hrpZ -induced ROS accumulation Psph  While it is an interesting possibility that SIPK might act through WIPK in affecting cell death, it is very likely that SIPK also has other routes through which it is also able to influence cell fate. Our results with respect to hrpZp h-induced ROS accumulation revealed that SIPK activation sp  by this elicitor requires the presence of ROS, as demonstrated by the ability of a general ROS scavenger, N A C , to largely block the activation (Figure 2.3). This is in contrast to published reports that catalase (an  H 2 O 2  scavenger), or diphenylene iodonium (DPI, an inhibitor of  N A D P H oxidase) both failed to inhibit the activation of AtMPK4/AtMPK6 in Arabidopsis cells treated with harpin (Desikan et al., 2001b). While it is not clear that exogenous application of a cell-impermeant reagent such as catalase would be expected to influence internal  H 2 O 2  levels,  these results led Desikan et al. (2001a) to suggest that harpin induces two defense response pathways, one involving activation of AtMPK4/AtMPK6 and the other leading to ROS production (Desikan et al., 2001b). However, this model of discrete pathways seems less likely since ROS (H 02) produced by ozone challenge is also capable of rapidly inducing SIPK 2  activation (Samuel and Ellis, 2002) and at least some ROS formation/accumulation appears necessary for SIPK and WIPK activation by harpin (Figure 2.3). Although hrpZpsph-induced ROS formation is a requirement for SIPK and WIPK activation in harpin-challenged cells, SIPK, in turn, appears to affect ROS (especially 41  H 2 O 2 )  levels. In fact,  SIPK-silencing renders tobacco plants prone to hrpZp h-induced H2O2 hyper-accumulation sp  (Figure 2.3), suggesting that SIPK plays a negative regulatory role in hrpZp h-induced H2O2 sp  accumulation. Given evidence that ROS scavenging attenuates (but does not eliminate) hrpZinduced H R (Desikan et al., 1996), SIPK (or WIPK) modulation of H 0 levels may provide one 2  2  possible mechanism of regulating cell fate. While it is uncertain whether the H2O2 burst or SIPK activation is first initiated by  hrpZ  Psp  h  treatment, a negative feedback loop involving SIPK and H2O2 seems plausible. Since WIPK expression appears more H202-affected than SIPK expression, and an increase in WIPK expression correlates with increased H 0 2  2  accumulation in the SIPK-Ri line, WIPK and H2O2  (or ROS in general) may be related in a positive feedback loop whereby WIPK activation contributes to increased ROS levels.  The sustained activation of WIPK following hrpZp h sp  treatment of tobacco tissues may provide a mechanism to enable the second stage of the biphasic ROS burst which has been observed to accompany elicitor-induced H R inductions (Baker et al., 1993; Dorey et al., 1999). Thus, ROS appear to play a significant role in mediating hrpZ  Psph  -  induced SIPK and WIPK activation, while SIPK also appears to negatively regulate ROS (H2O2) accumulation. This model would predict that silencing of WIPK would inhibit elicitor-induced ROS accumulation, particularly at time points that coincide with the onset of HR.  2.4.4  SIPK negatively regulates SA accumulation  HrpZp -induced H R appears to be SA-dependent as demonstrated by the abolition of lesion spn  formation in the NahG genotype. While these findings agree with the general observation that SA is a requirement for H R and SAR, crosses of lesion-mimic mutants with NahG lines in Arabidopsis have established that cell death can occur independently of SA (Lam, 2004). Potentially, SIPK and S A are acting on separate, discrete pathways leading to hrpZ h-induced Psp  cell death. Given that loss of SIPK has such a dramatic effect on lesion formation, while lesion formation is as suppressed in the N G R i double mutant as strongly as in the NahG line, SA is 42  likely acting downstream (and not independently) of SIPK activation in mediating HR. In fact, we find that SIPK and WIPK activation profiles for NahG plant are effectively identical to those o f W T (Figure 2.4B). Cryptogein (Phytophthora cryptogea) also induced strong SIPK and WIPK activation in an SA-independent manner (Lebrun-Garcia et al., 2002). As Lebrun-Garcia et al. (2002) suggest, this result is perhaps not surprising given that the rapid activation of SEPK and WEPK occurs well in advance of elicitor-induced SA accumulation, which occurs several hours to days later. Ironically, these findings contradict the first characterization of the SIPK protein as "salicylic acid-induced" (Zhang and Klessig, 1997). In support of the hypothesis that SEPK is upstream of SA accumulation in mediating hrpZ hPsp  induced HR, SEPK-silencing leads to a marked increase in elicitor-induced S A accumulation, relative to WT, over a 24 hour period. Interestingly, SA accumulation was not reduced (from levels induced in WT) through over-expression of SEPK (data not shown), suggesting that SEPK must negatively influence commitment to cell death in the context of hrpZ h challenge (as Psp  observed in Figure 2.1) through means other than reduction of SA accumulation.  2.4.5  Conclusion  In conclusion, SEPK activation has a striking impact on whether or not tobacco leaves challenged with harpin commit to HR. This impact can be partly explained by the effect that SEPK-silencing has on activation of the HR-associated M A P kinase, WIPK. SEPK and WEPK likely act together in regulating elicitor-induced cell death, but this relationship is difficult to dissect without access to genotypes with stably-modified WEPK expression and/or activation. Consistent with the proposed negative regulatory role for SEPK during hrpZ h-induced HR, and Psp  the apparent positive regulatory roles of ROS and SA on HR, SEPK-silencing led to hyperaccumulation of both ROS and S A in elicitor-challenged tobacco tissues. It thus seems 43  likely that SIPK mediates HR, at least partly, through manipulation of these two cell death determinants.  44  3 Microarray Analysis of SIPK-dependent Transcriptional Events in Biotic Stress-induced Tobacco Cell Suspension Cultures  3.1  Introduction  Plants exhibit many stress-specific defense responses, including involvement of particular signalling pathways, unique transcriptional and metabolic changes, and, ultimately, cell fate outcomes.  Parallel investigations of two or more different stresses can provide insights into  both common and different components of the associated stress-induced responses (Chinnusamy et al., 2004). Bacterial hrpZ h and oomycete p-megaspermin (an alkaline 98 amino acid-long Psp  glycoprotein elicitin derived from Phytophthora megasperma) represent two distinct biotic stresses against which plants respond with a variety of defenses, both common and distinct. Both P-megaspermin and hrpZp h induce a hypersensitive response (HR) as well as production sp  of phytoalexins, PR proteins and ethylene in tobacco tissues (Baillieul et al., 2003). However, hrpZp h has been shown to bind to a plasmalemma-associated receptor in tobacco plants, while sp  P-megaspermin does not appear to bind to this receptor (Lee et al., 2001a).  In addition, P-  megaspermin requires an influx of calcium ions from the extracellular space to the cytosol in order to activate the tobacco M A P kinase, SIPK, and induce the expression of the H R marker, HIN1, whereas hrpZp  spn  does not (Lee et al., 2001a). Thus, while both elicitors induce rapid  activation of M A P kinase signalling, and both can trigger superficially similar defense responses, they appear to do so via mechanisms that may be at least partially distinct. While stress-specific transcriptional response profiles have been revealed in some multiplestress studies (Schenk et al., 2000; Cardinale et al., 2002), the signalling pathways that potentially distinguish these responses remain largely unexplored. For the most part, the cellular targets of activated M A P K s in plants remain undefined, although knowledge of the regulatory 45  roles played by activated M A P K s in mammalian cells suggests that, in tobacco, activated SIPK is likely to affect specific transcription factors.  This suggestion is consistent with the  observation that over-expression or silencing of SIPK in tobacco has striking impacts on cell fate, as well as ROS and SA accumulation (Chapter 2). Direct action on transcriptional events is also consistent with the finding that SIPK (and its Arabidopsis orthologue, AtMPK6) appears to migrate from the cytosol to the nucleus upon perception of stress (Clegg, 2004; Ahlfors et al., 2004). If SIPK is a modulator of transcriptional activity, comparison of the transcriptome in short-term elicited wildtype and SIPK-suppressed tobacco cells would be predicted to reveal the primary transcriptional events controlled by elicitor signals passing through SIPK. Furthermore, by using two different elicitors, it should be possible to establish which transcriptional responses are common to both elicitors and which are unique to a specific agent. Finally, the comparison of each elicitor across both tobacco genotypes enables information to be gained as to which of the elicitor-specific responses may be reliant on SIPK signal transduction. To address these issues, I assayed the transcriptional responses of WT and SIPK-Ri tobacco cell suspension cultures to either P-megaspermin or hrpZp h at two time points (4 and 8-hours). sp  No large-scale transcript profiling tools are publicly available for tobacco, but The Institute for Genomic Research (TIGR) has developed a "Solanaceae" cDNA microarray printed with 10,000 duplicate EST clones that were derived from EST libraries prepared from diverse potato tissues. Because of the close evolutionary relationship between Nicotiana and Solarium genera, these printed potato cDNAs are expected to hybridize efficiently to most homologous sequences of the corresponding tobacco genes (Ronning et al., 2003).. M y results demonstrate that SIPK-silencing has a significant impact on the tobacco transcriptome, even prior to stress, but that SIPK-silencing had very different effects on the transcriptional responses arising from treatment of cell suspension cultures with either hrpZp hsp  and P-megaspermin.  A filtering regime was developed that made it possible for me to identify 46  genes whose behaviour in response to hrpZp h and P-megaspermin could be distinguished. sp  Finally, gene ontology analysis then permitted me to initiate characterization of SIPK-dependent functional differences in these responses.  3.2 Materials 3.2.1  and  Methods  Cell culture and treatment WT and SIPK-Ri cell cultures were derived and maintained as described in Chapter 2.  Four-day old (log-phase) cell suspension cultures of WT and SIPK-Ri line were used for all experiments. To test the effects of hrpZp h and P-megaspermin treatment, suspension-cultured sp  cells of the same genotype were pooled into 60 ml portions, divided into 25 ml homogeneous replicates for each prescribed treatment time (including control), and kept on the orbital shaker at 60 oscillations per minute for a 30 minute acclimation period before elicitor solutions were administered by pipette. After the prescribed treatment period, the cells and medium were separated by vacuum filtration. The cells were frozen in liquid nitrogen and stored at -80°C to await further analysis.  3.2.2  Preparation of megaspermin  P-megaspermin (217nM) purified from Phytophthora megasperma (H20) was generously donated by the Serge Kauffmann Laboratory (University of Reims, France) (Baillieul et al., 2003).  3.2.3  Medium alkalinization  At prescribed time points, the pH of each 25 ml portions of 4-day old cell cultures (maintenance and treatment previously described) was measured by maximal submersion of the Accumet pH electrode (Fisher Scientific) after the reading stabilized.  47  3.2.4  RNA  total  isolation (for expression profiling)  After the prescribed treatment time had elapsed,  RNA t i t o  a  was extracted from control and  treated tissues for WT and SIPK-Ri using a modified TRIZOL extraction method, as follows. Approximately 1.5 g plant material was ground in liquid nitrogen using a mortar and pestle, resuspended in 4 ml TRIZOL reagent (Invitrogen, cat#l 5596-026), vortexed and incubated at room temperature for 5 minutes with regular mixing. The sample was then vigorously mixed with 0.8 ml chloroform, and allowed to sit at room temperature for 2-3 minutes before cell debris was pelleted by centrifugation for 30 minutes at 12,000 g and 4°C. The aqueous phase was recovered and R N A was precipitated at room temperature for 10 minutes with 2 ml isopropanol. After centrifugation for 10 minutes at 12,000 g, the pellet was washed with 2 ml 75% ethanol and re-centrifuged. The  RNA tai to  pellet was re-centrifuged for 5 minutes at 10,000  g before being air dried for 5-10 minutes (until white) and resuspended in 50 ul RNAse-free water.  3.2.5  RNA labeling, microarray hybridization, and scanning  Following a spectrophotometric determination of R N A concentration (GeneQuant 260/280) and assessment of R N A t o t a i degradation (Agilent bioanalyzer), 120 ug  RNAtotai  (per sample) was  submitted in triplicate to The Institute for Genomic Research (TIGR) Solanaceae Gene Expression Profiling Service (http://www.tigr.org/tdb/potato/ RNAtotai  profiling service2.shtml) for  labelling, microarray hybridization, and microarray scanning. For SIPK-Ri / WT co-  hybridizations, replicate WT and SIPK-Ri samples were randomly paired (see Results section 3.3.4 for experimental design details). For WT sample was co-hybridized to pooled  WT  untr  tre  ated  eated  / WT  sample.  untr  eated  co-hybridizations, each  WT ated tre  Following labeling of mRNA using a  direct labeling method (TIGR Protocol 1), c D N A were then hybridized on to potato cDNA (10K) microarrays  (http://www.tigr.org/tdb/potato/microarray comp.shtml)  according  to  TIGR  Protocol 1. In accordance with TIGR policy, the WT samples were labeled with cyanine-5 dye, 48  while the SIPK-Ri samples were labelled with cyanine-3 dye. A dye swap on two of my samples was performed with no replication. Dried slides were scanned with the GenePix Array Scanner (www.axon.com).  3.2.6  Global expression profiling analysis (microarray)  Spots were identified and quantified by TIGR using the GenePix Pro software, and settings recorded in supplementary raw data files associated with this experiment (link to TIGR data set ".gpr" files). Poor spots that were manually flagged (ID=0) were not used in further data analyses and remaining potato array elements were analyzed using customized scripts for R and Bioconducter (The R Development Core Team, www.r-project.org).  Background correction and signal normalization  For background correction, we defined the mean background signal of all spots in a particular subgrid as the background for that subgrid. This mean background value was then subtracted from the signal value for each spot in the subgrid. I normalized each array using the robust local-linear regression algorithm LOWESS (or 'LOESS') included in the R package, with a span of 0.7 (Yang and Speed, 2002).  Detection of significant differential expressions within treatments  For each element, I first used the log2-transfomed, LOWES S-normalized ratios for each duplicate spot (array element) from the three replicate arrays (total of 6 observations) for each sample to perform a paired Student's t-test using the Welch approximation to the degrees of freedom. Log2 mean ratios for each array element were calculated for each set of 6 spots for an array clone. This method is applied in Results sections 3.3.4 and 3.3.5 (WT eated/WT treated tr  hybridizations only) where experimental design details are provided.  49  un  Statistical tests between gene expression ratios of different treatments  The standard R script for analysis of variance ( A N O V A ) was employed to identify genes with significant differences in mean gene expression ratios between at least two of control and treatments (n=6; 2 duplicate spot/hybridization, 3 hybridizations/treatment). To determine the significant differences in SEPK-Ri/WT ratios between two treatments in particular (as in Section 3.3.5), two-sample t-tests were performed using the standard t-test algorithm adjusted for two samples. Experimental design details are provided in Results section 3.3.4. For all statistical tests, p-values were adjusted to restrict the false discovery rate (the probability that a gene that has been called differentially expressed is indeed not differentially expressed) to <0.01 using the R method "fdr" (Benjamini and Hochberg, 1995).  Pearson  correlation coefficients between the stated slide or treatment mean expression ratios were calculated in Excel.  Experimental design details are provided in Results section 3.3.5 (SIPK-  Ri/WT hybridizations only).  Gene annotation  For the ANOVA-filtered set, annotations were derived from raw data (.gpr) files obtained from TIGR for the experiment (link to web). For Table 1 and 2, annotations were revised (with alignment scoring) by aligning the array element tentative consensus (TC) sequence against the NCBI nucleotide database using B L A S T n (http://www.ncbi.nlm.nih.gov/BLAST/) and default settings.  