vivian van oosten induced pathogen and insect resistance ...vivian van oosten induced pathogen and...

resistance in Arabidopsis:Induced pathogen and insect

Vivian van Oosten

Transcriptomics and specificity of defense

Vivian van Oosten

Induced pathogen and insect resistance in Arabidopsis:Transcriptomics and specificity of defense.

PROMOTOREN:

Prof. Dr. M. Dicke, hoogleraar in de Entomologie, Wageningen Universiteit

Prof. Dr. Ir. C.M.J. Pieterse, hoogleraar in de Plant-Microbe Interacties, Universiteit Utrecht

Prof. Dr. Ir. L.C. Van Loon, hoogleraar in de Fytopathologie, Universiteit Utrecht

PROMOTIECOMMISSIE:

Prof. Dr. L.H.W. Van der Plas Wageningen Universiteit

Dr. Ir. N.M. van Dam Nederlands Instituut voor Ecologie

Prof. Dr. M.A. Haring Universiteit van Amsterdam

Dr. P. Reymond University of Lausanne, Switzerland

Dit onderzoek is uitgevoerd binnen de onderzoekschool Experimental Plant Sciences (EPS).

Vivian van Oosten

Induced pathogen and insect resistance in Arabidopsis:Transcriptomics and specificity of defense.

Proefschrift

ter verkrijging van de graad van doctor

op gezag van de rector magnificus

van Wageningen Universiteit,

Prof. Dr. M.J. Kropff,

in het openbaar te verdedigen

OP WOENSDAG 18 APRIL 2007

des namiddags te vier uur in de Aula.

Van Oosten, Vivian R. (2007)

Induced pathogen and insect resistance in Arabidopsis:

Transcriptomics and specificity of defense

Thesis Wageningen University – with references – with summary in Dutch

ISBN: 978-90-8504-627-1

ContentsVII Abstract

IX Voorwoord

Chapter 1 15 General introduction

Chapter 2 29 Signal signature and transcriptome changes of Arabidopsis during pathogen and insect attack

Chapter 3 63 Analysis of genome-wide transcriptome changes of Arabidopsis during pathogen and insect attack using the visualization tool MapMan

Chapter 4 93 Differential effectiveness of microbially induced resistance against herbivorous insects in Arabidopsis

Chapter 5 111 Microarray analysis of insect herbivore-induced genes in Arabidopsis plants expressing rhizobacteria-induced systemic resistance

Chapter 6 133 General discussion

144 References

153 Nederlandse samenvatting

159 Curriculum vitae

160 List of publications

161 Education statement of the Graduate School Experimental Plant Sciences

AB

ST

RA

CT

Abstract

VII

An important question in plant defense signaling research is: how are plants capable of integrating

signals induced by pathogenic micro-organisms and herbivorous insects into defenses that are speci-

fically active against the attacker encountered? Plant defenses against pathogens and insects are dif-

ferentially regulated by cross-communicating signaling pathways in which salicylic acid (SA), jasmonic

acid (JA) and ethylene (ET) play key roles. To understand how plants integrate pathogen- and insect-

induced signals into specific defense responses, we monitored the dynamics of SA, JA, and ET signa-

ling and genome-wide transcriptome changes in Arabidopsis after attack by a set of microbial pathogens

and herbivorous insects with different modes of attack. Arabidopsis plants were exposed to a pathoge-

nic leaf bacterium (Pseudomonas syringae pv. tomato), a pathogenic leaf fungus (Alternaria brassici-

cola), tissue-chewing caterpillars (Pieris rapae), cell-content-feeding thrips (Frankliniella occidentalis),

or phloem-feeding aphids (Myzus persicae). Monitoring the signal signature in each plant-attacker com-

bination showed that the kinetics of SA, JA, and ET production varies greatly in both quantity and timing.

The transcriptional alterations were predominantly attacker-specific, but the processes affected sur-

prisingly similar. This indicates that the different attackers induce changes in similar plant processes

through largely non-overlapping transcriptional alterations. Yet, infestation by M. persicae induced a

transcriptional response that was opposite to those induced by the other attackers or exogenous

application of MeJA. We concluded that SA, JA, and ET play a primary role in the orchestration of the

plant’s defense response, but other regulatory mechanisms, such as pathway cross-talk and additional

attacker-induced signals, eventually shape the highly complex attacker-specific defense response.

Little is known about the spectrum of effectiveness of the different types of induced resistance

that are expressed upon attack by pathogens or insects. Because the signaling pathways that control

induced resistance against pathogens and insects partly overlap, we decided to investigate the effec-

tiveness of microbially induced resistance against the tissue-chewing herbivorous insects Pieris rapae

and Spodoptera exigua in Arabidopsis. Two types of microbially induced resistance were studied: sys-

temic acquired resistance (SAR), which is induced upon predisposal infection by necrotizing pathogens,

and rhizobacteria-mediated induced systemic resistance (ISR), which is triggered by selected strains

of non-pathogenic, root-colonizing rhizobacteria. No effect of SAR or ISR was evident on herbivore-

induced attractiveness of the parasitic wasp Cotesia rubecula, indicating that neither type of induced

resistance influenced indirect defense against these insects. In feeding experiments on whole plants,

induction of SAR and ISR significantly reduced growth and development of the generalist herbivore

S. exigua, whereas the performance of the specialist P. rapae was unaffected. The JA- and ET-respon-

sive genes PDF1.2 and HEL, which were activated upon feeding by either of the two herbivores,

showed a strongly potentiated expression pattern in SAR- and ISR-expressing plants upon feeding by

S. exigua, but not upon feeding by P. rapae. This differential priming for enhanced herbivore-induced

gene expression was confirmed in microarray experiments using a dedicated cDNA microarray con-

taining 111 insect-responsive Arabidopsis genes. These results suggest that the effectiveness of micro-

bially induced SAR and ISR against S. exigua feeding is associated with priming for enhanced defense-

related gene expression.