To obtain homologous Arabidopsis gene models, I used TAIR  BLASTn  (http://www.arabidopsis.org/Blast/) with default settings against the ' A G I Transcripts' (introns ,+UTRs) dataset. GO SLIMs summaries for clusters of AGIs were obtained from TAIR Bulk  Downloads  (http://www.arabidopsis.orR/tools/bulk/)  Categorization' tab.  50  through  the  'Functional  Filtering and clustering  A l l genes reported as significantly expressed in "Chapter 3 - Results" have been filtered in Excel on the basis of a minimum mean expression ratio while the associated minimum [fdradjusted] p-value is stated.  A l l hierarchical linkage clustering was performed in Genesis  (Institute for Biomedical Engineering, Graz University of Technology, add url) using the 'average linkage clustering' agglomeration rule, with the maximum colour saturation occurring at +/- 8-fold (log value=3) (Sturn et al., 2002). 2  3.2.7  Reverse transcription and quantitative P C R  Following DNAse 1 (Invitrogen, cat#l80680-015)  treatment according to the standard  Invitrogen protocol, 50 ul reverse transcription (RT) reactions were performed using a modified first-strand c D N A protocol from Invitrogen with 3 ug total R N A (volume calculated from GeneQuant 260/280) for each sample as the starting material. Samples were incubated in a nuclease-free microcentrifuge tube with 2.5 ul 01igo(dT)i2-i8 (500 ug/ml), 2.5 ul dNTP Mix (10 m M each), and the remaining volume adjusted to lOul with distilled water, for 5 minutes at 65°C followed by quick chill on ice. Each sample was then mixed with 12.5 ul 5X First-Strand Buffer and 5 pi 0.1 M DTT, and incubated for 2 minutes at 42°C. The sample was then incubated with 500 units (2.5 ul) of Superscript II RT (Invitrogen, cat# 18064-022) for 90 minutes at 42°C in a total reaction volume of 50 ul. Twenty-mer  primer  pairs  were  designed  with Primer3  (http://frodo.wi.mit.edu/cgi-  bin/primer3/primer3_www.cgi) to amplify gene-specific sequences 130-150bp in length (default settings selected).  Primer pair sequences for SIPK and the genes featured in Figure 5.1 are  summarized in Appendix Table 5. For quantitative PCR reactions, 2 ul cDNA (the cDNA equivalent of 200 ng total RNA) as incubated with 10 pi 2X QuantiTect S Y B R Green Mastermix (Qiagen, cat# 204143) and 5 u M 51  each of a forward and a reverse primer in a total volume of 20 pi. For each primer set (including tobacco actin (TOB54)), samples (n = 3 per treatment/genotype) were amplified by quantitative PCR using the M J Research D N A Engine Opticon 2 System programmed for an initial denaturation step at 95°C for 15 min, followed by 40 cycles of denaturation (95°C for 30 seconds), annealing (55°C for 30 seconds), and extension (68°C for 45 seconds). After a final extension phase, melting curve analysis was performed from 60 to 90°C, reading every 0.2°C for 5 minutes, to confirm the target-specificity of primer pairs. For each run, Opticon Monitor 2 Software (MJ Research) was used to manually adjust the threshold level was above background fluorescence. The resulting cycle thresholds C(T) were corrected in Excel by dividing the C(T) value for a sample by the mean of triplicate readings of actin C(T) values for that sample, thus generating normalized C(T) values. For comparison with microarray expression ratios, SEPK-Ri and WT samples for each treatment were randomly paired, and C(T) ratios generated analogous to SIPK-Ri / WT expression ratios. Eighty percent confidence intervals were calculated for both microarray mean expression ratios and qRT-PCR mean C(T) ratios as a measure of replicate variability.  52  3.3 Results  3.3.1  SIPK-silencing has differential effects on hrpZ and p-megaspermininduced defense responses in tobacco plants and cell suspension cultures Psph  As shown in the previous chapter, silencing of SIPK pre-disposes N.tabacum var. Xanthi to increased sensitivity to hrpZ h, indicating that SIPK likely plays a negative regulatory role in Psp  harpin-induced HR. Like harpins, megaspermins are known to induce H R in tobacco (cv Samsun NN) (Pellegrini et al., 1993), but little is known about the role of SIPK in mediating (3megaspermin-induced defense responses. To determine the level of sensitivity of N.tabacum var. Xanthi plants to megaspermin, lesion formation was monitored in WT leaf tissues infiltrated with 0, 9, 90 and 900 ng/ml Pmegaspermin. Lesions were induced by the two higher concentrations (Figure 3.1), but not at concentrations < 9 ng/ml (data not shown). When a similar experiment was conducted with SIPK-Ri lines, no obvious differences in lesion formation were apparent between WT and SIPKRi lines at either 24 or 48 hours. As in WT, lesions did not result from treatment of SIPK-Ri with P-megaspermin concentrations < 9 ng/ml (data not shown). For comparison, 500 pg/ml and 50 ug/ml hrpZ h infiltrations were performed opposite the Psp  P-megaspermin infiltrations on the same leaves (Figure 3.1).  The contrast between the  differential H R response to these harpin treatments observed in WT and SIPK-Ri tissues, compared to the WT and SIPK-Ri responses to p-megaspermin treatment indicates that SIPK does not play the same role in the latter cell death induction process.  53  WT  SIPK- RI  • 5 mM M E S Control A 500 ug/ml h r p Z • 50 ug/ml h r p Z S 9 Mg/ml p-megaspermin f 0.9 ug/ml p-megaspermin Psph  Psph  Figure 3.1 Lesion formation in W T and SIPK-Ri after infiltration with h r p Z megaspermin.  Psph  and p-  Icons are positioned over leaf injection sites. Dotted lines mark the outer boundary of the infiltration zone. Images shown are representative of genotypic responses (n = 3).  3.3.2  SIPK and WIPK transcription and activation in cell suspension culture  Transcriptome profiling can provide valuable insights into the extent of SIPK involvement in regulating stress-induced transcriptional changes, while also enabling the identification of putative SEPK transcriptional targets. The uniformity of undifferentiated cells in suspension  54  culture provides an experimental system that is less prone to the cell-to-cell transcriptional heterogeneity present in planta, a background that could potentially obscure observation of SIPK-dependent transcriptional responses.  Furthermore, the cell culture system permits the  synchronous treatment of large populations of cells, allowing more precise temporally-specific, elicitor-induced phenomena such as ion fluxes, kinase activation and early modification of gene transcription (Hahlbrock et al., 2003). Since SIPK appears to influence cell fate, I chose to compare SIPK-Ri and WT responses to elicitor concentrations that have been demonstrated to induce significant cell death in tobacco cultured cells. Treatment of tobacco B Y 2 cells with 20-100 ug/ml harpinp has previously been sp  shown to induce up to 30% cell death (from <5%) over a period of 24 hours (Ichinose et al., 2001) while N.tabacum var. Xanthi cells responded to 4.8 ug/ml harpin (E.amylovora) with 30% and 60%) cell death over 24 and 50 hours, respectively (Xie and Chen, 2000). Likewise, 50nM (~9 ug/ml) P-megaspermin has been shown to elicit greater than 4-fold increase in cell death of B Y 2 cells (Dorey et al., 1999). Thus, I selected 40 ug/ml hrpZp h and 9 ug/ml p-megaspermin sp  as treatment concentrations to be assessed at two time points (4 and 8 hours) that are likely to capture distinct stages of cell fate determination. Prior to investigation of harpin and megaspermin transcriptional responses, I wanted to first confirm that the reduction in SIPK transcription observed in planta for the SIPK-Ri line (R5) (Samuel and Ellis, 2002) was also maintained in the SIPK-Ri suspension cultures derived from this line. Comparisons of SIPK transcript levels were made between SIPK-Ri and WT for control and elicitor treatments of cell culture. Quantitative real-time polymerase chain reaction (qRT-PCR) using SIPK-specific primers and R N A obtained from two independently-derived WT/SIPK-Ri sets of cell suspension cultures (Figure 3.2).  These comparisons have been  summarized as [log2] mean ratios of SIPK-Ri/WT signals (3 biological replicates per comparison). The clear bars in the graph display values derived from the total RNAs used in the 55  microarray experiments and cell culture work described in Chapter 2, while the hatched bars represent values obtained from a different set of cell cultures in which the SIPK-Ri suspension culture was established from explants of new plants derived from different seeds of the same original [Tl] line.  In both cell culture sets, control and treatment [log2] mean ratios were  negative, demonstrating that suppression of SIPK transcript accumulation is maintained in the SIPK-Ri line relative to WT, in both control and elicitor-treated cells.  -4 -4.5  C  H4  M4  Figure 3.2 Relative levels of SIPK transcripts in WT and SIPK-Ri cell cultures following control and elicitor treatments. Expression ratios (log (SIPK-Ri/WT)) from gene-specific qRT-PCR of total R N A for control, 4hours 40 ug/ml harpin treatment (H4) and 4-hours 9 ug/ml P-megaspermin treatment (M4) (clear bars). Hatched bars represent values similarly-obtained from treatments of independentlyderived WT and SIPK-Ri cell suspension cultures. Error bars represent 80% confidence intervals (n = 3). 2  To demonstrate that the responses elicited by hrpZp h and P-megaspermin include sp  M A P K activation, proteins were extracted from cell suspension cultures of both WT and SIPKRi that had been either untreated (5 m M MES) or treated for 30 minutes with either 40 ug/ml hrpZpsph or 9 ug/ml P-megaspermin (Figure 3.3). Treatment of WT cell cultures with either elicitor resulted in the activation of two M A P K s , previously identified as SIPK and WIPK. In the SIPK-Ri line, WIPK activation was observed in response to either elicitor treatment. 56  Surprisingly, despite the presence of the SIPK-Ri construct in the R i line some SEPK activation, observed as a 48kDa band, was still detectable in response to harpin elicitation, although it was markedly reduced relative to that seen in WT cultures. This result could not be attributed to unequal loading of the lanes (which was not apparent from Coomassie staining). The same was true for SEPK-Ri leaves that had been treated with 0.5 ug/ml (Figure 2.2). For both elicitortreated SEPK-Ri samples, the WEPK activation band was proportionally more intense than the SIPK activation band, in contrast to the pattern seen in the WT culture response. In fact, WIPK activation appears to be higher in the SIPK-Ri line than in WT for both elicitors.  SIPK-Ri  WT  C  H  M  C  H  M  C= 5mM M E S control  - SIPK  30 min  ~ WIPK  Figure 3.3 SIPK and WIPK activation in wild-type (WT) and SIPK-silenced (SIPK-Ri) tobacco cell suspension cultures 30 minutes after treatment with 40 ug/ml hrpZ h (H) and 9 ug/ml p-megaspermin (M), immunoblotted with anti-pERK antibody. Coomassie blue staining of corresponding membrane portion below indicates loading (40ug/lane). Psp  3.3.3  Confirmation of defense response elicitation in cell suspension culture  Before transcriptional analysis of cell culture responses to harpin and megaspermin treatments, it was necessary to establish that both WT and SEPK-Ri cell cultures were responding to hrpZ h and P-megaspermin treatments with defense responses. Psp  One of the  earliest physiological responses to pathogen challenge is a rapid and often sustained burst of ROS production in the apoplastic space, where it has been proposed to contribute to cell wall strengthening through oxidative cross-linking of cell wall proteins (Levine, 94; Jabs, 99; Lam, 57  04).  In cell cultures, this phenomenon is typically accompanied by an alkalinization of the  culture medium, which occurs within minutes to hours after treatment of tomato cells with P.syringae preparations (Felix, 1999), of tobacco cells with harpin preparations (Baker et al., 1993), and of poplar cells with a flagellin-derived peptide (flg22) or chitosan (Haruta, 2003). To confirm that hrpZp h and P-megaspermin do elicit typical defense responses in both WT and sp  SIPK-Ri N. tabacum var. Xanthi cell suspension cultures, the pH of WT and SIPK-Ri cell culture medium was measured before treatment and again 4 hours after treatment with either 40 ug/ml hrpZpsph  or 9 ug/ml p-megaspermin (Figure 3.4). For each genotype, significant increases in pH  occurred for both elicitor treatments (n = 3), although the pH shifts detected in the WT and SIPK-Ri cultures were not significantly different from each other in the case of either elicitor. The elicitor treatments also resulted in a higher proportion of dead cells in response to both elicitors 24-hours post-treatment, relative to untreated cultures but this result is inconclusive due to the low numbers of cell death observations (see Appendix, Figure 4.2).  58  6.50  6.00  5.50  PH 5.00  4.50  4.00  H4 Treatment  Figure 3.4 Cell culture responses to h r p Z p  sph  and P-megaspermin treatments.  Medium alkalinization of WT (blue) and SIPK-Ri (grey) tobacco cell suspension culture in response to 40 ug/ml hrpZp h and 9 ug/ml p-megaspermin. Values are provided for a 5 m M MES control ' C , hrpZp h 4hrs (H4) and P-megaspermin 4hrs (M4). Error bars represent 95% confidence intervals (n = 3). sp  sp  3.3.4  The effects of silencing S I P K on gene transcription i n elicited and nonelicited tobacco cell suspension cultures  Transcriptional reprogramming controlled by up-stream stress signals such as activated SIPK is likely to play a critical role in defining defense outcomes. Preliminary to the assessment of the impact of silencing SIPK specifically on the elicitor-induced transcriptional changes observed between the control and the treatments, it was necessary to determine the effects of silencing SIPK on the transcriptome prior to stress. For this purpose, cDNAs representing R N A from 5mM MES-treated SIPK-Ri tissues were co-hybridized with cDNAs representing R N A from similarly treated WT tissues onto potato cDNA (10K) microarrays. This allowed me to 59  directly determine the effects of silencing SIPK on the transcriptome prior to application of any stress.  For comparison of WT and SIPK-Ri in the elicitor-induced states, a similar co-  hybridization was carried out using tissue from WT- and SIPK-Ri-treated cultures at 4 and 8hours after hrpZpsph (H4 and H8, respectively) and P-megaspermin (M4 and M8, respectively) treatments.  For each treatment (and control), 3 biological replicates from each of WT and  SIPK-Ri genotypes were randomly paired to establish 3 co-hybridizations. The relative transcript abundance between WT and SIPK-Ri derived from the microarray co-hybridizations was subsequently validated by means of qRT-PCR performed on these same R N A preparations for 12 annotated tobacco genes (see Appendix Figure 4.1 for details). The A N O V A statistical model provides a means of isolating genes with the most dramatic variation in SIPK-Ri/WT expression ratios amongst treatments. Using this model, 1000 genes were retrieved that exhibited the most significantly altered expression ratios amongst two or more treatments (control, H4, H8, M4, or M8) (p-value<5.35E-08). Using the program Genesis (see methods), these genes were hierarchically arranged on the basis of similarity in gene expression profile across treatments (complete linkage clustering), and represented as a hierarchical cluster image (heat map) (Figure 3.5). These genes are listed in the order in which they appear in Figure 3.4, along with their expression data, in Supplemental list 1. In order to identify groups of genes that are co-regulated by SIPK in the context of elicitor treatments, six groups of SIPK-Ri/WT expression profiles were identified. Cluster I contains 85 genes with higher expression in the SIPK-Ri line compared to WT (positive expression ratios) after harpin treatment but lower expression in SIPK-Ri than WT (negative expression ratios) in the megaspermin treatments. By contrast, Cluster II (62 genes) and Cluster V (99 genes) were generally negative ratios in the harpin treatments but positive ratios in the megaspermin treatments. Clusters IV (53 genes) and Cluster V I (112 genes) appear to have predominantly negative ratios in either of the megaspermin and harpin treatments, respectively. A small subset 60  of genes highlighted as Cluster III (75 genes) appear to have positive ratios at the 8-hour but not at the 4-hour time points for both stresses. Gene function for these clusters was not investigated, although gene annotation information, provided by the TIGR Solanum tuberosum gene index (StGI), is included in Supplemental list 1.  H4 H8 M4 M8  C  Figure 3.5 Hierarchical average cluster image of 1000 transcripts with significant differences in SIPK-Ri/WT ratios between treatments. Genes are assembled by rows according to similarity of expression profile across all treatments (excluding control). Purple and green represent up- and down- regulation of a gene between SIPK-Ri and WT samples for 4 and 8-hours hrpZp treatments (H4 and H8, respectively), and 4 and 8-hours P-megaspermin treatments (M4 and M8, respectively). Associated control (C) values are represented on the right. A N O V A p-value = 5E-03. See Supplemental list 1 for complete list of these genes and their expression values. sph  61  Of these 1000 genes, 477 have significant positive or negative expression ratios in the control (5mM MES, or "prior to treatment"), indicating that SIPK-silencing affects transcription of this subset of genes independently of stress application. While H8 exhibits a similar number of displaced expressions (420), values for H4 (120), M 4 (115), and M8 (55) are considerably lower, reflecting fewer genes displaced by the silencing of SEPK in these treatments (p-value<lE-03). The majority of the 22 genes that have >5-fold up-regulation in the SEPK-Ri line prior to treatment have been associated with defense, including genes annotated as polyphenol/catechol oxidases or their precursors (8 genes), tospovirus-resistance protein A , P R l b , PR1 precursor, Sadenosyl-L-methionine:salicylic acid carboxyl methyltransferase, and UDP-glucose:salicylic acid glucosyltransferase.  Within this subset of stress genes, the average SIPK-Ri/WT fold  difference declined from 5.81 in the control to 0.55 (st. dev. = 0.21), averaged across all elicitor treatment classes. The predominant group of genes amongst the 17 that were down-regulated >5-fold in the SEPK-Ri line prior to stress consists of those associated with cell wall degradation, namely 3 occurrences of pectate lyase, and 2 occurrences of xyloglucan endo-transglycosylase. Several antioxidant enzymes are also down-regulated in the SIPK-Ri line, including garnmahydroxybutyrate dehydrogenase, a putative N A D H dehydrogenase, and a proline-rich protein (GPP1). Similar to the up-regulated gene set, the average SEPK-Ri/WT fold difference for this set was attenuated from -8.64 to -3.37 (st. dev = 1.6), averaged across all elicitor treatment classes. To quantify the heterogeneity that is apparent between treatment classes in the cluster image (Figure 3.5), pair-wise Pearson correlations were performed between all classes (including control) based upon the 1000 ANOVA-filtered differential expressions for each class (Figure 3.6). Pearson correlation coefficients of determination (r values) provide more detail than do 2  dendogram-based clustering of experiments, summarizing the extent of profile similarity 62  amongst treatments rather than reporting this similarity as a branch length or proximity. In this case, the analysis indicates that treatment classes vary considerably from the control, ranging from most similar, H4:Control (r = 0.64), to least similar, M8:Control (r = 0.20). For each 2  2  elicitor, the 4-hour treatment classes are more similar to the control than the 8-hour treatment classes. The greatest similarities occur between the megaspermin (M4:M8 r = 0.66) and harpin 2  treatment classes (H4:H8 r = 0.65). 2  To assess whether or not these correlations represent  significant similarities in transcriptional response, the variation within each class was assessed by performing correlation analysis between the biological replicates in each treatment for the same set of genes.  As expected, expression were most consistent within the control  hybridizations (r = 0.97) indicating that cell cultures separated for 8 hours maintained a very similar gene expression profile. Since no intra-class correlation values were less than 0.90 for these ANOVA-filtered values, the vast majority of inter-class variation (maximum r = 0.66) is 2  likely to reflect expression differences that exist between treatments.  Control  0 97  H4  0.64  091  H8  0.61  0.65  0 96  M4  0.46  0.57  0.39  0.90  M8  0.20  0.37  0.40  0.66  0 90  Control  H4  H8  M4  M8  Figure 3.6 Transcription profile comparison between SIPK-Ri/WT expression ratios for control and treatments. Pearson's correlation coefficients (r values) are based upon 1000 ANOVA-filtered differentials per class ( A N O V A p-value = 5E-03). Inter-class coefficients represent comparison of 1000 mean ratios between two treatment classes, while intra-class values (shaded) represent means of all possible pair-wise correlations between biological replicates. 2  63  3.3.5  SIPK-silencing affects distinct sets of genes in hrpZ induced responses  Psph  - and p-megaspermin-  The role of SIPK can be more directly inferred, on a gene-for-gene basis, from the impact that SIPK-silencing has on the elicitor-induced transcriptional responses (responsive to SIPK, or 'RS'), as determined by comparison of SIPK-Ri/WT treatment group ratios to the SIPK-Ri/WT control ratio (the reference).  This subtractive method assumes that gene  transcripts that are not related to SIPK will respond similarly to treatments in both genotypes, thereby not affecting the relative expression ratio between control and treatment resulting in a non-significant expression change. Thus, comparison of the control ratio to that of the harpin 4hour ratio (H4), or harpin 8-hour ratio (H8) yields the SIPK-affected responses (H4 s and H8 s, R  R  respectively), measured as a [log2] fold-change with associated significance value (p-value of ttest).  Similarly, the SIPK-affected changes in the P-megaspermin-induced responses can be  summarized as M 4 s (4-hours) and M8RS (8-hours). Gene ontology summaries, utilizing TIGR R  StGI gene annotations of nearest homologs to potato ESTs, were generated for each response, yet do not fully describe these gene sets due to the high incidence of missing or incomplete annotations (Appendix, Figure 4.3). To gauge the similarity of these responses (H4 s, H8 s, M 4 , and M8 s), short-lists R  R  R S  R  were developed for each response, composed of genes whose expressions were most significantly altered by SIPK-silencing (>2-fold, 1.87E-03). These short lists were then compared in a series of pair-wise correlations on the basis of ' R S ' values, considering genes that were significantly affected by SIPK-silencing in either treatment (Figure 3.7A). The correlation coefficient between harpin and megaspermin responses was reduced from r = 0.77 for the 4hour responses (H4 s and M4 s) to r = 0.69 for the 8-hour responses (H8 s and M8 s), 2  R  R  R  R  suggesting a divergence in SIPK-affected transcriptional response over time between these stresses. The correlation coefficient between H 4  and H8 s was similarly low (r = 0.70), while 2  R S  64  R  the M 4  and M8 s coefficient was high (r = 0.90), implying that much of the dissimilarity 2  R S  R  between H8RS and M8RS had arisen by a shift in the transcriptional profile of the harpin response over the course of 8 hours. Consistent with these observations, M4 s and H8RS are the most R  divergent (r = 0.52), suggesting that transcriptional response was affected by both the type of 2  treatment and the time that had elapsed since treatment initiation.  A  B  -0.20 t  \ 0.52  0.69  0.80  t  t  0.77  0.86  1  0.90  Figure 3.7 Correlation between sets of genes that are most significantly affected by (A) SIPK-silencing in the elicitor responses and (B) most significantly induced by elicitor treatment in WT. For ' A ' , values in boxes represent the total number of genes whose elicitor-induced transcriptional response profiles, at 4 and 8-hours for either hrpZp h or P-megaspermin, were responsive to SIPK-silencing ('RS') (>2-fold, p-value<1.87E-03). For ' B ' , values in boxes represent the total number of genes that were up- or down-regulated (p-value<5E-05) 4-hours after hrpZp and P-megaspermin treatments (H4 T and M 4 T , respectively). Values within arrows represent the Pearson's correlation coefficients between treatment ratios for the genes present in either of the indicated lists. sp  spn  W  W  While it is unclear exactly how transcriptional responses that are dependent upon SIPKsilencing reflect endogenous roles for SIPK, SIPK-dependent transcriptional changes are most likely to be a sub-set of the transcriptional changes that occur in WT plants as a result of stress. 65  To determine the WT responses to either agent, c D N A / R N A from untreated W T tissue was cohybridized on potato c D N A (10K) arrays with WT samples treated with either hrpZp h- or (3sp  megaspermin (Figure 3.7B). experiments in this thesis.  For consistency, the same tissue was for all microarray  For each elicitor response, a short-list was generated containing  genes with significant transcriptional responses (>2-fold, p-value = 5E-05) (H4WT and M4 T, W  respectively). Correlation of gene expression differentials between these two lists yielded a coefficient of r = 0.86 which was 0.09 and 0.28 higher than either of the H4 :M4RS or 2  rs  H8RS:M8RS comparisons (respectively), indicating that SIPK-silencing may be modifying these elicitor-induced responses in a stress-specific manner that increases over an 8-hour period. Direct comparison of H4WT to H4 s allows validation of SIPK-affected transcriptional R  modifications in the context of the elicitor-induced WT background. If SIPK plays a role in regulating gene expression during stress response, then silencing of SIPK should have an effect on the stress-induced transcriptional modification of a given gene that is equal and opposite to the effect that SIPK has on this gene's stress-induced transcriptional response in the WT background.  A comparison of H4WT to H4 s encompassing 691 genes yields a negative R  correlation (r = -0.20), consistent with the net transcriptional effects that silencing of SIPK can 2  be anticipated to have on SIPK-dependent genes. Only twenty (non-redundant) genes were in common with H4 T and H4 s, representing approximately 4% and 10% of those lists, W  R  respectively where r = -0.43 (Table 1). Thirteen out of 20 of these transcriptional responses clearly differ between H4WT and H4 s in a manner that suggests that SIPK plays a role in their R  harpin-induced response, since silencing of SIPK impacts transcription of those genes significantly in an opposite orientation, observed as a difference in sign ('+' versus '-') between the H4RS and H4WT values. SIPK-silencing has a positive impact on the transcription of three genes that were downregulated in WT in response to harpin. These are xyloglucan endo-transglycosylase (cell wall 66  loosening), a putative pectate lyase (cell wall degradation), and fibrillarin (possible nucleolar targeting signal). SIPK-silencing also down-regulated the expression of 10 stress-associated transcripts that were up-regulated in the WT harpin response.  These included defense-  associated genes annotated as shaggy-related protein kinase (wound-response), putative kinesin light chain gene (cell death), and an elicitor-induced protein (TC67573). Other genes downregulated by SIPK-silencing include the redox-associated genes,  monodehydroascorbate  reductase and peroxidase, as well as hormone-signalling-related inositol polyphosphate 5phosphatase II. The remaining genes in Table 1 are associated with disparate primary metabolic functions. However, seven of those genes have fold-changes that are similarly-modified in both H 4 H4 T W  R S  and  (i.e. have the same sign), suggesting that SEPK-silencing may be enhancing the net  response of these genes to stress.  67  Table 1. 20 genes whose expressions are significantly induced 4 hrs after hrpZ treatment of WT cell suspension cultures (H4WT) (p-value<5E-05) and also significantly affected by SIPK-silencing in the hrpZ-induced response (H4 ) (>2-fold,pvalue< 1.87E-03 _ RS  Tentative Annotation  H4wr  H4  RS  Array Clone Function  TC73295  a  Accession  Species ^  trehalose-6-phosphate phosphatase homolog  Arabidopsis thaliana  At4g39770.1  C  d E-value  8.0E-15  fold-change  f p-value  e  g fold-change  3.0Er06  2.43  p-value  5.14  1.6E-05 4.2E-04  TC58326  xyloglucan endotransglucosyiase-hydroiase  Lycopersicon escuientumAY497477  O.OE+00  -1.71  3.8E-05  3.18  TC73791  putative pectate-lyase {Capsicum annuum}  Capsicum annuum  AF492632  O.OE+00  -6.73  3.2E-05  2J8  1.5E-03  2.71  5.6E-05  TC66452  fibrillarin 2  Arabidopsis ihaliana  AY142647  1.0E-110  -4.66  6.0E-06  TC58816  ATP citrate lyase b-subunit  Lupinus albus  AJ344108  O.OE+00  1.30  2.3E-05  2.59  5.5E-05  TC58058  lipoxygenase (POTLX-1)  Solanum tuberosum  U60200  0.0E+00  2.16  1,6E4)5  2.21  3.6E-05  TC61962  mini-chromosome maintenance 7  Pisum sativum  AY371199  5.0E-79  1.61  5.4E-05  2.07  6.8E-06  TC63755  FS-38 kinesin-like protein  Mirabilis jaiapa  AY589697  2.0E-12  7.77  2.5E-0.6  -2.02  3.0E-04  TC68399  clone 134017R  Lycopersicon escuientumBT014563  1.0E-121  1.32  3.8E-05  -2.03  9.9E-04  TC66092  Shaggy-related protein kinase NtK-1  (4icotiana tabacum  X77763  O.OE+00  1.94  2.0E-05  -2.05  4.4E-04  TC57715  Sucrose synthase  Solanum tuberosum  AY205084  0.0E+00  3.55  1.2E-05  -2.07  7.1E-04  TC67260  serine carboxypeptidase precursor  Gossypium hirsutum  AY072822  2.0E-31  -1.52  5.0E-05  -2.16  2.2E-04  TC57778  endo-1,4-beta-glucanase  Lycopersicon escuientumAY205084  0.0E+00  -1.70  2.3E-05  -2.20  4.8E-04  TC68658  T-complex protein 11 contains Pfam Pf 05  ArabidopsisIhaliana  3.0E-10  4.13  5.5E-06  -2.25  8.9E-04  TC66537  monodehydroascorbate reductase  Lycopersicon escuientum141345  0.0E+00  3.52  7.7E-06  -2;42  9.7E-04  TC65924  pre-pro-cysteine proteinase  Lycopersicon escuientumZ14028  O.OE+00  •2.40  1.3E-05  -2.82  1.1E-03  TC66671  peroxidase (PER9-6)  Nicotiana tabacum  •1.0E-.18  7.80  5.6E-07  -2.91  4.7E-04  TC58418  SNF1 kinase complex anchoring protein  Lycopersicon escuientumAY245177  1.0E-158  2.74  3.5E-05  -2.98  4.8E-05  -328  3.0E-04 8.2E-04  7.6E-07  -4.78  7.1E-05  NM_116984  AY032674  TC59335  inositol polyphosphate 5-phosphatase II  Arabidopsis- thaliana  NM_179071  4.0E-07  2.48  TC67573  wound-inducible gene wun1  Ipomoea batatas  X17554  3;0E-37  4.60  P u t a t i v e f u n c t i o n s for a r r a y c l o n e s ( T C s ) w e r e d e r i v e d f r o m c l o s e s t g e n e h o m o l o g b y N C B I B L A S T S of T C s e q u e n c e s a n d listed w i t h a s s o c i a t e d 8  s p e c i e s n a m e s , g e n e a c c e s s i o n n u m b e r s , a n d E - v a l u e s c o r e s " . F o l d - c h a n g e arid p - v a l u e s c o r e s ' a r e r e p o r t e d f o r H 4 w r a n d H 4 b  0  6  e l e m e n t ( s e e s u p p l e m e n t a l list 2 for c o m p l e t e s e t of v a l u e s for all other c o - h y b r i d i z a t i o n s ) . G e n e s a r e r a n k e d by M 4  R S  R S  for e a c h a r r a y  fold-change . Four genes 9  h a v e b e e n r e m o v e d b e c a u s e t h e y w e r e m u l t i p l e i n c i d e n c e s of the s a m e g e n e , often with v e r y s i m i l a r e x p r e s s i o n v a l u e s  Direct comparison of M 4 W T to M4 s also yields a negative correlation (-0.39), with 30 genes R  significantly modified by P-megaspermin treatment in the WT background (5.6% of M 4 T ) and W  also responding in a SEPK-affected manner (12.7% of M 4 ) (Table 2). SIPK-silencing leads to RS  sign flips (+/-) in 28 of 30 transcriptional responses induced by P-megaspermin treatment of tobacco where r = -0.68. SIPK-silencing results in up-regulation of transcripts related to stress 2  response, including guanylate kinase (pathogen recognition), heat shock cognate 70kDa protein 1 (stress response), and lecithine cholesterol acyltransferase-like protein (cholesterol ester synthesis), while down-regulating stress-related transcripts associated with nam (no apical meristem)-like proteinlO and AVR9/CF-9-rapidly elicited protein 189.  SEPK-silencing also  appears to down-regulate several transcripts related to redox control in the context of megaspermin-induced stress, including thioredoxin (H-type 1), catalase isozyme, and peroxidase (prxl4) precursor.  SIPK appears to down-regulate several genes associated with the  transcriptional machinery (eukaryotic translation initiation factor 3, histone deacetylase, while up-regulating a stress-related transcription factor ( W R K Y 22) and adenosylhomocysteinase (osmotic balance).  Detailed analyses of the 8-hour time points were not performed due to  insufficient resources for hybridizations of treated with untreated WT tissues.  69  Table 2.  3 0 g e n e s w h o s e e x p r e s s i o n s a r e significantly i n d u c e d 4 hrs after b e t a - m e g a s p e r m i n treatment of W T cell s u s p e n s i o n cultures (M4WT) (P-  v a l u e < 5 E - 0 5 ) a n d a l s o significantly affected by S I P K - s i l e n c i n g in the b e t a - m e g a s p e r m i n - i n d u c e d r e s p o n s e ( M 4 ) (>2-fold,p-value< 1 . 8 7 E - 0 3 ) . R S  Tentative Annotation  Array Clone Function guanylate kinase (GK-1) TC58502 histone deacetylase-like protein TC57721 TC59348 TC69618 TC66818 TC69478 TC67262 TC57689 TC68231 TC66452  o  TC65910 TC57911 TC57595TC57805 TC67604 TC67327 TC67462 TC65912 TC59725 TC58050 TC66780 TC73120 TC72942 TC68658 TC68546  a  .  