In conclusion, this thesis highlights the complexity of the defense signaling interactions between

plants, pathogens and insect herbivores.

Voorwoord

IX

VoorwoordAantoonbare desinteresse voor projectvoorstellen

“Gezocht: moleculair bioloog met sterke belangstelling voor interdisciplinair onderzoek enaantoonbare interesse voor insecten”. Ik - moleculair bioloog met sterke belangstelling voorinterdisciplinair onderzoek – schreef mijn sollicitatiebrief, het aspect aantoonbare interessevoor insecten negerend. De entomologen bleken wel aantoonbare interesse voor mij te hebben:niet alleen werd ik uitgenodigd op gesprek, ook mijn vliegticket – ik studeerde in Zwitserland –werd betaald.

Aan de vooravond van de sollicitatie ging ik met lang-niet-geziene vrienden aan detapas en de wijn. Om de aandacht af te leiden van de knoflookkegel en het ontbreken vanaantoonbare interesse voor insecten trok ik de volgende dag een rok respectievelijk mijnstoute schoenen aan.

Ik maakte kennis met het comité en was verbaasd: is deze vrolijk besproete man degeneaan wie ik kort geleden een email stuurde met de aanhef “geachte professor”? (Een verma-nend bericht bovendien, waarin ik hem erop wees dat ik een maand na het insturen van mijnbrief nog geen reactie had ontvangen.) En die ander was dus een studiegenoot van mijn vader.De derde zag eruit als een enorme lolbroek. Het klikte meteen.

Met veel enthousiasme lichtte ik mijn belangstelling voor interdisciplinair werk toe.Echter, naarmate het gesprek vorderde, bleek meer en meer dat er nogal wat schortte aanmijn inhoudelijke kennis van de zaak. “Heb je het projectvoorstel eigenlijk wel gelezen?,”vroeg de besproete professor. “Projectvoorstel?,” ik wist van niks. Ai, daar werd op tongengebeten en druk “aantoonbare desinteresse voor projectvoorstellen” opgekrabbeld.

Tactisch werd van onderwerp gewisseld.Wat doe je over tien jaar? “Dan werk ik in elkgeval niet meer in het wetenschappelijk onderzoek”, flapte ik eruit, doelend op mijn bredeinteresse, maar mijzelf slechts minder populair makend. Pijnlijk, ook nog aantoonbare aver-sie tegen de wetenschap!

Als ze maar wel aantoonbare interesse voor insecten heeft, zag ik de drie denken. “Jebent toch niet bang van insecten?,” vroeg de een. “Haha, welnee,” oef, dat viel mee. “Hoedoen rupsen het eigenlijk?”, vroeg de lolbroek zich af. Iedereen lachte. Dat brak de spanning.“En je vindt het niet vervelend om langdurig experimenten te doen in de windtunnel?,”wilde de derde weten. “Natuurlijk niet!,” riep ik zelfverzekerd uit, terwijl ik me afvroeg wateen windtunnel was.

Opgeluchte blikken. Het projectvoorstel kreeg ik mee voor in het vliegtuig. Drie dagenlater ging de telefoon: ik was aangenomen!

Zo begon, inmiddels dik vijf jaar geleden, mijn carrière in de plant-insect interacties.De besproete professor, Marcel Dicke, bleek net zo vrolijk als de eerste indruk die hij destijdshad gemaakt. De lolbroek, Corné Pieterse, heeft me regelmatig in mijn broek laten pissenvan het lachen. Marcel en Corné, zonder twijfel zijn jullie de belangrijkste mensen geweestvoor het slagen van mijn promotieonderzoek. Jullie enthousiasme voor wetenschap en hetgrote vertrouwen dat jullie in me hebben uitgesproken, hebben voor mij van begin tot eind

veel betekend. Kees, L.C. van Loon, de studiegenoot van mijn vader, ik ben dankbaar voorje gedetailleerde commentaren op mijn teksten in de eindfase. Soms begon ik aan mezelf tetwijfelen als je het woord “interestingly” weer eens uit alle zinnen had geschrapt. Gelukkigwist je altijd goed te onderbouwen waarom wat ik had geschreven bepaald niet “interesting”was.

Martin, mijn twin-AIO, en Remco, post-doc op dit plant-pathogeen/plant-insect project,met jullie heb ik intensief samengewerkt in de beginfase. Diep in de nacht monsters nemenin de kas van de Botanische Tuinen in Utrecht, stiekem in de ballenbak van de MacDonaldsin Veenendaal na vergaderingen in Wageningen, wespen sexen voor experimenten in de wind-tunnel. Stuk voor stuk onvergetelijke momenten!

Ento’s, bedankt voor het mij opnemen in jullie grote Ento-familie, ook al was ik er alleende eerste twee jaar echt vaak. Ik heb me altijd welkom gevoeld. Tjeerd, bedankt voor hetregelmatig meescoopen van een nasibakje met rupsen van Wageningen naar Utrecht, wat mijvele trein-plus-OV-fiets-ritjes heeft bespaard. Dick Peters, veel dank voor het tijdrovendeopkweken van tripsen, speciaal voor mijn experimenten. Frodo, Linde, ik vind het leuk dat ik met jullie bevriend ben geraakt, en hoop dat dit nog lang zo blijft. Office-mates Silas andMidori, we had good fun together, I enjoyed showing you a bit of Holland outside Wageningen.I hope to meet you again, whether in Africa, Japan, or elsewhere in this small world.

Fyto’s, jullie hebben me gehersenspoeld met schunnigheden. Ik vrees dat ik er voor het leven mee ben besmet, en houd jullie daarvoor direct verantwoordelijk. Gelukkig heb ikmet mijn kamergenootjes een wat serieuzere band gekregen. Mareike, ik ben er trots op datik paranimf mocht zijn bij je promotie. Tita en Rogier, jullie gezelligheid en interesse hebben die laatste maanden stressend achter de computer een stuk draaglijker gemaakt. Antonio,I hope we will dance again sometime. Hans, bedankt voor je geduld bij het fotograferen vanbeweeglijke beestjes. Annemart, onze avonturen met de billenman mogen niet onvermeld blijven. Ze lijken mij een uitstekend argument om niet meer te sporten, en zeker niet ’s och-tends om half acht.

Dear Natacha, I had a fantastic time in Lausanne, thanks for inviting me to stay in yourapartment, ride a real bike, discuss career perspectives for women in science and drink a goodlot of apero’s. Marco, it was great meeting you in Utrecht, and then again in Switzerland. Hipplaces will bond us forever.

Lieve vrienden, jullie waren er altijd om gezellig te bieren, te kubben of te kuieren.Peggy, jij staat niet voor niks naast me als paranimf tijdens mijn promotie. Kloostergenoot en kuiervriendin, we hebben elkaar de afgelopen jaren intensief leren kennen. Zelfs toen jein Nicaragua zat, ben ik nog even op de koffie gekomen. Dat zegt genoeg, lijkt me. Ik hoopdat we elkaar nog vaak zullen zien, ook al ben je inmiddels verhuisd naar Tilburg. Mireille,als je niet zo nodig had moeten bevallen, was jij ook een uitstekende kandidaat geweest voor de rol van paranimf. Ook met jou heb ik een sterke band opgebouwd, die begon tijdens deorganisatie van de EPS-PhD-day 2003. Onder jouw leiding heb ik regelmatig sessies vanbiertherapie ondergaan, en zo promotie- en privé- lief en leed gedeeld. Laura, Edith, Edmé,ik wens jullie veel succes bij jullie omzwervingen over de aarde, en ik geniet van die paarmomenten per jaar dat we elkaar weer treffen op Oerol of in een Chinese karaoke-bar.

Lieve familie, ik ben heel blij met jullie. Joris, je bent een heel fijne broer.We hebben

X

altijd lol met elkaar en kunnen ook met de lastigere zaken des levens bij elkaar terecht. Nietvoor niks dus dat je als paranimf naast me staat tijdens mijn promotie! Papa en mama, julliezijn altijd oprecht geïnteresseerd, en ik vind het bijzonder hoe goed we elkaar hebben lerenkennen tijdens mijn promotietijd. Bedankt dat jullie er altijd voor me zijn. Oma, ik vind hetgeweldig om met jou samen cocktails te drinken in de hipste bar van Utrecht!

Mijn aangewakkerde aantoonbare interesse voor insecten heeft me op de laatste nippervan mijn promotietijd ook nog iets meer dan een doctorstitel opgeleverd. Op 23 september2006 vond in Wageningen het festival “City of Insects” plaats, georganiseerd door de vak-groep Entomologie.Vlak nadat ik enkele gefrituurde sprinkhanen en een meelwormenquichehad gegeten, ben ik Diaz tegen het lijf gelopen, en sindsdien zijn wij uitermate onafscheidelijk.Lieve Diaz, ik hoop dat dit avontuur nog een flink lange staart krijgt!

Voorwoord

XI

15

chapter 1

GENERALINTRODUCTION

Microorganisms and insects as plant attackers

Heterotrophic organisms depend on the availability of organic nutrients and utilize eitherdead organic materials or exploit living organisms. A broad range of microbial pathogens and arthropod herbivores are adapted to utilize plants as a nutrient source. Plant pathogensinclude viruses, bacteria, fungi and oomycetes, whereas phytophagous insects range from tissue-chewing caterpillars to phloem-sucking aphids. Each microbial or insect attacker hasdeveloped sophisticated strategies to exploit either one or a few host plants (specialists) or a wide range of different plants (generalists).

Broadly, plant pathogens can be divided into those with a biotrophic and those with anecrotrophic lifestyle (Agrios, 1997). Biotrophic pathogens retrieve nutrients exclusively fromliving cells, and have usually developed a highly specialized relationship with their hosts. Incontrast, necrotrophic pathogens are mostly generalists that kill host cells in order to extractnutrients from the dead plant tissue. Since biotrophic and necrotrophic pathogens have highlydifferent lifestyles, they have developed specialized mechanisms either to keep their hostalive during infection and tissue colonization, or to kill the plant. An example of a biotrophis the oomycete Hyaloperonospora parasitica that causes downy mildew on e.g. Arabidopsis.Upon germination of sporangiospores of H. parasitica on the leaf surface, the growinghyphae penetrate the epidermal cells. Subsequently, specialized feeding structures, calledhaustoria, are formed in the mesophyll cells in order to extract nutrients from the plant cellsthat remain alive during the infection process (Koch and Slusarenko, 1990). In the earlystages of infection by the bacterial leaf pathogen Pseudomonas syringae (causing bacterialleaf spot), the host cells also remain alive, but later necrosis develops (Glazebrook, 2005).Therefore, this pathogen starts off as a biotroph, but develops into a necrotroph, a lifestyleoften referred to as hemi-biotrophic (Thaler et al., 2004). P. syringae bacteria enter leavesthrough stomata or wounds and subsequently multiply in the intercellular spaces, where theypromote nutrient release from the cells. The truly necrotrophic fungal pathogens Alternariabrassicicola (the causal agent of black leaf spot) and Botrytis cinerea (grey mould) producephytotoxins with a likely role in killing host cells (Otani et al., 1998; MacKinnon et al.,1999; Colmenares et al., 2002). After entering the plant, the fungi colonize the dead tissueand spread as soon as surrounding cells have succumbed. Thus, pathogens have developed a wide range of strategies to successfully infect and exploit plants.