Species  b  eukaryotic initiation factor 3E subunit Avr9/Cf-9 rapidly elicited protein 189 glucose-6-phosphate/phosphate translocator, putative,  1.0E-58 0.0E+00 0.0E+00  -2.70 -2.84  Nicotiana tabacum  AF493159 X54029 Y11996  Arabidopsis thaiiana Avicennia marina Arabidopsis thaiiana Arabidopsis thaiiana  AT4G31120.1 AF056316 AY042858 AY142647  2.0E-19 1.0E-33  Lycopersicon esculentum Arabidopsis thaiiana Nicotiana tabacum Arabidopsis thaiiana  AY456411 AT1G50920.1 D16138 AF285832 AY220479 At5g17630.1 AJ225172  Nicotiana tabacum Arabidopsis thaiiana Solarium tuberosum Solanum tuberosum  magnesium dependent soluble inorganic pyrophosphatase Sucrose synthase (Sucrose-UDP glucosyltransferase) 4,5-DOPA dioxygenase extradiol  REV3C mRNA for histone H1Mk T-complex protein 11 contains Pfam PF05 thioredoxin  TC60158  octicosapeptide/Phox/Bem1 p (PB1) domain-containing protein  TC57712 TC67271  catalase (CAT2) short-chain dehydrogenase/reductase  TC70318 TC67023  nam-like protein 10 peroxidase prx14 precursor  2.0E-35 1.0E-110 0.0E+00 1.0E-47 0.0E+00 5.0E-89 1.0E-111 5.0E-24  -2.34 -2.45  -3.29 -3.40 -3.42 -2.81 -4.09 -4.86 -3.09 -2.98 -2.60 10.57  f p-value 4.8E-05 2.45E-05 1.9E-05 3.4E-06 3.9E-05 6.7E-06 1.1E-05 6.35E-06 1.9E-07 3.9E-05 1.7E-07 5.5E-06 8.78E-07 3.2E-05 4.8E-05 1.8E-05  M18745 AJ583017  0.0E+00 0.0E+00 7.0E-07  -3.19 2.70 -1.77 2.13  AT5G22640.1 AF130253  4.7E-02 0.0E+00  1.87 11.55  Capsella rubella  AY166720  6.0E-44  Arabidopsis thaiiana Arabidopsis thaiiana Capsicum annuum  X62461 NMJ16984. AY496104  2.0E-05 3.0E-10 10E-105  4.85 2.50 5.33  NM_112707  2.0E-04  6.27  3.9E-07  AY500290 At5g18210.1 AF509873 AF244923  0.0E+00 4.1E-02  3.24 1.97  1.0E-128 0.0E+00  7.17 3.53  1.01E-06 3.6E-06 4.5E-07 4.0E-05  Beta vulgaris Arabidopsis thaiiana Arabidopsis thaiiana  MORN repeat-containing protein, membrane related protein CP5 WRKY transcription factor 22  e  fold-change  0.0E+00 3.0E-03 0.OE+00  Capsicum annuum  fibritlarin 2 ribosomal protein L3 (RPL3) GTP-binding protein-related S-adenosyl-l-homocysteine hydrolase  d E-value  C  AF205130 AT3G27730.1 AF354454  Medicago truncatula Lycopersicon esculenlum  PRT1 (elF3 subunit) Skbl methyltransferase family protein 40S ribosomal protein S7 phosphoglycerate dehydrogenase  Accession  -1.49  Niootiana tabacum Arabidopsis thaiiana  putative acyi-CoA synthetase lecithine cholesterol acyltransferase-like protein hsc-1 mRNA for heat shock protein 70 kD  M4  M4WT  Arabidopsis thaiiana Solanum tuberosum Arabidopsis thaiiana Petunia x hybrida Spinacia oleracea  1.73  1.3E-05 7.8E-06 7.8E-07 2.0E-05 2.3E-07 3.2E-05 2.7E-05 3.6E-06 9.1E-07  R S  g fold-change 4.57 4.27 3.66 3.30 3.15 2.41 2.41 2.38  p-value 1.8E-06 1.0E-05 7.9E-05 1.6E-04 1.9E-04 1.3E-04 5.8E-05 9.9E-05 5.1E-04  2.35 2.23 2.20 2.18  7.9E-04 4.9E-04 2.6E-04  2.14 2.02 -2.03 -2.16  8.7E-04 1.7E-04 1.3E-03 1.2E-03  -2.20 -2.24 -2.26  8.4E-05 1.0E-03 4.3E-05 5.5E-04 5.1E-04  -2.26 -2.44 -2.58 -2.59 -2.61 -2.91 -3.01 -3.16 -4.10 -4.93 -10.67  1.0E-05 6.0E-04 5.5E-05 2.8E-05 4.9E-06 5.7E-05 6.0E-05 1.6E-05 2.8E-05  Putative functions for array clones (TCs) were derived from closest gene homolog by NCBI BLASTS of TC sequences and listed with associated species names , gene 8  accession numbers , and E-value scores". Fold-change and p-value scores' are reported for M4wr and M 4 0  8  RS  for each array element (see supplemental list 2 for complete set of  values for all other co-hybridizations). Genes are ranked by M4R fold-change . Twelve genes have been removed because they were multiple incidences of the same gene, often 9  S  with very similar expression values.  .  ^  3.3.6  Functional analysis of SIPK-dependent transcriptional modifications for hrpZ - and p-megaspermin responses Psph  In the previous section, SIPK-dependent responses were identified that appear most likely to represent proximal down-stream targets of SIPK-mediated regulation. potentially  excludes  subtle  yet  biologically  important  Such a compilation  SIPK-dependent transcriptional  modifications that are more exclusive to either stress response. To gain a better understanding of how SIPK may differentially regulate these elicitor-induced responses, I devised a filtering system to identify genes that respond significantly (>3-fold, p-value<2E-04, n = 6) to either hrpZp ph or P-megaspermin, but not to both. To focus on those genes that respond in a SIPKS  dependent manner only upon stress, genes were removed that already showed significant differential expression prior to elicitor treatment (>1.5-fold, p-value<8E-03, n = 6).  The  filtering resulted in a set of 239 genes that respond uniquely to application of either hrpZp h or sp  P-megaspermin. This gene set was ordered by hierarchical linkage clustering on the basis of 'RS' values for the 4- and 8-hour treatment responses to both elicitors, and formatted using Genesis (Figure 3.8). Several clusters emerge from the hierarchical analysis. These include sets of genes that are significantly down-regulated by SIPK-silencing in the hrpZ h-induced (cluster A) or PPsp  megaspermin-induced (cluster B) responses, and corresponding sets that are up-regulated by SIPK-silencing in the hrpZp  sph  (cluster C) or P-megaspermin (cluster D) responses.  A small  fraction (44) of the overall gene set is excluded from these continuous clusters, providing evidence of the effectiveness of the filter in isolating elicitor-specific genes. A l l 239 genes are arranged as in Figure 3.8, with cluster designations and associated ' R S ' values (and p-values), in Supplemental list 2.  71  Figure 3.8 Hierarchical linkage clustering of 239 SIPK-dependent genes that respond most exclusively to hrpZp or p-megaspermin treatment. sph  124 genes respond in a SIPK-dependent manner 4 and/or 8 hrs after hrpZp h treatment only (H), while 114 genes respond in a SIPK-dependent manner 4 and/or 8-hours after P-megaspermin treatment only (M) (>3-fold, p-value< ,n = 6). Genes with significant differentials prior to any treatment are excluded (> 1.5-fold, p-value<8E-03, n = 6). Columns represent the transcriptional effects of silencing SIPK at 4 and 8-hours for each elicitor treatment (purple and green represent up- and down-regulation of a gene by SIPK-silencing). Vertical bars denote clusters of genes that are exclusively modified by one particular elicitor at either 4 or 8-hours (H or M). Refer to Supplemental list 2 for details. sp  72  To determine i f genes that respond similarly to particular stresses also share similar function, clusters A to D were compared on the basis of the occurrence of gene ontology terms. Since there was insufficient GO information in the StGI database to adequately annotate the genes within these clusters, I used GO Slim terms of the related Arabidopsis gene loci (identified by A G I locus identifiers). These A G I numbers were then used to obtain GO SLIM term summaries for 'cellular component', 'molecular function', and 'biological process' for the putative Arabidopsis homologues.. The resulting G O S L I M terms, which provide generalized, non-overlapping G O information for each cluster, were then summarized as relative proportions and absolute occurrences (Table 3). A set of 3000 randomly chosen AGIs were submitted as a representative of the genomic proportions of these GO SLIMS. In  terms  of  'cellular  component',  a  disproportionately  low  number  of  "chloroplast/plastid" genes (5%) appear in Cluster A (down-regulated by SIPK-silencing in the hrpZp p -induced response) compared to a mean rate of appearance of this GO term across all S  h  clusters (12.5%) and the random set (11%).  On the other hand, SIPK-silencing results in up-  regulation of relatively high proportion "ribosome" genes (18%), as compared to 6.75% (cluster mean) and 3% (random set). A high proportion of genes (12%) with "nucleus" annotation were found in Cluster C (up-regulated by SIPK-silencing in the hrpZp h response) in contrast to the mean proportion of sp  this GO term in all response types (7.75%) and in the random set (8%).  Clusters appear most  uniform in the proportion of "biological process" GO SLIM occurrences, with the exception of a high occurrence of "electron transport" genes (7%) being down-regulated by SIPK-silencing in the P-megaspermin response (Cluster B) where the cluster mean = 3%, and the random set = 3%. By contrast, "molecular function" ontologies appeared very heterogeneous between clusters. Cluster A contains a disproportionately high fraction of "kinase activity" genes (16%, 73  cluster mean = 7.5%, random = 8%) while Cluster C possesses no "kinase activity" genes, suggesting that SIPK-silencing down-regulates, but does not up-regulate, this class of genes. Similarly, Cluster A also possesses a high proportion of "transferase activity" genes (24%, cluster mean = 12.5%, random = 11%) while Cluster C possesses a low number (6%); no such difference is evident between the P-megaspermin-induced response clusters (B and D). Cluster C appears to be high in both " D N A or R N A binding" (11%, cluster mean = 3.8%), random = 7%) and "transcription factor activity" (11%), cluster mean = 5.75%, random = 3%), indicating that SIPK-silencing selectively up-regulates these classes of genes in the hrpZ h-induced response. Psp  Genes annotated as having "structural molecular activity" are more abundant in Clusters A and C, indicating that this class of genes is preferentially up- or down-regulated by SIPK-silencing in the hrpZp h,, but not in the P-megaspermin response. sp  74  Table 3: Gene ontology summaries of hierarchical clusters of genes uniquely-affected by SIPK-silencing in either of hrpZ or beta-megaspermin treatment. See Supplemental List 2 for expression data on Clusters A-D PSDh  HCL C l u s t e r GOSIimTerm  b  Cellular Component cellular component unknown chloroplast/plastid cytosol ER mitochondria nucleus other cellular components other cytoplasmic components other intracellular components other membranes plasma membrane ribosome  A 0  B  /  %, total*  0  „ c Random  a  C  /  %,total*  0  /  %, total*  D 0  /  %, total*  %, total* 20 ,872  17 ,70 13 ,8 5 ,3 2 ,1 5 ,3 12 ,7 2 ,1 8 ,5 7 ,4 18 ,17 3 ,2 8 ,5  28 13 9 0 6 9 0 11 4 15 0 6  ,73 ,6 ,4 ,0 ,3  14,3 23 ,5 0 ,0 18 ,4  21 ,12 19 ,11 3 ,2 3 ,2 9 ,5 5 ,3 3 ,2 10 ,6 2 ,1 19,71 2 ,1 3 ,2  Molecular Function DNA or RNA binding hydrolase activity kinase activity molecular function unknown nucleic acid binding nucleotide binding other binding other enzyme activity other molecular functions protein binding structural molecule activity transcription factor activity transferase activity transporter activity  3 ,7 8 ,3 16 ,6 11 A 0 ,0 5 ,2 8 ,3 5 ,2 5 ,2 3 ,7 11 ,4 3 ,7 24 ,9 0 ,0  2 ,2 15 ,75 6 ,6 11 ,71 2 ,2 4,4 6 ,6 25 ,26 10 ,70 1 ,1 2 ,2 3 ,3 10 ,70 4,4  11 ,7 13 ,8 0 ,0 8 ,5 0 ,0 2 ,7 8 ,5 13 ,8 9 ,6 3 ,2 8 ,5 11 ,7 6 ,4 9 ,6  3 14 8 2 7 9 17 11  ,3 ,72 ,7 ,2 ,6 ,8 ,75 ,70  Biological Process biological process unknown cell organization and biogenesis developmental processes DNA or RNA metabolism electron transport or energy pathwa; other biological processes other cellular processes other metabolic processes other physiological processes protein metabolism response to abiotic or biotic stimulus response to stress signal transduction transcription transport  6 3 3 3 0 3 6 38 3 21 3 3 6 0 3  9 ,12 1 ,2 6 ,8 0 ,0 7 ,9 5 ,7 3 ,4 37 ,50 10 ,13 13 ,17 1 ,2 2 ,3 2 ,3 2 ,3 1 ,2  6 ,7 4 ,5 6 ,7 1 ,1 2 ,3 7 ,8 3 ,4 27 ,33 13 ,16 11 ,14 2 ,3 4 ,5 4 ,5 5 ,6 5 ,6  3 ,3 1 ,1 2 ,2 6 ,6 3 ,3 6 ,6 2 ,2 37 ,37 9 ,9 10 ,10 2 ,2 2 ,2 5 ,5 5 ,5 6 ,6  14 5 5 0 5 5 0  ,3 ,7 ,7  ,0  ,7 ,1 ,0  14.3  ,2 ,1 ,1 ,1  ,0 ,1 ,2 ,13 ,1 ,7 ,1 ,1 ,2 ,0 ,1  A ,0 ,5 ,2 ,7 ,0 ,3  2,2 1 ,7 2 ,2 6 ,5 10 ,9 6 ,5  11 ,428 1 ,25 1 ,29 9 ,363 8 ,324 9 ,376  4 ,772  8 ,377 26 ,7070 1 ,47 3 ,774  7 ,440  11 8 12 2 7 6 15 4 2 2 3 11 8  ,646 ,470 ,736 ,746 ,409 ,345 ,907 ,241 ,147 ,137 ,193 ,674 ,446  7 ,794 2 ,170 1 ,145 2 ,230 3 ,287 3 ,357 19 ,2746 19 ,2182 22 ,2443 7 ,766 2 ,187 2 ,177 1 ,165 3 ,338 7 ,821  GO Slim summaries are provided for cellular component, molecular function, and biological process. Values in bold represent the frequency of occurrence of gene ontology (GO) Slim terms for Arabidopsis homologs of array clones b  contained within each cluster (A-D) , while italicized values indicate the number of occurrences of that term within each a  cluster. Summary for random set of 3000 AtGs . c  75  While SIPK-silencing has a dramatic impact on the transcription of many tobacco genes and there appears to be a basis for SIPK-dependent functional differences between elicitor responses, the mechanism by which SIPK affects the transcription of these genes is entirely unknown.  Promoter regions display a high degree of conservation amongst sets of genes  responding to stress (Chen et al., 2002; Mahalingam et al., 2003), likely most homologous in the proportion and position of diverse sets of cis-acting elements.  Important clues about gene  regulatory mechanisms involving SIPK may be gained through the exploration of cis-elements of genes whose expressions are affected by SIPK in a similar manner. Since the potato EST database lacks this sequence information, I used Arabidopsis genes as proxy, limiting my promoter investigation to 5' regions of Arabidopsis homologs that were obtained from TAIR BLASTs of TC clone (EST) sequences, where Eigen values were less than 1E-04 indicating a high degree of similarity. With the intention of determining i f Clusters A - D (co-expressed genes) possess a greater proportion of over-represented cis-elements within a lOOObp region upstream (5') of the start codon, the TAIR tool "Motif Analysis" was used to estimate the probability of 6-mer sequences occurring within that cluster by chance alone (stated as a p-value). Given the difficulty of establishing a meaningful p-value cut-off, four random lists of 44 genes (matching the average number of non-redundant genes in each cluster) were similarly analyzed. The fourteen most over-expressed 6-mers from Clusters A - D (coloured) and the four random sets (uncoloured) have been collated, ranked by increasing p-value in Table 4. Over-representation seems most likely for eight motifs in Cluster D, six in Cluster C, and two in Cluster B, since four consecutive motifs from random sets 1 and 4 rank just below these. Motifs from the random lists occur as frequently (56%) as those from clusters for the remaining 96 motifs, although their frequency is higher towards the highest p-values. This analysis cannot rule out any cis-elements as false-positives. To determine i f the detected motifs represent known plant regulatory motifs, 76  the first 20 6-mers were submitted to the Cis-acting regulatory elements (CARE) database resulting in one putative identification; A G T A T G corresponds to an endosperm regulation element. To verify that Motif Analysis accurately represents the motif representation in these clusters, the same sequences to were submitted to Motif Sampler (Thijs et al., 2001), resulting in motif over-representation probability scores for Clusters A - D that were largely indistinguishable from the random sets (data not shown).  77  Table 4: Most over-represented 6-mer base sequences in 5' regions (lOOObp) of Arabidopsis gene models associated with co-expressed tobacco genes in Clusters A - D Four random sets of Arabidopsis gene models are provided for comparison. Colour coding matching clusters is provided to assist interpretation.  Cluster D D D D C  c  D D B B C  c c c  D D R4 R4 R1 R1 B B B B D D R4 R4 A A R1 R1 R1 R1 D D R3 R3 A A R4 R4 D D A R1 R1 A A B B R3 R3 R1 R1 R2 R2  a  6-mer AGCCCA TGGGCT AGAATG CATTCT AGTTAA TTAACT GGGCTT AAGCCC AAGAGT ACTCTT AGTATG CATACT CAAGAG CTCTTG CCTAAC GTTAGG CACAGA TCTGTG CAGCAA TTGCTG AACTCT AGAGTT TATCAA TTGATA ATCTGG CCAGAT CCTGTC GACAGG CAGAAA TTTCTG TTCTAC GTAGAA ATCCAC GTGGAT AGCGTA TACGCT ATATGC GCATAT CGTATG CATACG GCCGCG CGCGGC TTATCG CGATAA TCATGA CGAGAC GTCTCG AGTAAG CTTACT GAGGTA TACCTC TAATAA TTATTA CAAGCA TGCTTG AGTCAC GTGACT  Cluster  p-value 6.17E-07 6.17E-07 1 43E-05 1.43E-05 1 75E-05 1.76E-05 4.28E-05 4.30E-05 1.36E-04 1 36E-04 2.02E-04 2.02E-04 2.58E-04 2.58E-04 2.70E-04 2.70E-04 2.75E-04 2.75E-04 3.38E-04 3.38E-04 4.10E-04 4.11E-04 4.86E-04 4.88E-04 4.93E-04 4.94E-04 5.05E-04 5.05E-04 5.47E-04 5.47E-04 5.86E-04 5 886-0* 5.91E-04 5.92E-04 6.66E-04 6.66E-04 6.84E-04 6.86E-04 7.08E-04 7.09E-04 7.48E-04 7.56E-04 7.64E-04 7.65E-04 8.53E-04 8.61E-04 8.61E-04 9.41E-04 9.43E-04 9.61E-04 9.61E-04 1 04E-03 1 04E-03 1 08E-03 1.08E-03 1.15E-03 1.15E-03  p-value  continued... A A R4 R4 R1 R1 R1 R1 R2 R2 R4 R4 B B R2 R2 B B B B R3 R3 R2 R2 A A A A C C A R3 R3 R2 R2 R4 R4 R3 R3 R3 R3 R2 R2 R4 R4 R3 R3 C C R2 R2 C C C C  AAATCG CGATTT GTCGAA TTCGAC CATTTC CGAACA TGTTCG GAAATG CTGATC GATCAG GGTTAA TTAACC ACCGAG CTCGGT CATGAG CTCATG ACTCCT AGGAGT GACCGA TCGGTC GATAAG CTTATC ACAAGC GCTTGT CGAAAC GTTTCG CTCGTA TACGAG ACATAA TTATGT ATCGAA GCTGAA TTCAGC TGTTGT ACAACA CGTGCG CGCACG TGCAAT ATTGCA ACACTT AAGTGT AACGAG CTCGTT AATGGA TCCATT CTGAAC GTTCAG CCTACT AGTAGG TCGGTG CACCGA AGAACT AGTTCT CTAGCT AGCTAG  1.19E-03 1.19E-03 1.22E-03 1.22E-03 1.26E-03 1.26E-03 1.26E-03 1.27E-03 1.43E-03 1.43E-03 1.47E-03 1.47E-03 1.50E-03 1.50E-03 1.62E-03 1.62E-03 1.73E-03 1.73E-03 1.73E-03 1.74E-03 1.77E-03 1.78E-03 1.88E-03 1.89E-03 1.90E-03 1.90E-03 2.08E-03 2.08E-03 2.19E-03 2.19E-03 2.48E-03 2.50E-03 2.50E-03 2.55E-03 2.56E-03 2.B5E-03 2.G6E-03 2.68E-03 2.69E-03 2.77E-03 2.78E-03 2.79E-03 2.79E-03 2.82E-03 2.82E-03 2.88E-03 2.88E-03 2.93E-03 2.94E-03 3.03E-03 3.04E-03 3.70E-03 3.70E-03 4.12E-03 4.13E-03  C l u s t e r d e s i g n a t i o n s match those of Figure 3.8. S e t s of 44 r a n d o m l y - s e l e c t e d g e n e s are d e s i g n a t e d  asR1-R4 b  6-mer  P r o b a b i l i t y of finding the 6-mer s e q u e n c e by r a n d o m c h a n c e  78  3.4  Discussion  3.4.1 SIPK has a different role in hrpZ responses  Psph  - and p-megaspermin-induced defense  Microarray analyses and bioinformatics approaches have been combined in this study to test the hypothesis that the tobacco M A P kinase, SIPK, plays a role in conditioning distinct defense response outcomes towards elicitors from evolutionarily diverse pathogens, Pseudomonas syringae pv. phaseolicola and Phytphthora megasperma. Lee et al. (2001a) earlier demonstrated that SIPK-activation was induced in tobacco B Y 2 cell suspension cultures by treatment with 9 ug/ml P-megaspermin, while low levels (2-100 ug/mi) of hrpZp h have previously been shown sp  to induce a range of defense responses (Lee, 2001a) including activation of SIPK and WIPK in cell suspension culture (Figure 2.2). Since WT plants infiltrated with 1 ug/ul hrpZ h displayed Psp  lesions that closely resemble those seen on SIPK-Ri plants challenged with lower concentrations of hrpZp h, SIPK may influence the cellular thresholds that regulate commitment to HR. sp  I first established that SIPK-silencing leads to increased sensitivity of tobacco plants to 0.5 ug/ul hrpZp h, while having no apparent effect on the extent of cellular collapse or lesion sp  formation arising from either 0.9 or 9 ug/ml p-megaspermin treatment (figure 3.1). However, this dramatic difference does not preclude SIPK involvement in P-megaspermin-induced defense responses, nor does it indicate that SIPK is necessarily restricted to cell fate determination in harpin-induced defense responses. As an experimental system, I employed WT and SIPK-Ri cell suspension cultures challenged with either hrpZ  Psph  or P-megaspermin. While my primary interest in this investigation was to  compare the transcriptional responses of tobacco to hrpZp  spn  and B-megaspermin, I was also  interested in the elicitor-specific transcriptional changes that may be critical in defining cell fate. Heath et al. (2000b) reported that resistant cowpea epidermal vein cells treated with rust fungus (Uromyces vignae) displayed irreversible cell death several hours after infection, marked by the 79  retention of the nucleus at the infection site (plasma membrane) followed by cessation of cytoplasmic streaming and extended (many hours) dismantling of the protoplast. In the absence of published data defining the reversibility window for cell fate in tobacco tissues treated with harpins or megaspermins, I assayed the transcriptome at 4 and 8-hours post-treatment. It is worth noting that the microarray data revealed significantly lower expression in the SEPK-Ri line as compared to the WT line across all treatments for a potato array element (TC67874) which was a reciprocal best hit (BLASTn) to tobacco SEPK (Appendix, Figure 4.1). However, the observed 2-fold down-regulation (average amongst control and treatments) is less than would be expected i f the RNAi-facilitated SEPK transcript degradation were complete in the SEPK-Ri line. The fact that array element TC67259 (reciprocal best hit to NtF4, a closely related tobacco M A P K ) exhibited similar expression patterns to TC67874 (reciprocal best hit to SEPK) (Appendix, Figure 4.1) suggests that either of these two array elements may hybridize with both SEPK and NtF4 transcripts. For consensus regions, TC67874 is 90% and 85% similar to SEPK and NtF4, respectively, while TC67259 is 75% and 87% similar to SEPK and NtF4. In contrast, the consensus region for WEPK was only 17% and 15% similar to TC67874 and TC67259, respectively. While cross-hybridization of either SIPK or NtF4 transcripts to either TC67874 or TC67259 cannot be ruled out, and there appears to be persistent SEPK transcription (mRNA) in the SEPK-Ri line, SIPK transcription was continually suppressed in this line, regardless of treatment or time point (Figure 3.2 and Appendix Figure 4.1), thus providing a functionally (if "leaky") SIPK-suppressed line. This SEPK-suppressed line (SEPK-Ri) was used to compare elicitor-induced transcriptional responses against those observed in the WT line to infer the subset of these responses that are SEPK-dependent. Immunoblotting of protein extracts from WT cell suspension culture, using anti-pERK, established that a 48kDa protein, confirmed as SIPK in Chapter 2 (Figure 2.2), was strongly active 30 minutes after application of either hrpZp h or P-megaspermin (Figure 3.3), whereas no sp  80  phospho-SIPK was detected in untreated cells. As expected, suppression of SIPK in the SEPKRi line led to much reduced levels of SEPK activation after treatment with either elicitor. Some pSEPK signal was also persistent in similar analyses of leaf tissue from elicitor-treated SIPK-Ri plants (Figure 2.2A). It should be noted that, while the immunoprecipitate in-gel kinase assay confirmed the activation of SEPK (Figure 2.2B), I cannot rule out the possibility that another protein of similar size and activation profile is also being detected by anti-pERK in the western blot. Nevertheless, these findings demonstrate that SEPK is activated in cell suspension cultures as a result of both hrpZ h and p-megaspermin treatments. Psp  Rapid SEPK activation in cell cultures is accompanied by other indicators that the plant tissues are responding to hrpZp h or p-megaspermin treatment. sp  In particular, medium  alkalinization was observed in cell suspension cultures of both WT and SEPK-Ri lines at 4 hours after treatment with either elicitor. While medium alkalinization in plant cell cultures can occur as early as 5 minutes after elicitor treatment (Haruta and Constabel, 2003), sustained elevations of extracellular pH have also been recorded 30 minutes to 6 hours after treatment of tobacco cells with harpin (4.1pg/ml) with no significant decline over that period (Baker et al., 1993). I also observed that the culture medium remained alkaline up to 8 hours after elicitor treatments (data not shown). In contrast to the increased cellular collapse (figures 2.1, 3.1) and associated increase in ROS burst (figure 2.3) resulting from SEPK silencing in elicited tobacco cells, medium alkalinization reached similar levels in both SEPK-Ri and WT cultures, indicating that SEPK has no apparent effect on this particular response. This would imply that the mechanisms that are likely to drive this extracellular alkalinization (activation of K+/H+ antiporters, activation of H+/solute cotransporters, and inhibition of plasma membrane H+ ATPase) (Haruta and Constabel, 2003) are distinct from ROS-generating mechanisms such as activation of N A D P H oxidase. In fact, the H2O2 burst induced by harpin has a biphasic temporal pattern while medium alkalinization 81  increases steadily to a plateau during the same response period (Baker et al., 1993; Dorey et al., 1999). I was unable to quantify H2O2 levels in the SIPK-Ri and WT cultures due to technical problems associated with the chosen chemiluminescence assay (luminol).  Similarly, due to  difficulties in obtaining consistent staining of sufficient numbers of cells, I was unable to either accurately quantify cell death or measure differences in cell death outcome resulting from the silencing of SIPK, although treatments with either elicitor did appear to result in an increase in the proportion of dead cells found in both W T and SIPK-Ri cell suspension cultures as compared to untreated cells (Figure 4.2). Although the response of WT and SEPK-Ri leaves to P-megaspermin appeared similar, it is still possible that p-megaspermin-induced H R in cell suspension cultures is influenced by SEPK-silencing in light of the substantial differences between complex, autotrophic plant tissues and undifferentiated, heterotrophic cells grown in suspension cultures. Dorey et al.(1999) demonstrated that only a subset of the P.megaspermaderived elicitors that induced H R in tobacco leaves (P-megaspermin included) remained effective in inducing defense responses in cell culture, implying that the growth environment and/or cell state can have a dramatic impact on defense response.  3.4.2  SIPK-silencing affects the pattern of gene transcription prior to, and during, stress  Given the substantial impact that SEPK-silencing has on hrpZp h-induced H R (Chapter 2), sp  and the requirement for de novo gene transcription to enact HR, SEPK is likely to modify HR, at least in part, through transcriptional change. However, SEPK is also very likely to influence gene transcription events associated with non-HR outcomes induced by these elicitors. After processing the data from the transcript profiling experiments comparing WT and SIPK-Ri, at least one thousand genes satisfied a robust A N O V A p-value (5E-03) criterion of statistically significant fold difference amongst all treatments and control.  82  The robustness of these  treatment-dependent differences in log2(SJJ K-Ri/WT) ratios established that SIPK does play a J  significant role in affecting either hrpZp h or p-megasperm-induced transcriptional changes. sp  Hierarchical linkage clustering of this ANOVA-filtered gene set revealed that genes clusters based on the type of stress, rather than the treatment time - Cluster III (grouping by 4 vs. 8-hour responses) was the sole exception to this pattern.  Pair-wise correlations between gene  expression values for these treatments (and control) more accurately quantified the differences between elicitor types and treatment times, revealing that the greatest similarities in transcriptional response (based on r values of gene expressions values for treatment pairings) 2  2  2  occurred between treatments of the same elicitor, where r = 0.66 (M4:M8) and r = 0.65 (H4:H8) (Figure 3.6). These were substantially higher than correlations between the two elicitor responses for the same time point (either 4 or 8 hours). This evidence suggests that SIPK is more relevant to establishing gene expression profiles that are specific to either elicitor than to reprogramming gene expression to suit particular stages in defense response. Interestingly, the P-megaspermin treatments were the most dissimilar of all to the control (r =0.20) despite in planta evidence that marginalizes the role of SIPK in P-megaspermin-, but not hrpZpsph-, induced lesion formation (Figure 3.1).  If transcriptional responses to P-  megaspermin do not involve SIPK, SIPK-Ri and WT cells should not differ in their megaspermin-induced gene expression patterns; i.e. these two data sets should yield r values 2  close to those for replicate within-treatment comparisons (approaching 0.95).  Since the  observed r =0.20, SIPK does, indeed, appear to affect transcriptional responses induced by P2  megaspermin. Elicitors can induce a range of responses that don't necessarily lead causally to an H R outcome, and it is unlikely that H R responses demand a higher degree of transcriptional reprogramming than other responses (such as phtyoalexin production). The lack of SIPK influence on p-megaspermin-induced H R in planta does not preclude SIPK involvement in transcriptional reprogramming arising from treatment of cells (or plant tissue for that matter) 83  with P-megaspermin. It may be possible to compare the cell culture profiles to those of leaf tissue, although transcriptional differences between cell types of leaf tissue may obscure SIPKdependent transcriptional effects. From this data, it is apparent that SIPK is relevant to both hrpZpsph-  and P-megaspermin-induced transcriptional responses, and that the impact that SEPK  has on the tobacco transcriptome differs by elicitor type.  3.4.3  Effects of SIPK-silencing on cell cultures prior to elicitor treatment  The large number of significant transcriptional differences between SIPK-Ri and WT in the control samples amongst the A N O V A set suggests that SEPK is involved in cellular processes that affect transcription even in the absence of elicitor application. This was unexpected since M A P K s are classically viewed as molecular switches whose function relies upon them being converted to their phosphorylated (active) forms. Both the WT and SEPK-Ri lines display very low levels of basal SEPK/WEPK activations prior to elicitor application which means that the observed transcriptome perturbations seen in un-elicited cells could reflect a reduction in the (small) basal pool of activated SEPK rather than physical absence of the un-activated protein. The SEPK-Ri cultures also show other phenotypic differences in contrast to WT cells grown under the same conditions. SEPK-Ri cultures produce a significantly more alkaline medium even in the absence of elicitation, and they also show increased cell clumping and strong cell discoloration (browning). It is likely, therefore, that these constitutive phenotypic differences between SIPK-Ri and WT cell cultures will be reflected in transcriptional differences.  It is  noteworthy that among the strongly over-expressed transcripts detected in the un-elicited SEPKRi cultures are several annotated as polyphenol/catechol oxidase (PPO), an enzyme known to catalyze oxidation and polymerization of phenolics. Since SEPK is constitutively expressed in WT tobacco cells (Zhang et al., 1998), it is possible that specific basal levels of non-activated SEPK are required for the maintenance of cellular homeostasis.  Alternatively, constitutive expression of the R N A i "hairpin" construct used to 84  suppress SIPK expression may be inducing a stress condition, since the R N A silencing machinery is effectively a defense mechanism. However, there is currently no documentation of such a response, despite widespread usage of R N A i technology for functional knockout studies in plants.  In fact, several lines of stably-transformed SIPK-silenced plants exhibited severe  suppression of SIPK transcription with no apparent phenotypic consequences for plants grown under normal greenhouse conditions (Samuel and Ellis, 2002).  Thus, the SIPK-Ri cultures  exhibit a phenotype (potentially including gene expression levels) that may be unique to the cell culture system. Given the large number of genes within the ANOVA-fdtered set that were significantly expressed in the control prior to treatment (477), and the difficulty in interpreting the relation of SIPK-silencing to these expressions, the A N O V A analysis and class correlations are primarily useful to demonstrate that SIPK-silencing affects transcription of many genes differently amongst treatments and control.  More sophisticated methods of analysis are required to and P-  distinguish SIPK-dependent transcriptional events that are induced by hrpZ ph Ps  megaspermin.  3.4.4 A subset of genes affected by silencing SIPK are most likely SIPK-dependent during elicitor-induced responses Comparison of each treatment to the control, producing ' R S ' values for each gene, were an effective means of isolating the most substantial SEPK-affected transcriptional responses. Pairwise correlations between ' R S ' lists indicate that 8-hour responses are less similar (based on r  2  values between paired treatments) than 4-hour responses suggesting that SIPK-mediated elicitor responses are diverging as a function of time; potentially establishing distinct metabolic and physiological end-points for each elicitor response. correlation exists between M 4  R S  and H 8  Consistent with this, the most dissimilar  (r = 0.52). However, H 4 , M 4 2  R S  RS  R S  and M 8  R S  appear  very similar while H8 s appears as the outlier. It is possible that the p-megaspermin-induced R  85  transcriptional changes are being modified from the control state at a slower rate than those of the hrpZpsph with the result that the H 4  and M8RS profiles are more closely aligned (r = 0.80). 2  RS  Additional later time points could possibly resolve this issue, although high levels of cell death within the treated cell populations would be expected to have an increasingly greater effect on transcription as time passes. While the subtractive ' R S ' method provides a means to remove responses that are common to both SIPK-Ri and WT (thus not of interest), I cannot exclude the possibility that these ' R S ' changes are not true SIPK-dependent responses, but rather, artefacts of the silencing mechanism, or of a refractory response to an initial SIPK-Ri stress state (Chandra et al., 2000). Negative correlations between the ' R S ' list (4-hour/control) and the 'WT-only' list (4-hour/control) for each stress support the prediction that silencing of SIPK should negatively impact the expression of a set of genes (presumably SIPK-dependent) that are normally up- or down-regulated upon stress treatment.  Not surprisingly, these correlation values appear to be low (r values -> 0 2  indicating no relation) since SIPK is unlikely to be the dominant influence on the expression of the majority of genes that respond most strongly to stress in WT (i.e. comprising the 'WT-only' list).  These negative correlations are improved (r values -> -1 indicating strong negative 2  correlation) when considering the more limited set of genes considered to be SIPK-dependent. Thus, correlations between the genes expressions of the ' R S ' list (4-hour/control) and the 'WTonly' list (4-hour/control) for hrpZp h (Table 1) and P-megaspermin (Table 2) yielded r values 2  sp  of -0.43 and -0.68, respectively. However, these data should be viewed with caution since they are based upon small numbers of genes (<30). In general, genes that are most likely to be SIPKdependent based on their co-association with the H4WT and H4RS lists also exhibit a negative correlation in their expressions between these lists, providing additional evidence that these genes are SIPK-dependent.  