Arthropod herbivores can be categorized based on their feeding style: tissue-chewing,cell-content feeding or phloem-sucking. Tissue-chewing insects, like caterpillars or beetles,can cause substantial tissue losses, such as may also occur during mechanical wounding. Lessextensive feeding damage is inflicted by stylet-feeding insects, which may have a long-lastinginteraction with plant cells. Herbivores that use a piercing/sucking mode of feeding consumelarge quantities of fluids (Walling, 2000). Cell-content feeding arthropods, like mites andthrips, use their stylets to pierce and empty mesophyll cells, leaving whitish spots of collapsedcells (Helle and Sabelis, 1985; Parker et al., 1995). The subtlest wounding is inflicted byphloem-feeding insects like aphids and whiteflies. Aphids hardly cause any tissue damage,because they carefully maneuver their stylets between the cells through the middle lamellae(Tjallingii and Hogen Esch, 1993).They occasionally puncture mesophyll cells, but once their

16

stylet has reached a sieve tube in the phloem, aphids establish a feeding site, where they canremain feeding for hours or even days (Raven, 1983; Miles, 1999). Thus, insect feeding canresult in substantial wounding of leaves, or in a more intimate, long-lasting interaction at specific feeding sites.

Evidently, for a plant to effectively defend itself against such a wide range of microbialand herbivorous attackers, it needs to be able to recognize the invader and respond with anappropriate defense. The first line of defense is formed by preexisting physical or chemicalbarriers, such as the presence of trichomes or toxic secondary metabolites. Often, preformedbarriers are sufficient to fend off both pathogens and insects. However, if this first line fails,plants can activate additional defense responses. These include physical defenses, such as thestrengthening of cell walls and chemical defenses, such as the production of anti-microbial or anti-feedant compounds. Thus, resistance against microbes and insects can be mediatedthrough defenses that are constitutively present, or through defense mechanisms that areinduced only upon attack (Van Loon, 2000; Dicke and Van Poecke, 2002). Induced defensescan be highly diverse and directed against a wide range of microbial and herbivorous attack-ers, but their regulation seems to involve only a limited number of plant signaling compounds:salicylic acid (SA), jasmonic acid (JA) and ethylene (ET) (Reymond and Farmer, 1998; Pieterseand Van Loon, 1999; Feys and Parker, 2000; Thomma et al., 2001; Kessler and Baldwin, 2002).Other plant hormones, such as abscisic acid, brassinosteroids, auxins and oxylipins, have beenreported to play roles in plant defense against pathogens and insects as well (Jameson, 2000;Farmer et al., 2003; Krishna, 2003; Thaler and Bostock, 2004; Mauch-Mani and Mauch, 2005).Since SA, JA and ET were shown to be the key players in plant defense, we will focus on thesethree signaling compounds.

Spectrum of signals: SA, JA and ET in plant defense againstmicrobes and insects

The signal molecules SA, JA and ET have been repeatedly shown to play key roles in theregulation of plant resistance responses. Genetic analyses of mutant and transgenic plantsthat are affected in the biosynthesis or perception of these compounds have provided com-pelling evidence for their role in plant defense against both microbes and insects. In general,it can be stated that SA-dependent defense is effective against biotrophic pathogens, whereasJA- and ET-dependent defenses are important for resistance against necrotrophic pathogensand herbivorous insects (Thomma et al., 2001; Dicke and Van Poecke, 2002; Glazebrook, 2005).Often, infection by microbial pathogens and attack by herbivorous insects results in the accu-mulation of SA, JA and/or ET and a concomitant activation of distinct sets of defense-relatedgenes (Maleck et al., 2000; Reymond et al., 2000; Schenk et al., 2000; Pieterse et al., 2001;Reymond et al., 2004). Pharmacological experiments revealed that exogenous application ofthese compounds often results in an enhanced level of resistance.

Plant defense against microbial pathogensThe central role of SA in plant defense became apparent with the use of transformed plants

Chapter 01: General introduction

17

constitutively expressing the bacterial NahG gene, encoding salicylate hydroxylase, whichconverts SA into inactive catechol. Tobacco and Arabidopsis NahG plants are unable to accu-mulate SA and show enhanced disease susceptibility to a broad range of oomycete, fungal,bacterial and viral pathogens (Delaney et al., 1994; Kachroo et al., 2000). Moreover, severalrecessive Arabidopsis mutants, including sid1, sid2 and pad4, do not show SA accumulationupon pathogen infection. These mutants are more susceptible to the oomycete pathogen H.parasitica (Zhou et al., 1998; Nawrath and Métraux, 1999) or the bacterial pathogenPseudomonas syringae pv. tomato (causing bacterial speck), confirming the importance ofSA in plant resistance against different types of pathogens.

Genetic evidence for the role of JA in plant defense came predominantly from analysesof mutant plants affected in the biosynthesis or perception of JA. The JA-insensitive mutantof Arabidopsis jar1 allows enhanced levels of growth of P. syringae pv. tomato in the leaves(Pieterse et al., 1998) and is more susceptible to Fusarium oxysporum (causing Fusarium wilt)(Berrocal-Lobo and Molina, 2004). Moreover, the Arabidopsis JA-response mutant coi1 dis-plays enhanced susceptibility to the necrotrophic fungi A. brassicicola and B. cinerea(Thomma et al., 1998), and the bacterium Erwinia carotovora (causal agent of bacterial softrot) (Norman-Setterblad et al., 2000). Conversely, the Arabidopsis cev1 mutant with consti-tutive JA-signaling showed less disease development upon infection with either the pathogenErysiphe cichoracearum (powdery mildew) or P. syringae (Ellis et al., 2002). These resultsindicate that JA-dependent defenses contribute to resistance against several pathogens. TheJA-biosynthesis mutant, fad3 fad7 fad8 of Arabidopsis is deficient in the biosynthesis of theJA-precursor linolenic acid and, like jar1, susceptible to normally non-pathogenic Pythiumspp. (causing root rot) (Staswick et al., 1998; Vijayan et al., 1998), indicating a role for JA in non-host resistance against microbial pathogens as well. This mutant is also impaired in the induced resistance against Cucumber mosaic virus (Ryu et al., 2004), further illustratingthe importance of JA in plant defense against different types of pathogens.

Using mutant and transgenic plants that are impaired in ethylene perception or signal-ing, ET was found to either promote or reduce disease, depending on the specific plant-attackercombination (reviewed in van Loon et al., 2006). Notably, hemi-biotrophic and necrotrophicpathogens were sensitive to ET-dependent defense mechanisms. By contrast, diseases causedby other fungi and bacteria with varying lifestyles were promoted by ET. In addition, severalplant pathogenic fungi and bacteria are capable of producing ET as a virulence factor, whichimproves their ability to colonize plant tissues.Thus, ET appears to modulate disease develop-ment in different ways, but when induced early after infection, it can play an important rolein resistance against pathogens.

Taken together, these observations illustrate the importance of SA, JA and ET in plantdefense against a wide spectrum of microbial attackers.

Plant defense against herbivorous insectsTo fend off insects, plants have adapted two distinct strategies: induced defense directedagainst the attacker, referred to as direct defense, and induced defense aimed at exploitingthe natural enemies of the attacker, referred to as indirect defense (Fig. 1). Both types ofdefense can be triggered upon insect feeding. Direct defense includes induced responses such

18

as the production of secondary chemicals or enzymes that act as toxins or feeding deterrents(Kessler and Baldwin, 2002; Howe, 2004), whereas indirect defense can involve productionof a blend of volatiles that attracts predatory or parasitic enemies of the herbivores (Turlingset al., 1995; Dicke, 1999).

FIGURE 1. Illustration of direct and indirect induced plant defense against insects.

Direct defense is aimed at the attacker, and involves the production of toxic compounds or feeding deterrents.Indirect defense is aimed at the attraction of enemies of the attacker, e.g. by the induction of volatiles that attract parasitic wasps that lay eggs in caterpillars (illustration by Remco van Poecke).

Direct defenseOne of the best-studied examples of induced direct defense against herbivores is the rapidand systemic induction of proteinase inhibitors (PIs) after wounding or insect feeding intomato (Lycopersicon esculentum) (reviewed in Howe, 2004). Upon consumption of inducedtissues by the herbivore, PIs bind to, and inhibit digestive proteases in the insect gut, leadingto reduced feeding (Farmer and Ryan, 1992). Several PI-inducing signals have been discovered,including oligogalacturonides (OGAs) and systemin (Ryan, 1992). In response to wounding,OGAs are produced from cell wall components, and the 18-amino acid peptide systemin isgenerated by cleavage from its precursor protein prosystemin. This eventually leads to JAsynthesis by the octadecanoid pathway and induction of PIs and other defense genes (Farmerand Ryan, 1992). However, the signal transduction events that couple the perception ofOGAs and systemin at the plasma membrane to the subsequent activation of JA synthesis in the chloroplast remain to be elucidated (Howe, 2004).

Recently, two other JA-inducible tomato proteins, arginase and threonine deaminase,were demonstrated to catabolize amino acids required for insect growth in the midgut of larvae of the tobacco hornworm (Manduca sexta) (Chen et al., 2005), further illustrating theimportance of JA-inducible defenses against insects. Indeed, Pieris rapae caterpillars (smallcabbage white butterfly) gained significantly more weight when they fed on the ArabidopsisJA-signaling mutant coi1 than on wild-type plants (Reymond et al., 2004). Likewise, thepopulation of the aphid Myzus persicae (green peach aphid) increased faster on coi1 than on wild-type Arabidopsis (Ellis et al., 2002). Conversely, on the constitutive JA-signaling


19

Arabidopsis mutant cev1, population levels of M. persicae were reduced (Ellis et al., 2002).The tomato mutant def1, deficient in JA biosynthesis, has a compromised resistance to thetissue-chewing M. sexta and Spodoptera exigua (beet armyworm) larvae, as well as the cell-content feeding two-spotted spider mite (Tetranychus urticae) and Western flower thrips(Frankliniella occidentalis) (Howe et al., 1996; Li et al., 2002; Thaler et al., 2002). The JA-biosynthesis mutant, fad3 fad7 fad8 of Arabidopsis, is extremely sensitive to larvae of thefungal gnat, Bradysia impatiens (McConn et al., 1997). Moreover, the ET-insensitiveArabidopsis mutant ein2 is less resistant to larvae of Spodoptera littoralis (Egyptian cottonworm) (Stotz et al., 2000). In addition, Arabidopsis mutants and transgenics that are com-promised in SA-dependent defense responses exhibit enhanced resistance against feeding by the cabbage looper Trichoplusia ni (Cui et al., 2002). Hence, whereas JA plays a mainrole, ET and SA also contribute to plant resistance against insects.