86  Comparison of the gene annotations of the hrpZ h and P-megaspermin short-lists reveals Psp  that both lists have many stress-related genes. SIPK appears to down-regulate the cell wall degrading enzymes, pectate lyase and xyloglucan endo-transglycosylase, in response to  hrpZ  Psp  h  treatment (and to a lesser extent to p-megaspermin treatment; see Supplemental list 3), which may indicate that SEPK modulates the cellular response to threats against cell wall integrity. SEPK also appears to enhance the transcription of a number of ROS scavenging agents, as reflected in both the hrpZ h and p-megaspermin responses, although different antioxidants are Psp  represented in each short-list, with the exception of peroxidase genes.  Inference of stress-  specific antioxidant mechanisms is made difficult since the appearance of a gene exclusively in one of the lists does not imply that this gene is not similarly expressed in the other response, only that it did not meet the stringency criteria that would allow it to be included in the filtered short-list. In addition, the lack of Solanaceae array homolog characterization, coupled with the complexity/redundancy of plant antioxidant machinery (Mittler, 2002), make it challenging to identify specific SIPK-mediated antioxidant mechanisms.  However, monodehydroascorbate  reductase (MDHAR), which functions as a plant-specific intermediate in the reduction of ascorbate, is of particular interest for its preferential up-regulation in H 4 T (2-fold>M4 T - data W  W  not shown) and its selected down-regulation by SEPK-silencing in H4 s (1.5-fold>M4Rs - data R  not shown). M D H A R has recently been characterized as a cell death marker in plants, based on microarray analysis which revealed up-regulation of its cytosolic form during both heat- and senescence-induced stress (Swidzinski et al., 2002). Interestingly, M D H A R is also a putative ortholog (reciprocal best hit) to a known mammalian cell death-related protein, apoptosis inducing factor (AEF). AIF is released from the mitochondria into the cytosol in stressed cells, where it triggers a caspase-independent  cell death pathway (Yu et al., 2002).  Based on  homology, it is tempting to speculate that M D H A R may also play a role in governing programmed cell death in plants. Considering the pattern of elicitor-induced responses observed 87  in this study, such a model presents the intriguing possibility that a cell death mechanism may exist in plants that is more relevant to hrpZp h -induced cell death than P-megaspermin-induced sp  cell death.. Comparison of the SIPK-dependent gene sets for hrpZp h and p-megaspermin short-lists also sp  revealed that only two genes, encoding fibrillarin (key nucleolar protein; pre-rRNA processing) and T-complex protein 11 (putative chaperone), are common to both short-lists and exhibit similar SIPK-affected responses (Tables 1&2).  Since viruses are believed to re-organize and  re-distribute host cell fibrillarin as a means of disrupting normal rRNA processing and ribosome biogenesis (Taliansky and Robinson, 2003), up-regulation of fibrillarin may be a core plant defense required to compensate for such disruption. The present experiments did not, of course, involve viruses, but it is interesting to speculate that these transcriptional responses might form part of a general defense response against a wide range of pathogens including viruses, bacteria, and oomycetes. Since SIPK silencing leads to enhancement of fibrillarin expression, this would suggest that SIPK normally acts as a negative regulatory checkpoint for the induction of general stress responses. There is little data on the function of Arabidopsis T-complex protein 11 (not to be confused with the A. tumefaciens T-complex) other than in silico characterization, although chaperones can potentially play stress-specific roles.  3.4.5  S I P K m o d u l a t e s t r a n s c r i p t i o n o f a subset o f genes i n a n e l i c i t o r - s p e c i f i c manner  While the genes found in Tables 1 and 2 are those whose expression during elicitation is most profoundly affected by suppression of SIPK (based on fold-change magnitude), and the genes populating Table 1&2 appear to differ from each other, they do not collectively define a set of genes for which SIPK-dependent responses are exclusive to either hrpZp responses.  spn  and P-megaspermin  In fact, many subtle yet biologically-significant responses were either eliminated by 88  the stringent A N O V A filtration (p-value<5.35E-08) or are difficult to recognize based on patterns within the hierarchical cluster image (Figure 3.5). The filter devised to expose these potentially subtle SIPK-dependent, elicitor-specific response differences yielded four data sets that largely conformed to the four possible elicitor-specific expression profiles: | H only, J,H only, f M only, -IM only. While the ANOVA-filtration revealed clusters (I-VI) that demonstrate heterogeneity of response amongst 1000 genes, this more focused approach yields a much shorter list of 239 genes (192 without redundancies) that behave disparately between hrpZpsph and P-megaspermin responses. The majority of these genes are grouped by hierarchical linkage clustering (HCL) into large contiguous clusters that responded significantly to either elicitor (but not both). Since H C L established four such response clusters with approximately the same population size (~40genes) it can be concluded that SIPK both up- and down-regulates genes in the induced responses to both elicitors, i.e. that SEPK activity underpins discrete, elicitor-specific transcriptional reprogramming. While the Solanaceae microarray is a very useful tool for probing expression of tobacco genes, there is only a limited amount of information available concerning the biological functions of most tobacco genes. On the other hand, the functional annotation of Arabidopsis genes is far better developed, and high through-put analyses of putative gene function are possible using resources such as the GO S L I M annotation tool. If we accept the assumption that many plant gene homologs play similar roles in different species, we should usually be able to infer tobacco gene function from GO SLIM annotations for the most homologous Arabidopsis gene models. Gene ontology characterization of Clusters A - D carried out in this manner revealed that SEPK preferentially up- and down-regulates functionally distinct classes of genes in response to each of the two elicitors tested. For example, hrpZp h-induced responses appear sp  89  to involve SIPK as a positive regulator for the expression of several "kinase activity" annotated genes, while this pattern was not observed for the P-megaspermin response. On the other hand, the SIPK-dependent P-megaspermin-specific response is marked by selective down-regulation of "transcription factor" annotated genes, which may indicate that the plant's response to P-megaspermin does not require extensive transcription factor expression as a means of enacting defense responses.  There is also some evidence that, compared to harpin,  P-megaspermin induces a more radical transcriptional response amongst a smaller number of genes. While both elicitors induced a similar magnitude of transcriptional change in WT (measured as the mean of the absolute values of all fold changes in each W T response; 1.3 (H4WT) and 1.28 (M4WT)), P-megaspermin has more significantly-induced transcriptional changes (Figure 3.7B). Taken together, this evidence suggests that hrpZp h and P-megaspermin sp  may be utilizing SIPK to enact very different transcriptional reprogramming towards divergent defense response outcomes. It might be predicted that genes within each of these elicitor-specific clusters (A-D) should be governed to some degree by a common SIPK-dependent gene regulation mechanism. However, motif analysis of the upstream sequences for the Arabidopsis homologs of the genes in each cluster failed to detect any pattern of cluster-correlated over-represented motifs. Despite this result, it cannot be concluded that the co-expressed tobacco genes lack over-represented motifs or that they are not co-regulated.  Approaches with web-based algorithms such as Motif  Analysis and Motif Sampler are currently incapable of detecting "higher-order" cis-element organization which is based upon both frequency and position of known plant cis-acting elements.  Perhaps more importantly, it is not clear to what extent it is possible to detect  meaningful cis-element similarities when substituting Arabidopsis gene models for tobacco genes whose expression patterns have been assayed via potato microarrays. This problem is particularly acute for genes that belong to large families in which family members have 90  developed divergent functions (reflected in their regulatory sequences) even for structurally very closely related coding sequences. In contrast to the broadly transferable GO SLIM information, Arabidopsis cis-region sequences are much less likely to provide meaningful clues about tobacco gene regulation due to the more rapid evolution of the regulatory regions.  3.4.6 Conclusion Overall, my study has provided evidence that  hrpZ  Psp  h  and P-megaspermin serve as distinct  phytopathogenic cues to which tobacco cell suspension cultures respond in a largely cue-specific manner. Investigation of differential expression patterns in the SIPK-Ri and W T cell suspension cultures, and gene ontology characterization based upon clustering of co-expressed genes, provide strong evidence that SIPK plays different roles according to the type of stress.  In  addition, the expression data generated in these experiments will likely provide a useful resource in the future, as more information is gathered about the detailed structure and biological functions of more tobacco genes.  91  Future Directions In an effort to learn more about plant M A P kinase signalling mechanisms, this research project examined the role(s) of SIPK in mediating stress responses induced by elicitors derived from two divergent non-host pathogens.  The results support earlier findings that SEPK and  WEPK act coordinately during response to elicitors. It is likely that there is an interdependence of function between these two M A P K s , making it difficult to fully characterize the roles of SIPK in defense response without knowing more about the role(s) of WEPK.  Stable silencing of  WEPK would enable demonstration of the relation of WEPK to stress signals, as was done with SEPK (Chapter 2), to ROS and S A in stress response.  Such investigations should then be  followed up by microarray analysis to assess the impact of silencing WIPK on the transcriptional response to several stresses. It was unexpected that the silencing of SIPK has little effect on B-megaspermin-induced lesion formation, while greatly increasing the sensitivity of tobacco leaves to hrpZp h infiltration. sp  Given the large impact that SIPK silencing has on B-megaspermin-induced transcriptional changes observed in cultured tobacco cells, it is likely that investigation of the causal relationships between SIPK, WIPK, ROS and SA in the context of B-megaspermin treatment would yield interesting contrasts to that of the hrpZ h treatment. Psp  It would also be worth characterizing the differences in stress response between autotrophic plant tissues and undifferentiated, heterotrophic cells grown in suspension cultures, when treated with either hrpZp  sph  or B-megaspermin. This could initially be addressed by transcriptome  profiling of elicitor-infiltrated plant tissues.  While in planta transcriptome profiling would  provide the ideal biological context to investigate elicitor responses of tobacco, transcriptional differences between cell types and/or defense response zones (such as HR, L A R , and SAR) could potentially obscure SIPK-dependent transcriptional changes important in cell fate determination. To circumvent this problem, future efforts to characterize the transcriptome of WT and SEPK-Ri leaf tissues might focus on particular cell types and defense zones.  92  To address a more complete range of SIPK (and WIPK) involvement, the approaches presented in this thesis can be extended to include a more diverse set of elicitors, wholepathogens, and abiotic stresses. Variation in the degree of tissue exposure to these stresses would likely be instructive, as SIPK function appears to affect the sensitivity of plants to HR. Such a multiple-stress study would offer tremendous potential to dissect the predictably complex SIPK-dependent defense response pathways. This thesis provides several short-lists of genes that are very likely to be SIPK-dependent in their response to hrpZp p and/or fJ-megaspermin. Thus, these are genes that are likely to be S  h  important in coordinating responses to pathogens. Despite the limitations inherent to a crossspecies hybridization where neither genome has been sequenced, such lists provide a practical entry point to the cloning, characterization, and genetic manipulation of tobacco genes that are likely to be the executors of SIPK-mediated stress signalling.  M y transcriptome profiling  suggests that SIPK has elicitor-specific roles and it seems likely that many of these genes could be useful for developing pathogen-specific resistance strategies for Solanaceae at some point in the future. For other angiosperm species, M A P K s orthologous to SIPK, such as A t M P K 6 (Arabidopsis thaiiana), L e M P K l  (Lycopersicon peruvianum), and ShMK (Medicago sativa), are also  involved in plant responses to biotic stimuli. Since orthologs to WIPK are also present, and stress-induced, it is possible to consider a comparative genomic study of stress-specific kinase roles of SIPK and WIPK orthologs.  While tobacco has been a useful model organism for  studying plant-pathogen interactions, M A P K roles in stress responses, such as transcriptional reprogramming, may be more precisely defined in Arabidopsis where genes within a fullysequenced genome have been far more extensively characterized. Insights into defense-related M A P K roles that are gained through study of tobacco and Arabidopsis will undoubtedly facilitate novel strategies in controlling plant pathogens.  93  4 Appendix 4.1 Validation of microarray expression values through qRT-PCR In order to validate the expression ratios derived from the co-hybridizations, quantitative reverse transcription polymerase chain reaction (qRT-PCR ) was employed as an alternate means to measure the transcript abundance. Since two-channel c D N A array design necessitates interpretation of expression ratios of the two hybridized samples, the absolute measures of transcript abundance from qRT-PCR for different samples were combined as in the microarray experiment to yield SIPK7WT ratios (log2) that were directly comparable to those of the potato c D N A array (Figure 4.1).  Most ratios suggest a similar relative abundance (SIPK-Ri < or >  WT), while inconsistencies are often obscured by large 80% C.I. intervals for qRT-PCR data. Notably, tobacco Prlb gene qRT-PCR ratios for each treatment are opposite in magnitude to the microarray ratios, suggesting that a similar, but distinct homolog of Prlb is producing this opposite result in the microarrays. The SIPK-affected responses appear to be consistent (up- or down-regulation) for the majority of genes, as measured by the difference between the treatments and the control ratio, (see the additional copies of the figure with the trend lines super-imposed for either elicitor response). Harpin responses were consistent between techniques for all but secretory peroxidase and possibly PRT1, G protein alpha subunit 3, and NtF4 (curiously not Prlb). B-megaspermin responses were consistent between techniques for all but P R l b and arguably thioredoxin, secretory peroxidase and NtF4.  94  SAGT AF (9063* (TC71881)  guanylate kinase /<lf206(30(TCei737)  shaggy protein kinase V08S0?) (TC67BQ0)  m [  14%  to  i' ' J j  '  t  ,.  1  3  PRT1 (elF-3 subunit) Q4055«(TC69741)  2 T  I 0  -  **  •I  •2  1  •3 •4 thioredoxin H-type P29449 (TC68S46)  25  G protein alpha subunit 3 AF249742(1C58S0S)  2 Ii r  .  0  I  \  k  1 OS  r [  f  05  J  -1  0  p  -2  I*  • 15  05  0  •05  -1 -2  -10 SIPK ,48098730 0X67874)  -I-15  -8  •2  plasmodesmal receptor /1F307095(TCSS305)  05  H  -4  15' 1  I  -6  •25  PR1I) S22531 (TC57935)  2  secretory peroxidase 4F1492SnTC66034)  •25 NtF4 (SIPK-homolog) Q40532 (TC672S9)  4  0  1.5  WIPK AB052964 CTC70862)  1 0.5  -05  0  -1  •0.5 •1  •15  -1.5 •2  -2  C  H4  H8 M4 M 8  C  C  H4 H8 M4 M8  H4 H8 M4 M 8  C  H4 H8 M4 M 8  Treatment  Figure 4.1 Validation of potato microarray expression ratios by qRT-PCR for 12 selected tobacco genes with common R N A | samples. tota  Clear bars represent normalized C(T) ratios while hatched bars represent log2 ratios. Each tobacco sequence is identified by common names (bold), accession numbers (italics) and TC #s (brackets) for associated potato microarray clones. Error bars represent 80% confidence intervals for qRT-PCR (n = 3) and microarray (n = 6) values.  