Indirect defenseInsect feeding induces the plant’s production of volatile chemicals, which are effective inattracting parasitic and predatory insects. JA is the major signaling molecule involved in theinduced production of plant volatiles (Dicke et al., 1999). Treatment of plants with JA leadsto the emission of a volatile blend that is similar, but not identical, to the blend of herbivore-infested plants. Moreover, the volatiles induced by JA treatment are attractive to carnivorousenemies of the herbivores (reviewed in Van Poecke and Dicke, 2004).

ET and SA can also play a role in indirect defense. ET was shown to enhance JA-mediatedvolatile emission in Lima bean (Horiuchi et al., 2001). Herbivores such as spider mites inducethe emission of methyl salicylate (MeSA) in many plant species (Takabayashi and Dicke,1996; Ament et al., 2004; De Boer and Dicke, 2004), which can lead to the activation of SA-inducible defense-related genes (Arimura et al., 2000; Kant et al., 2004). In line with theseresults, feeding by P. rapae larvae induced MeSA production in Arabidopsis (Van Poecke andDicke, 2002). In Arabidopsis NahG plants, MeSA was not produced upon P. rapae feeding,leading to a decreased attractiveness of the induced volatile blend to the parasitoid waspCotesia rubecula (Van Poecke, 2002; Van Poecke and Dicke, 2002). Similarly, the volatilesinduced upon feeding of P. rapae in the transgenic Arabidopsis S-12 line with reduced JAbiosynthesis were less attractive to C. rubecula (Van Poecke, 2002; Van Poecke and Dicke,2002). These results clearly illustrate that JA, ET and SA all play a role in induced indirectdefense against insects.

Microbially induced systemic resistance mechanisms in plants

Once plants have been attacked and have effectively resisted the invader, they develop anenhanced defensive capacity against further attack.Two inducible systemic resistance mechanismsagainst pathogens have been well investigated: pathogen-induced systemic acquired resistance(SAR) (Ross, 1961) and non-pathogenic rhizobacteria-mediated induced systemic resistance(ISR) (Fig. 2) (Pieterse et al., 1996; Pieterse et al., 1998).

20

FIGURE 2. Signal transduction pathways leading to the induction of ISR, SAR and the wound response (adaptedfrom Pieterse and Van Loon, 1999).

SARSAR is activated upon a predisposal infection, particularly with necrotizing pathogens, andrenders distant uninfected plant parts more resistant towards a broad spectrum of virulentpathogens, including viruses, bacteria and fungi (Kuc, 1982; Sticher et al., 1997). The onsetof pathogen-induced SAR is associated with increased levels of SA both at the infection siteand systemically (Mauch-Mani and Métraux, 1998). Moreover, SAR is associated with largetranscriptional reprogramming (Maleck et al., 2000). The activation of a specific set of genesencoding pathogenesis-related (PR) proteins is thought to contribute to the state of resistanceattained, because several PR-proteins possess antimicrobial activity (Van Loon and Van Strien,1999; Van Loon et al., 2006). Exogenous application of SA, or its functional analogues 2,6-dichloroisonicotinic acid (INA) or benzothiadiazole (BTH), induces SAR and activates thesame set of PR genes (Ryals et al., 1996). Transgenic NahG plants that cannot accumulateSA, and the recessive mutants sid1, sid2 and pad4, that are compromised in pathogen-inducedSA accumulation, are incapable of developing SAR and do not show activated PR gene expres-sion upon pathogen infection (Gaffney et al., 1993; Lawton et al., 1995; Zhou et al., 1998;Nawrath and Métraux, 1999).Thus, SA is a necessary intermediate in the SAR signaling path-way.Accordingly, SAR is predominantly effective against pathogens that are resisted by SA-dependent basal resistance mechanisms in uninduced plants (Ton et al., 2002).Transduction ofthe SA signal requires the function of the regulatory protein NPR1 (also known as NIM1 andSAI1) (Cao et al., 1994; Delaney et al., 1995; Shah et al., 1997). Interaction of NPR1 with theb-ZIP transcription factor TGA2 is required for the activation of the SA-regulated gene PR-1,suggesting that NPR1 acts by altering the activity of transcription factors (Fan and Dong, 2002).


21

ISRSelected strains of non-pathogenic, root-colonizing bacteria, notably of the genus Pseudomonas,are capable of inducing a systemic resistance (ISR) that is phenotypically similar to SAR (Van Loon et al., 1998). However, unlike SAR, rhizobacteria-mediated ISR in Arabidopsis is not associated with the induction of SA and PR gene expression (Pieterse et al., 1996).Instead, ISR requires responsiveness to both JA and ET, but the levels of these hormonesremain unchanged (Pieterse et al., 1998; Pieterse et al., 2000). Transcriptome analysis didnot reveal major changes in gene expression after establishment of ISR (Verhagen et al.,2004). However, upon pathogen challenge, ISR-expressing Arabidopsis showed potentiatedexpression of JA-responsive defense genes, indicating that these plants were sensitized toactivate these defenses in a faster and stronger fashion (Van Wees et al., 1999; Verhagen etal., 2004). This phenomenon is known as “priming” and has been demonstrated to underliedifferent types of induced resistance, including SAR (Conrath et al., 2002; 2006). Thus, theobserved effectiveness of ISR against pathogens that are resisted through JA- and ET-dependentbasal defenses in uninduced plants is based on the enhancement of JA- and ET-dependentdefenses (Ton et al., 2002). Priming of pathogen-induced genes probably allows a plant toreact more effectively to the invader encountered, because once induced the plant respondsfaster and stronger to pathogen attack (Conrath et al., 2006).

Although ISR is not associated with an increase in SA content or induction of PR-proteins,the Arabidopsis npr1 mutant was unable to mount an enhanced defense against pathogensnormally resisted by ISR. This demonstrates that, similar to SAR, effective ISR depends on afunctional NPR1 protein (Pieterse et al., 1998).Thus, NPR1 is not only required for SA-mediatedinduction of PR genes during SAR, but also for the JA- and ET-dependent activation of genesinvolved in ISR.

Cross-talk between SA-, JA- and ET-dependent signalingpathways

Ample evidence exists that the SA-, JA- and ET-dependent signal transduction path-ways cross-communicate during plant defense (Felton and Korth, 2000; Feys and Parker,2000; Pieterse et al., 2001; Kunkel and Brooks, 2002; Rojo et al., 2003; Bostock, 2005).A commonly accepted hypothesis is that the interacting signaling pathways enableplants to mount responses specifically tailored to the inducing attacker, and to optimizethe defense response when multiple attackers are present. Pathway crosstalk may be synergistic or antagonistic. The signaling compounds JA and ET have been demonstrated to act in a synergistic manner in the regulation of the plant defensin gene PDF1.2(Penninckx et al., 1998; Lorenzo et al., 2003). Moreover, SA-dependent SAR and JA-and ET-dependent ISR are additive in conferring disease resistance to pathogens,indicating that the activation of different signaling pathways can enhance plant defense(Van Wees et al., 2000). In contrast, SA has been shown to attenuate JA-dependent defenseresponses in several plant species, suggesting that SA-dependent defense against pathogenssuppresses JA-dependent defense against insect herbivores. The negative crosstalk between

22

SA and JA has received much attention. Therefore, this antagonistic interaction will bedescribed into more detail.

Antagonistic interactions between the SA and JA pathwaysIn tomato, SA has been repeatedly associated with inhibition of the expression of wound-and JA-responsive genes, such as those encoding PIs (Doherty et al., 1988; Peña-Cortés et al.,1993; Doares et al., 1995). Similarly, in Arabidopsis SA application repressed the expressionof the JA-activated geneVSP2 (Van Wees et al., 1999), encoding a vegetative storage proteinimplicated in defense against insects (Liu et al., 2005). The regulatory protein NPR1 plays animportant role in this antagonistic effect of SA on JA-responsive gene expression (Spoel et al.,2003). Furthermore, the transcription factor WRKY70 has been shown to have a dual role inArabidopsis: it acts both as an activator of SA-responsive genes and as a repressor of JA-induciblegenes, thereby integrating signals from these two pathways (Li et al., 2004).

In several cases, the suppressive effect of SA on JA-dependent genes was shown to correlate with reduced resistance against attackers that are sensitive to JA-dependent plantdefense. S. exigua larvae developed better on SA-treated tomato and Arabidopsis than onuntreated control plants, indicating that JA-dependent resistance was repressed (Thaler et al.,1999; Cipollini et al., 2004). Likewise, the SA-analogue BTH reduced resistance of tomatoagainst the corn earworm (Helicoverpa zea) (Stout et al., 1999). Thus, pharmacological ex-periments indicated that SA reduced JA-inducible gene expression as well as JA-dependentdefense. In line with these results, transgenic tobacco overexpressing phenylalanine ammo-nia lyase (PAL) and accumulating higher levels of SA than control plants (Felton et al.,1999), showed increased susceptibility to Heliothis virescens larvae (tobacco budworm). ThePAL over-expressing tobacco also had increased resistance to tobacco mosaic virus (TMV).Conversely, transgenic tobacco expressing low levels of PAL had low levels of SA and highlevels of JA, resulting in increased resistance to H. virescens and reduced resistance to TMV.These results clearly demonstrate the inverse relationship between induced resistance againstan insect and a pathogen.

Cross-resistance between pathogens and insects

The observations that SA-induced defense and the JA-signaling pathway can negatively affecteach other led to the intriguing question whether inverse resistance against pathogens and insectswould be the standard outcome of multiple infections. Is microbially induced resistance specific-ally directed against microbial pathogens, or are insect herbivores affected as well? Is herbivore-induced resistance specifically effective against insect herbivores, or does it also affectpathogens? Exogenous application of chemical inducers might not accurately reflect what happensin planta during biological induction of defense in terms of the compartmentation, concentrationand timing of signaling compounds.A number of studies have investigated the effect of microbiallyinduced defense on insects and vice versa (reviewed in Stout et al., 2006). However, due to varia-tion in the choice of plant, insect and pathogen species, the results varied, making it difficult todraw general conclusions about the response of one plant against multiple attackers.