95  Table 5: Primer sequences for 13O-150bp regions of known tobacco genes for qRT-PCR validation of microarray expression data Tobacco Gene  Accession  SIPK  NTU94192  WIPK  AB052964  NtF4  AB062139  UDP^glucoseisalicylic acid glucosyltransferase Shaggy protein kinase6 Guanylate kinase Eukaryotic translation initiation factor 3 subunit 10 (elF-3theta) Thioredoxin H-type 1 (TRX-H-1) G prptein alpha subunit 3 Secretory peroxidase  tm  THH1_TOBAC AY063128 AF149251  S22531  gc%°  :percent of G or C bases in the primer  seq  :primer sequence (5' to 3')  d  IF3A_TOBAC  Pathogenesis-related: protein 1b  jmelting temperature of primer (°C)  b  AF205130  AF307095  :primer length (base pairs)  a  T03601  Non-cell-autonomous protein pathway^  Actin (control)  len  AF190634  TOB54  Primer Orientation LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT LEFT RIGHT  le'if  tin  gC%  24 21 20 20 20 20 18 20 20 20 19 20 21 20 20 20 25 20 20 18 20 21 20 19 22 24  59.3 59.6 60.3 59.5 60.0 60.5 59.3 59.2 60.4 58.5 60.2 58.6 57.8 58.4 59.5 59.7 59.5 60.6 59.1 58.8 59.9 60.0 59.3 60.1 59.5 58.0  41.7 47.6 45.0 45.0 45.0 45.0 61.1 55.0 45:0 45.0 47.4 55.0 38.1 50.0 45.0 45.0 40.0 45:0 50.0 61.1 50.0 52.4 40.0 57.9 50.0 41.7  d  C  seq  CCATGTATATATGTGTGCACTTCG CAGAACCAATACAAGCGATCC CCCTGAI I I ICCTTCGGTTT CAATTCCATAAGCACCACGA CAACAACACGCACAI I I ICC TGAGCTGGACCATCCAI I 11 CGAGGCTATCTCCGATGG GCAACTCACAGGACAACCAC AATGGAACGGAAACTGGTCA AGCACTTCGCCTGAAATACA CCAAAAGCGTCTCCGAAAT CGAGCAGTCTTCACACCTTC GCACTTACGAACAAAAATGGA CGAGCCACTACCAAATGAAG AAGGTTGAGGAATGGAACGA ATGGGGCATCTTCTTAGCAA CCTATGCTGTGTGTGGTATTAGAAA TATGCTTGTCCGCCTTTGTT CTGGAGTTGTTGCCTTGCTA GTCTGGGATTGGGTCAGG TGATGGTGAAGAAGGGTTCC TGTTCCAGTAAGGGTGGTGAG TCI I I ICACAAATGCCCTCA ATAGGCTGCTACCCCGTTG CAGGTATTGTGTTGGACTCTGG ATCTTCATCAGGCTATCAGTAAGG  4.2 HrpZ and P-megaspermin-induced cell death in WT and SIPKRi cell suspension culture Psph  O  H Treatment WT  S I P K - R i Figure 4.2 H r p Z  P s p h  and P-megaspermin affects on cell viability.  Cell death (%) in WT and SIPK-silenced (black) tobacco cell suspension cultures 24-hours after treatment with 40 ug/ml harpin (H) and 9 ug/ml P-megaspermin (M). Values for each bar are the number of cells counted (2 biological replicates per treatment)  97  4.3 Gene ontology summaries for shortlists ofpotato array sequences  Figure 4.3 Frequency of occurrence of molecular function gene ontology (GO) terms for considered gene lists (see section 3.3.5) based upon TIGR Solanum tuberosum gene index (StGI). "Microarray" summarizes the occurrences of GO terms for the entire array. Inset values for each pie-chart indicate the number of genes in each list. 'H4WT A N D M 4 W T " summarizes the gene ontologies for the genes common to both H4WT and M4WT- G O annotations included in the array gene ID file were updated with current annotations available from the TIGR Solanum tuberosum gene index (StGI) (as of October, 2 0 0 4 ) .  98  4.4 Explanation of correlations performed on short-lists Filters, such as fold-change or p-value cut-offs, can be used to find the most dramatic responses induced by either treatment. Several possible methods can be used to assess the similarity of transcriptome responses to these two different treatments by directly comparing the content of the filtered shortlists derived from those treatments.  In the model case shown here  (Figure 4), 12 genes exhibit expression changes that are more than 2-fold (up or down-) in each H4  R S  and M 4  R S  responses (considered separately). " " R S  • " R S  a  a1 b1  c d e  d  b  d1 e1  fl gi hi  f  9  h  i  i1  ii ki  j k  d  1  Figure 4: Two arbitrary short-lists of genes with >2-fold differentials for each treatment.  By direct comparison of the content, we see that only 1 gene ('d') is common to both lists. However, this apparent dissimilarity may be a stochastic artifact. That is, many genes may be responding similarly to both treatments, but the response may fail to pass the filter (fold cut-off and/or p-value) in one treatment, yet pass it in the other treatment data set. This would not be surprising in situations where stringent filters have been applied, given the extent of stochastic variation in biological responses. B y relaxing the filtering stringency we can see i f the number of genes that agree between both lists is increased (Figure 5).  99  a b  c  d  d  d1 e1 fl  e f  g h i i k  |ratio|>2 fold  I  gi hi i1 J1 k1 d  m  p  n o P q r  I b n e r k  i1 b1  hi  " |ratio|>1.5-fold  a1 b1  di k1  h g  i  I q  Figure 5: Two arbitrary short-lists of genes with >1.5-fold differentials for each treatment. Many genes are now common to both lists, suggesting (perhaps) that transcriptional reprogramming induced by these two treatments is very similar, whereas the responses looked very different for a 2-fold cut-off. Of course, i f we relax the filtering too much, than we will admit all genes to both lists. In general, the less strict the filtering becomes, the less meaningful the content similarity will become between the two lists (Figure 6).  100  H4RS  M4RS  a b  el BI  «  «j very dissimilar  t  n  n  til  h  t>1  P  n  ?  [  ht di  g  less dissimilar  I  more similar  i  indistinguishable  no cut-off  Figure 6: Effect of relaxing fold-change (or p-value) cut-off on list similarity The list similarity as a function of filtering cut-off should provide an estimate of the similarity of transcriptional reprogramming across treatment responses; one that is robust against stochastic influences.  The analysis would proceed by modifying the cut-off  (independent variable) and then observing the % similarity (dependent variable). However, I am not aware of such a comparison method, from literature on microarray analysis. It is more practical, common, and intuitive to assess similarity of transcriptional responses of two treatments given a fixed cut-off.  As in Figure 7, the lines match a gene's position  between the rankings of the two lists, and line angle provides a measure of rank dissimilarity. Rank dissimilarity can also be simply computed as a difference. Large numbers (steep slopes in the figure) represent more dissimilar responses.  One can generalize about the overall  dissimilarity of two shortlists (our main focus) by generating a sum of differences for all genes 101  found in both lists, for a given cut-off. This method is restricted to addressing genes that are common to both lists at a given cut-off. In addition, this approach is further restricted to considering absolute values that do not retain the important information about whether the gene was up or down regulated, only that it responded significantly.  H4  M4  a  a1  f  P:  Figure 7: Finding commonalities across treatment short-lists. Lines connect genes that are common to both lists. However, there is a more powerful and easy way of comparing these transcriptome-level responses at a fixed cut-off that circumvents these problems. In this method, we consider all the genes that occur in either shortlists and assess the similarity of transcriptional reprogramming; that is, we include the transcriptional response of a gene even i f that gene is interesting (based upon a pre-determined cut-off) in only one response.  Obviously, it is difficult to make a strong  statement about dissimilarity between a response that we believe is significant and one that we observe as being less significant (failed our cut-off). However, Pearson correlations based on fold-change values can compare the general agreement between these two elicitor responses 102  based on a large group of genes; those that appear in either list (and not just both).  This  'collective' observation should be fairly robust despite the low level of confidence we may have in some values, but a rigorous test of this robustness would require a large set of modeled transcriptional responses that span a continuum of similarity. In Figure 8, genes found in either list (using equivalent pre-determined cutoffs) have been arranged by fold-change from most up-regulated to most down-regulated in both responses. Scatter plots (right) represent this same data, where treatment 1 (H4 s) has been represented on R  the ' x ' axis (ranked by fold-change) and corresponding treatment 2 ( M 4 ) values for each gene RS  are plotted on the ' y ' axis. Such comparisons demonstrate the extent that such comparisons fit three possible scenarios; 1) no similarity transcriptome response to both stresses [r approaches 0], 2) the genes respond 2  similarly to both treatments [r approaches 1], or 3) the genes respond in a completely opposite 2  fashion [r approaches -1]. 2  Figure 8 : AH genes occurring in either treatment short-lists, ranked by most up-regulated to most down-regulated in each treatment. Scatter plots present the three possible scenarios of list similarity 103  4.4.1  Conclusion  Several possible methods can be used to assess the similarity of transcriptome responses to these two different treatments by directly comparing the content of the filtered shortlists derived from those treatments. However, I have chosen a method of comparing filtered lists by Pearson correlation that I believe is suitable for the biological questions posed in the main body of the thesis.  104  List of supplemental material (provided on CD) Supplemental list 1: A N O V A 1000 genes altered by SIPK-silencing in at least one treatment response of harpin or P-megaspermin (>2-fold, p-value<8E-03). Include mean expression values for all treatments Supplemental list 2: Hierarchical linkage clustering of 239 genes that are significantly responsive to SIPK-silencing 4 or 8 hrs after treatments with either of hrpZp h (H) (124) or Pmegaspermin (M) (114), but not both (>3-fold, p-value< ,n = 6). Genes are labelled according to the elicitor they responded significantly to, the cluster they belong to (if any), and data for their expression corresponding to heat map intensities are provided. sp  Supplemental list 3: Raw and transformed,normalized data for all microarrays (these will be in a MGED-friendly format) TIGR Protocol 1 ( S G E D S O P 6 . 1 . 1 ) : Hybridization with Direct Labeled mRNA (Effective Date: 08/01/03) TIGR Protocol 2 (SGED_SOP_6.2.1) Hybridization with Indirectly Labeled Aminoallyl Probes (Effective Date: 08/01/03)  105  Bibliography Ahlfors, R., Macioszek, V., Rudd, J., Brosche, M . , Schlichting, R., Scheel, D., and Kangasjarvi, J. (2004). Stress hormone-independent activation and nuclear translocation of mitogen-activated protein kinases in Arabidopsis thaliana during ozone exposure. Plant J 40,512-522. Armengaud, P., Breitling, R., and Amtmann, A. (2004). The potassium-dependent transcriptome of Arabidopsis reveals a prominent role of jasmonic Acid in nutrient signalling. Plant Physiol 136, 2556-2576. Asai, T., Tena, G., Plotnikova, J., Willmann, M.R., Chiu, W.L., Gomez-Gomez, L., Boiler, T., Ausubel, F.M., and Sheen, J. (2002). M A P kinase signalling cascade in Arabidopsis innate immunity. Nature 415, 977-983. Baillieul, F., De Ruffray, P., and Kauffmann, S. (2003). Molecular Cloning and Biological Activity of alpha-, P-, and gamma- Megaspermin, Three Elicitins Secreted by Phytophthora megasperma H20. Plant Physiol 131, 155-166. Baker, C.J., Orlandi, E.W., and Mock, N.M. (1993). Harpin, A n Elicitor of the Hypersensitive Response in Tobacco Caused by Erwinia amylovora, Elicits Active Oxygen Production in Suspension Cells. Plant Physiol 102, 1341-1344. Benjamini, Y., and Hochberg, Y. (1995). Controlling the False Discovery Rate - a Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B-Methodological 57, 289-300. Bestwick, C.S., Brown, I.R., Bennett, M.H., and Mansfield, J.W. (1997). Localization of hydrogen peroxide accumulation during the hypersensitive reaction of lettuce cells to Pseudomonas syringae pv phaseolicola. Plant Cell 9,209-221. Blume, B., Nurnberger, T., Nass, N., and Scheel, D. (2000). Receptor-mediated increase in cytoplasmic free calcium required for activation of pathogen defense in parsley. Plant Cell 12, 1425-1440. Bolwell, G.P., Blee, K.A., Butt, V.S., Davies, D.R., Gardner, S.L., Gerrish, C , Minibayeva, F., Rowntree, E . G . , and Wojtaszek, P. (1999). Recent Advances in Understanding the Origin of the Apoplastic Oxidative Burst in Plant Cells. Free Radical Research 31, S137-S145. Cardinale, F., Meskiene, I., Ouaked, F., and Hirt, H. (2002). Convergence and divergence of stress-induced mitogen-activated protein kinase signalling pathways at the level of two distinct mitogen-activated protein kinase kinases. Plant Cell 14, 703-711. Cardinale, F., Jonak, C , Ligterink, W., Niehaus, K., Boiler, T., and Hirt, H. (2000). Differential activation of four specific M A P K pathways by distinct elicitors. J Biol Chem 275, 36734-36740.  106  Chandra, S., Cessna, S.G., Yahraus, T., Devine, R., and Low, P.S. (2000). Homologous and heterologous desensitization and synergy in pathways leading to the soybean oxidative burst. Planta 211, 736-742. Chen, W.Q., Provart, N.J., Glazebrook, J., Katagiri, F., Chang, H.S., Eulgem, T., Mauch, F., Luan, S., Zou, G.Z., Whitham, S.A., Budworth, P.R., Tao, Y., Xie, Z.Y., Chen, X., Lam, S., Kreps, J.A., Harper, J.F., Si-Ammour, A., Mauch-Mani, B., Heinlein, M . , Kobayashi, K., Hohn, T., Dangl, J.L., Wang, X., and Zhu, T. (2002). Expression Profile Matrix of Arabidopsis Transcription Factor Genes Suggests Their Putative Functions in Response to Environmental Stresses. Plant Cell 14, 559-574. Chinnusamy, V., Schumaker, K., and Zhu, J.K. (2004). Molecular genetic perspectives on cross Clegg, A. (2004). Titles. In Agricultural Sciences (University of British Columbia), pg. 108 Cordelier, S., de Ruffray, P., Fritig, B., and Kauffmann, S. (2003). Biological and molecular comparison between localized and systemic acquired resistance induced in tobacco by a Phytophthora megasperma glycoprotein elicitin. Plant M o l Biol 51,109-118. de Jong, C.F., Laxalt, A.M., Bargmann, B.O., de Wit, P.J., Joosten, M.H., and Munnik, T. (2004). Phosphatidic acid accumulation is an early response in the Cf-4/Avr4 interaction. Plant J 39, 1-12. Delledonne, M . , Xia, Y., Dixon, R.A., and Lamb, C. (1998). Nitric oxide functions as a signal in plant disease resistance. Nature 394, 585-588. Desikan, R., A-H-Mackerness, S., Hancock, J.T., and Neill, S.J. (2001a). Regulation of the Arabidopsis transcriptome by oxidative stress. Plant Physiol 127, 159-172. Desikan, R., Clarke, A., Atherfold, P., Hancock, J.T., and Neill, S.J. (1999). Harpin induces mitogen-activated protein kinase activity during defense responses in Arabidopsis thaiiana suspension cultures. Planta 210, 97-103. Desikan, R., Hancock, J.T., Coffey, M.J., and Neill, S.J. (1996). Generation of active oxygen in elicited cells of Arabidopsis thaiiana is mediated by a N A D P H oxidase-like enzyme. FEBS Lett 382, 213-217. Desikan, R., Hancock, J.T., Ichimura, K., Shinozaki, K., and Neill, S.J. (2001b). Harpin induces activation of the Arabidopsis mitogen-activated protein kinases AtMPK4 and AtMPK6. Plant Physiol 126, 1579-1587. Desikan, R., Reynolds, A., Hancock, J.T., and Neill, S.J. (1998). Harpin and hydrogen peroxide both initiate programmed cell death but have differential effects on defense gene expression in Arabidopsis suspension cultures. Biochem J 330 ( Pt 1), 115-120.  107  Di Gaspero, G., and Cipriani, G. (2003). Nucleotide binding site/leucine-rich repeats, Ptolike and receptor-like kinases related to disease resistance in grapevine. M o l Genet Genomics 269, 612-623. Dong, H., Delaney, T.P., Bauer, D.W., and Beer, S.V. (1999). Harpin induces disease resistance in Arabidopsis through the systemic acquired resistance pathway mediated by salicylic acid and the NIM1 gene. Plant J 20, 207-215. Dorey, S., Kopp, M . , Geoffroy, P., Fritig, B., and Kauffmann, S. (1999). Hydrogen peroxide from the oxidative burst is neither necessary nor sufficient for hypersensitive cell death induction, phenylalanine ammonia lyase stimulation, salicylic acid accumulation, or scopoletin consumption in cultured tobacco cells treated with elicitin. Plant Physiol 121, 163-172. Durner, J., Wendehenne, D., and Klessig, D.F. (1998). Defense gene induction in tobacco by nitric oxide, cyclic G M P , and cyclic ADP-ribose. Proc Natl Acad Sci U S A 95, 1032810333. Dutilleul, C , Garmier, M . , Noctor, G., Mathieu, C , Chetrit, P., Foyer, C.H., and De Paepe, R. (2003). Leaf Mitochondria Modulate Whole Cell Redox Homeostasis, Set Antioxidant Capacity, and Determine Stress Resistance through Altered Signalling and Diumal Regulation. Plant Cell 15, 1212-1226. Farmer, E.E., and Ryan, C A . (1990). Interplant communication: airborne methyl jasmonate induces synthesis of proteinase inhibitors in plant leaves. Proc Natl Acad Sci U S A 87, 7713-7716. Farmer, E.E., and Ryan, C A . (1992). Octadecanoid Precursors of Jasmonic Acid Activate the Synthesis of Wound-Inducible Proteinase Inhibitors. Plant Cell 4, 129-134. Felix, G., Duran, J.D., Volko, S., and Boiler, T. (1999). Plants have a sensitive perception system for the most conserved domain of bacterial flagelhn. Plant J 18, 265-276. Galan, J.E., and Collmer, A. (1999). Type III secretion machines: Bacterial devices for protein delivery into host cells. Science 284, 1322-1328. Gomez-Gomez, L . , and Boiler, T. (2000). FLS2: A n L R R receptor-like kinase involved in the perception of the bacterial elicitor flagellin in Arabidopsis. Mol Cell 5, 1003-1011. Hahlbrock, K , Bednarek, P., Ciolkowski, I., Hamberger, B., Heise, A., Liedgens, H., Logemann, E., Nurnberger, T., Schmelzer, E., Somssich, I.E., and Tan, J. (2003). Nonself recognition, transcriptional reprogramming, and secondary metabolite accumulation during plant/pathogen interactions. Proc Natl Acad Sci U S A 100 Suppl 2, 14569-14576. Hammond-Kosack, K.E., and Jones, J.D. (1996). Resistance gene-dependent plant defense responses. Plant Cell 8, 1773-1791.  108  Haruta, M., and Constabel, C P . (2003). Rapid alkalinization factors in poplar cell cultures. Peptide isolation, cDNA cloning, and differential expression in leaves and methyl jasmonate-treated cells. Plant Physiol 131, 814-823. Heath, M.C. (2000a). Nonhost resistance and nonspecific plant defenses. Curr Opin Plant Biol 3,3i5-319. Heath, M.C. (2000b). Hypersensitive response-related death. Plant M o l Biol 44, 321-334. Holley, S.R., Yalamanchili, R.D., Moura, D.S., Ryan, C.A., and Stratmann, J.W. (2003). Convergence of signalling pathways induced by systemin, oligosaccharide elicitors, and ultraviolet-B radiation at the level of mitogen-activated protein kinases in Lycopersicon peruvianum suspension-cultured cells. Plant Physiol 132, 1728-1738. Holub, E.B., and Cooper, A. (2004). Matrix reinvention in plants: how genetics is unveiling secrets of non-host disease resistance. Trends Plant Sci 9,211-214. Hoyos, M.E., and Zhang, S. (2000). Calcium-independent activation of salicylic, acidinduced protein kinase and a 40-kilodalton protein kinase by hyperosmotic stress. Plant Physiol 122, 1355-1363. Hrabak, E . M . , Chan, C.W.M., Gribskov, M . , Harper, J.F., Choi, J.H., Halford, N., Kudla, J., Luan, S., Nimmo, H.G., Sussman, M.R., Thomas, M., Walker-Simmons, K., Zhu, J.K., and Harmon, A.C. (2003). The Arabidopsis CDPK-SnRK superfamily of protein kinases. Plant Physiology 132, 666-680. Ichimura, K. (2002). Mitogen-activated protein kinase cascades nomenclature 7, 301-308.  in plants: a new  Ichimura, K., Mizoguchi, T., Irie, K., Morris, P., Giraudat, J., Matsumoto, K., and Shinozaki, K. (1998). Isolation of A T M E K K 1 (a M A P kinase kinase kinase)-interacting proteins and analysis of a M A P kinase cascade in Arabidopsis. Biochem Biophys Res Commun 253, 532-543. Ichimura, K., Mizoguchi, T., Yoshida, R., Yuasa, T., and Shinozaki, K. (2000). Various abiotic stresses rapidly activate Arabidopsis M A P kinases A T M P K 4 and A T M P K 6 . Plant J 24, 655-665. Ichinose, Y., Andi, S., Doi, R., Tanaka, R., Taguchi, F., Sasabe, M . , Toyoda, K., Shiraishi, T., and Yamada, T. (2001). Generation of hydrogen peroxide is not required for harpin-induced apoptotic cell death in tobacco B Y - 2 cell suspension culture. Plant Physiology and Biochemistry 39, 771-776. Jabs, T. (1999). Reactive Oxygen Intermediates as Mediators of Programmed Cell Death in Plants and Animals. Biochemical Pharmacology 57, 231-245.  109  Jia, Y., McAdams, S.A., Bryan, G.T., Hershey, H.P., and Valent, B. (2000). Direct interaction of resistance gene and avirulence gene products confers rice blast resistance. Embo J 19, 4004^4014. Jin, H., Liu, Y., Yang, K.Y., Kim, C.Y., Baker, B., and Zhang, S. (2003). Function of a mitogen-activated protein kinase pathway in N gene- mediated resistance in tobacco. Plant J33,719-731. John, M . , Rohrig, H., Schmidt, J., Walden, R., and Schell, J. (1997). Cell signalling by oligosaccharides. Trends in Plant Science 2,111-115. Jonak, C , and Hirt, H. (2002). Glycogen synthase kinase 3/SHAGGY-like kinases in plants: an emerging family with novel functions. Trends Plant Sci 7, 457-461. Jonak, C , Okresz, L., Bogre, L., and Hirt, H. (2002). Complexity, Cross Talk and Integration of Plant M A P Kinase Signalling. Curr Opin Plant Biol 5, 415. Koricheva, J. (2004). Meta-analysis of Trade-offs among Plant Antiherbivore Defenses: Are Plants Jacks-of-All-Trades, Masters of All? A m Nat 163, E64-75. Kovtun, Y., Chiu, W.L., Tena, G., and Sheen, J. (2000). Functional analysis of oxidative stress-activated mitogen-activated protein kinase cascade in plants. Proc Natl Acad Sci U S A 97, 2940-2945. Kumar, D., and Klessig, D.F. (2000). Differential induction of tobacco M A P kinases by the defense signals nitric oxide, salicylic acid, ethylene, and jasmonic acid. M o l Plant Microbe Interact 13, 347-351. Lahaye, T. (2004). Illuminating the molecular basis of gene-for-gene resistance; Arabidopsis thaliana RRS1-R and its interaction with Ralstonia solanacearum popP2. Trends Plant Sci 9, 1-4. Lam, E . (2004). Controlled cell death, plant survival and development. Nat Rev M o l Cell Biol 5, 305-315. Lamattina, L., Garcia-Mata, C , Graziano, M . , and Pagnussat, G. (2003). NITRIC OXIDE: The Versatility of an Extensive Signal Molecule. Annual Review of Plant Biology 54, 109-136. Lebrun-Garcia, A., Chiltz, A., Gout, E . , Bligny, R., and Pugin, A. (2002). Questioning the role of salicylic acid and cytosolic acidification in mitogen-activated protein kinase activation induced by cryptogein in tobacco cells. Planta 214, 792-797. Lee, J., Klessig, D.F., and Nurnberger, T. (2001a). A harpin binding site in tobacco plasma membranes mediates activation of the pathogenesis-related gene HIN1 independent of extracellular calcium but dependent on mitogen-activated protein kinase activity. Plant Cell 13, 1079-1093.  110  Lee, J., Klusener, B., Tsiamis, G., Stevens, C., Neyt, C., Tampakaki, A.P., Panopoulos, N.J., Noller, J., Weiler, E.W., Cornelis, G.R., Mansfield, J.W., and Nurnberger, T. (2001b). HrpZ(Psph) from the plant pathogen Pseudomonas syringae pv. phaseolicola binds to lipid bilayers and forms an ion-conducting pore in vitro. Proc Natl Acad Sci U S A 98, 289-294. Levine, A., Pennell, R.I., Alvarez, M.E., Palmer, R., and Lamb, C. (1996). Calciummediated apoptosis in a plant hypersensitive disease resistance response. Curr Biol 6, 427437. Li, X., Zhang, Y., Clarke, J.D., Li, Y., and Dong, X. (1999). Identification and cloning of a negative regulator of systemic acquired resistance, SNI1, through a screen for suppressors ofnprl-1. Cell 98,329-339. Liu, Y., Jin, H., Yang, K.Y., Kim, C.Y., Baker, B., and Zhang, S. (2003). Interaction between two mitogen-activated protein kinases during tobacco defense signalling. Plant J 34, 149-160. Mahalingam, R., Gomez-Buitrago, A., Eckardt, N., Shah, N., Guevara-Garcia, A., Day, P., Raina, R., and Fedoroff, N.V. (2003). Characterizing the Stress/Defense Transcriptome of Arabidopsis. Genome Biology 4, art. no.-R20. Miles, G.P., Samuel, M.A., and Ellis, B.E. (2002). Suramin Inhibits Oxidant Signalling in Tobacco Suspension- Cultured Cells. Plant Cell and Environment 25, 521-527. Mittler, R. (2002). Oxidative stress, antioxidants and stress tolerance. Trends Plant Sci 7, 405-410. Mittler, R., Herr, E.H., Orvar, B.L., van Camp, W., Willekens, H., Inze, D., and Ellis, B.E. (1999). Transgenic tobacco plants with reduced capability to detoxify reactive oxygen intermediates are hyperresponsive to pathogen infection. Proc Natl Acad Sci U S A 96, 14165-14170. Montesano, M . , Brader, G., and Palva, E.T. (2003). Pathogen derived elicitors: searching for receptors in plants. M o l Plant Pathol 4, 73-79. Neill, S.J., Desikan, R., Clarke, A., Hurst, R.D., and Hancock, J.T. (2002). Hydrogen peroxide and nitric oxide as signalling molecules in plants. J Exp Bot 53, 1237-1247. Nimchuk, Z., Eulgem, T., Holt, B.F., 3rd, and Dangl, J.L. (2003). Recognition and response in the plant immune system. Annu Rev Genet 37, 579-609. Nuhse, T.S., Peck, S.C, Hirt, H., and Boiler, T. (2000). Microbial elicitors induce activation and dual phosphorylation of the Arabidopsis thaiiana M A P K 6. J Biol Chem 275, 7521-7526. Nurnberger, T., and Scheel, D. (2001). Signal transmission in the plant immune response. Trends Plant Sci 6, 372-379.  I  l  l  Ortega, X., and Perez, L . M . (2001). Participation of the phosphoinositide metabolism in the hypersensitive response of Citrus limon against Alternaria alternata. Biol Res 34,43-50. Orvar, B.L., McPherson, J., and Ellis, B.E. (1997). Pre-activating wounding response in tobacco prior to high-level ozone exposure prevents necrotic injury. Plant J 11, 203-212. Overmyer, K., Brosche, M . , and Kangasjarvi, J. (2003). Reactive oxygen species and hormonal control of cell death. Trends Plant Sci 8, 335-342. Pellegrini, L . , Geoffroy, P., Fritig, B., and Legrand, M . (1993). Molecular cloning and expression of a new class of ortho-diphenol-O-methyltransferases induced in tobacco (Nicotiana tabacum L.) leaves by infection or elicitor treatment. Plant Physiol 103, 509-517 Pontier, D., Tronchet, M . , Rogowsky, P., Lam, E., and Roby, D. (1998). Activation of hsr203, a plant gene expressed during incompatible plant-pathogen interactions, is correlated with programmed cell death. Mol Plant Microbe Interact 11, 544-554. Popham, P.L. (1995). The effect of harpin from Erwinia amylovora on the plasmalemma of suspension-cultured tobacco cells. Physiological and molecular plant pathology., v. () ( A B B R E V TITLE = Physiol, mol. plant pathol.) 47. Potin, P., Bouarab, K., Kupper, F., and Kloareg, B. (1999). Oligosaccharide recognition signals and defense reactions in marine plant-microbe interactions. Curr Opin Microbiol 2, 276-283. Ramonell, K.M., and Somerville, S. (2002). The genomics parade of defense responses: to infinity and beyond. Curr Opin Plant Biol 5, 291-294. Romeis, T., Piedras, P., Zhang, S., Klessig, D.F., Hirt, H., and Jones, J.D. (1999). Rapid Avr9- and Cf-9 -dependent activation of M A P kinases in tobacco cell cultures and leaves: convergence of resistance gene, elicitor, wound, and salicylate responses. Plant Cell 11, 273-287. Ronning, C M . , Stegalkina, S.S., Ascenzi, R.A., Bougri, O., Hart, A.L., Utterbach, T.R., Vanaken, S.E., Riedmuller, S.B., White, J.A., Cho, J., Pertea, G.M., Lee, Y., Karamycheva, S., Sultana, R., Tsai, J., Quackenbush, J., Griffiths, H.M., Restrepo, S., Smart, C D . , Fry, W.E., Van Der Hoeven, R., Tanksley, S., Zhang, P., Jin, H., Yamamoto, M.L., Baker, B.J., and Buell, C R . (2003). Comparative analyses of potato expressed sequence tag libraries. Plant Physiol 131,419-429. Ross, A.F. (1961). Systemic acquired resistance induced by localized virus infections in plants. Virology 14, 340-358. Ryals, J.A., Neuenschwander, U.H., Willits, M.G., Molina, A., Steiner, H.Y., and Hunt, M.D. (1996). Systemic Acquired Resistance. Plant Cell 8, 1809-1819. Samuel, M.A. (2002). Ozone-induced signal transduction in tobacco. In Faculty of Agricultural Sciences (Vancouver, B.C., Canada: University of British Columbia), pp. 161. 112  Samuel, M.A., and Ellis, B.E. (2002). Double jeopardy: both overexpression and suppression of a redox- activated plant mitogen-activated protein kinase render tobacco plants ozone sensitive. Plant Cell 14, 2059-2069. Samuel, M.A., Hall, H., Krzymowska, M., Drzewiecka, K., Hennig, J., and Ellis, B.E. (2005). SIPK signalling controls multiple components of harpin-induced cell death in tobacco. Plant Journal OnLineEarly. Samuel, M.A., Miles, G.P., and Ellis, B.E. (2000). Ozone treatment rapidly activates M A P kinase signalling in plants. Plant J 22, 367-376. Schenk, P.M., Kazan, K., Wilson, I., Anderson, J.P., Richmond, T., Somerville, S.C, and Manners, J.M. (2000). Coordinated plant defense responses in Arabidopsis revealed by microarray analysis. Proceedings of the National Academy of Sciences of the United States of America 97, 11655-11660. Scrase-Field, S., and Knight, M.R. (2003). Calcium: just a chemical switch? Current Opinion in Plant Biology 6, 500-506. Shah, K., and Shokat, K.M. (2002). A chemical genetic screen for direct v-Src substrates reveals ordered assembly of a retrograde signalling pathway. Chem Biol 9, 35-47. Shiu, S.H., Karlowski, W.M., Pan, R.S., Tzeng, Y . H . , Mayer, K.F.X., and L i , W.H. (2004). Comparative analysis of the receptor-like kinase family in Arabidopsis and rice. Plant Cell 16, 1220-1234. Stone, J.M., and Walker, J . C (1995). Plant protein kinase families and signal transduction. Plant Physiol 108, 451-457. Sturn, A., Quackenbush, J., and Trajanoski, Z. (2002). Genesis: cluster analysis of microarray data. Bioinformatics 18, 207-208. Swidzinski, J.A., Sweetlove, L.J., and Leaver, C J . (2002). A custom microarray analysis of gene expression during programmed cell death in Arabidopsis thaiiana. Plant J 30, 431446. Taiz, L., and Zeiger, E. (1998). Plant physiology. (Sunderland, M A : Sinauer Associates). Takahashi, Y., Berberich, T., Miyazaki, A., Seo, S., Ohashi, Y., and Kusano, T. (2003). Spermine signalling in tobacco: activation of mitogen-activated protein kinases by spermine is mediated through mitochondrial dysfunction. Plant Journal 36, 820-829. Taliansky, M.E., and Robinson, D.J. (2003). Molecular biology of umbraviruses: phantom warriors. J Gen Virol 84, 1951-1960. Tampakaki, A.P., Fadouloglou, V.E., Gazi, A.D., Panopoulos, N.J., and Kokkinidis, M . (2004). Conserved features of type III secretion. Cell Microbiol 6, 805-816.  113  Tang, X., Frederick, R.D., Zhou, J., Halterman, D.A., Jia, Y., and Martin, G.B. (1996). Initiation of Plant Disease Resistance by Physical Interaction of AvrPto and Pto Kinase. Science 274, 2060-2063. Thijs, G., Lescot, M., Marchal, K., Rombauts, S., De Moor, B., Rouze, P., and Moreau, Y. (2001). A higher-order background model improves the detection of promoter regulatory elements by Gibbs sampling. Bioinformatics 17,1113-1122. Thomma, B.P., Nelissen, I., Eggermont, K., and Broekaert, W.F. (1999). Deficiency in phytoalexin production causes enhanced susceptibility of Arabidopsis thaiiana to the fungus Alternaria brassicicola. Plant J 19, 163-171. Tiwari, B.S., Belenghi, B., and Levine, A. (2002). Oxidative Stress Increased Respiration and Generation of Reactive Oxygen Species, Resulting in A T P Depletion, Opening of Mitochondrial Permeability Transition, and Programmed Cell Death. Plant Physiology 128, 1271-1281. Torres, M.A., Dangl, J.L., and Jones, J.D. (2002). Arabidopsis gp91phox homologues AtrbohD and AtrbohF are required for accumulation of reactive oxygen intermediates in the plant defense response. Proc Natl Acad Sci U S A 99, 517-522. Van der Biezen, E.A., and Jones, J.D. (1998). Plant disease-resistance proteins and the gene-for-gene concept. Trends Biochem Sci 23, 454-456. Vandenabeele, S., Van Der Kelen, K., Dat, J., Gadjev, I., Boonefaes, T., Morsa, S., Rottiers, P., Slooten, L., Van Montagu, M., Zabeau, M., Inze, D., and Van Breusegem, F. (2003). A comprehensive analysis of hydrogen peroxide-induced gene expression in tobacco. Proc Natl Acad Sci U S A 100, 16113-16118. VanEtten, H.D., Mansfield, J.W., Bailey, J.A., and Farmer, E.E. (1994). Two Classes of Plant Antibiotics: Phytoalexins versus "Phytoanticipins". Plant Cell 6,1191-1192. Wang, K . L . , Li, H., and Ecker, J.R. (2002). Ethylene biosynthesis and signalling networks. Plant Cell 14 Suppl, S131-151. Ward, E.R., Uknes, S.J., Williams, S.C, Dincher, S.S., Wiederhold, D.L., Alexander, D . C , Ahl-Goy, P., Metraux, J.P., and Ryals, J.A. (1991). Coordinate Gene Activity in Response to Agents That Induce Systemic Acquired Resistance. Plant Cell 3, 1085-1094. Wei, Z.M., Laby, R.J., Zumoff, C.H., Bauer, D.W., He, S.Y., Collmer, A., and Beer, S.V. (1992). Harpin, elicitor of the hypersensitive response produced by the plant pathogen Erwinia amylovora. Science 257, 85-88. Wendehenne, D., Binet, M.N., Blein, J.P., Ricci, P., and Pugin, A. (1995). Evidence for Specific, High-Affinity Binding-Sites for a Proteinaceous Elicitor in Tobacco PlasmaMembrane. Febs Letters 374, 203-207.  114  Xie, Z., and Chen, Z. (2000). Harpin-induced hypersensitive cell death is associated with altered mitochondrial functions in tobacco cells. M o l Plant Microbe Interact 13, 183-190. Yalpani, N . , Silverman, P., Wilson, T.M., Kleier, D.A., and Raskin, I. (1991). Salicylic acid is a systemic signal and an inducer of pathogenesis-related proteins in virus-infected tobacco. Plant Cell 3, 809-818. Yang, K.Y., Liu, Y., and Zhang, S. (2001). Activation of a mitogen-activated protein kinase pathway is involved in disease resistance in tobacco. Proc Natl Acad Sci U S A 98, 741746. Yang, Y.H., and Speed, T. (2002). Design Issues for c D N A Microarray Experiments. Nature Reviews Genetics 3, 579-588. Yu, S.W., Wang, H., Poitras, M.F., Coombs, C , Bowers, W.J., Federoff, H.J., Poirier, G.G., Dawson, T.M., and Dawson, V . L . (2002). Mediation of poly(ADP-ribose) polymerase-1-dependent cell death by apoptosis-inducing factor. Science 297, 259-263. Zhang, S., and Klessig, D.F. (1997). Salicylic acid activates a 48-kD M A P kinase in tobacco. Plant Cell 9, 809-824. Zhang, S., and Klessig, D.F. (1998a). The tobacco wounding-activated mitogen-activated protein kinase is encoded by SEPK. Proc Natl Acad Sci U S A 95, 7225-7230. Zhang, S., and Klessig, D.F. (1998b). The Tobacco Wounding-Activated Mitogen-Activated Protein Kinase Is Encoded by Sipk. Proceedings of the National Academy of Sciences of the United States of America 95, 7225-7230. Zhang, S., and Klessig, D.F. (2000). Pathogen-induced M A P kinases in tobacco. Results Probl Cell Differ 27, 65-84. Zhang, S., and Liu, Y. (2001). Activation of salicylic acid-induced protein kinase, a mitogen- activated protein kinase, induces multiple defense responses in tobacco. Plant Cell 13, 1877-1889. Zhang, S., Du, H., and Klessig, D.F. (1998). Activation of the tobacco SIP kinase by both a cell wall-derived carbohydrate elicitor and purified proteinaceous elicitins from Phytophthora spp. Plant Cell 10, 435-450. Zhang, S., Liu, Y., and Klessig, D.F. (2000). Multiple levels of tobacco WEPK activation during the induction of cell death by fungal elicitins. Plant J 23, 339-347. Zottini, M . , Formentin, E., Scattolin, M . , Carimi, F., Lo Schiavo, F., and Terzi, M . (2002). Nitric Oxide Affects Plant Mitochondrial Functionality in Vivo. Febs Letters 515, 75-78.  115  


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items