23

Pathogen-induced defense against insectsAs a result of pathogen infection, increased susceptibility towards insect herbivore feedingwas observed in several cases. For instance, tobacco infected with SAR-inducing TMV wasmore prone to feeding by M. sexta caterpillars (Preston et al., 1999). Likewise, S. exigualarvae fed more and developed faster on peanut infected with the necrotrophic white moldfungus, Sclerotium rolfsii (Cardoza et al., 2003b). However, TMV-infected tobacco was moreresistant to the green peach aphid (M. persicae) (McIntyre et al., 1981) whereas no effect onthe tobacco aphid (Myzus nicotianae) was observed (Ajlan and Potter, 1992). Also in othercases, pathogen-induced resistance was effective or had no effect against insects. Infectionwith P. syringae pv. tomato rendered tomato more resistant to H. zea larvae, whereas infec-tion with the hemi-biotrophic oomycete Phytophthora infestans (causing late blight) had noeffect (Stout et al., 1999). Spider mite performance on tomato was reduced upon infectionwith the vascular wilt pathogen Fusarium oxysporum f.sp. lycopersici (Jongebloed et al.,1992). Similarly, spider mites performed less well on cotton infected with Verticillium dahliae(Karban et al., 1987). These results indicate that the effect of pathogen infection on resis-tance against insects varies with plant, pathogen and insect species. A plant infected by dif-ferent pathogens can become more, equally or less susceptible to the same insect.

Insect-induced defense against pathogensInsect-induced defense has also been shown to render plants more, equally and less resistantagainst pathogens, depending on the interaction. Feeding by H. zea caterpillars induced resis-tance against P. syringae in tomato (Stout et al., 1999). Previous infestation with silverleafwhitefly (Bemisia argentifolii) increased resistance of tomato against powdery mildew, butnot against TMV (Mayer et al., 2002). Rice gained resistance to rice blast, caused by Magnaporthegrisea, upon infestation with the white-backed planthopper, Sogatella furcifera (Kanno andFujita, 2003; Kanno et al., 2005). These results indicate that insects with different feedingmodes can induce resistance against various pathogens. In contrast, alfalfa infested with thealfalfa hopper, Spissistilus festinus, was more susceptible to F. oxysporum (Moellenbeck et al.,1992). Moreover, feeding by the willow leaf beetle (Plagiodera versicolora) caused undamagedleaves of willow to become less resistant to the rust Melampsora allii-fragilis than undamagedplants (Simon and Hilker, 2003). Therefore, both sap-sucking and tissue-chewing insects canrender plants more susceptible to pathogens. However, several plant-insect interactions havebeen described in which resistance to pathogens was unaffected (reviewed in Stout et al.,2006).

The question whether and how indirect defense is influenced by pathogen infection hasreceived little attention. One report shows evidence for increased attraction of the parasitoidCotesia marginiventris towards peanut infected with S.rolfsii (Cardoza et al., 2003a). Thisbehavioral difference could be related to qualitative and quantitative differences in volatilecompounds. In contrast, in fungus- (Setosphaeria turcica) infected versus uninfected maize, nodifferences in attraction of two different parasitoids were observed, even though the volatileblend was quantitatively much reduced in the infected maize (Rostas et al., 2006). This resultsuggests that pathogen infection of a host plant does not necessarily interfere with the attractionof a parasitoid.

24

In conclusion, the current literature provides varying results, showing positive, negative or no interactions between pathogen- and insect-induced defense. Since each research group hasadopted their own favorite plant-attacker model system, it has become difficult to draw generalconclusions from the diverse results obtained. Despite the fact that many studies investigatedthe effectiveness of microbe- or insect-induced defense, most of them did not determine thesignaling pathways involved. In order to understand how a plant responds and how this maylead to enhanced effectiveness against the primary and subsequent attackers, a thorough studyshould be performed on a single model plant with a range of attackers. Cross-effectivenessshould be determined using the same model plant and the same spectrum of microbial andherbivorous attackers. To understand the molecular mechanisms underlying the effectiveness,both the phytohormone signals involved and the genes expressed should be investigated. Tothis end, we started an interdisciplinary collaboration between the groups of Entomology,Wageningen University and Phytopathology, Utrecht University. Dr. Remco van Poecke startedoff establishing Arabidopsis as a model plant for studying indirect defense against insects,investigating the tritrophic interaction between Arabidopsis, several insect herbivores, amongstwhich the specialist P. rapae, and its specialist parasitic wasp C. rubecula. Making use of wild-type and mutant or transgenic Arabidopsis plants with defects in SA or JA signaling, he studiedthe effectiveness of indirect defense and correlated his findings with the genes expressed andthe volatiles synthesized (Van Poecke et al., 2001; Van Poecke, 2002; Van Poecke and Dicke,2002, 2003; Van Poecke et al., 2003). Martin de Vos studied the effectiveness of herbivore-induced resistance in Arabidopsis against the microbial pathogens A. brassicicola, P. syringaepv. tomato, Xanthomonas campestris pv. armoraciae, and turnip crinkle virus (TCV) (De Voset al., 2006). I performed a reciprocal study, in which I investigated the effect of microbiallyinduced resistance (SAR and ISR) in Arabidopsis on direct and indirect defense against twochewing insects: the specialist P. rapae and the generalist S. exigua. Together with De Vos,Van Poecke and others, I investigated the molecular signals involved and the genes expressedin Arabidopsis during pathogen infection or insect infestation (De Vos et al., 2005).

Outline of the thesis

The literature described above indicates that in general SA-dependent defense is associatedwith resistance against biotrophic pathogens, whereas JA-dependent defense is effective againstnecrotrophic pathogens and arthropod herbivores. Since SA and JA often have antagonisticeffects, it was proposed that pathogen-induced defense would make plants more susceptibleto herbivore attack. Conversely, insect-induced resistance was hypothesized to render plantsless protected against invading pathogens. Although results of some studies are in favor ofthis hypothesis, other reports show no effects or even protection against both pathogens andinsects.This indicates that biological attackers induce responses in the plant that are not iden-tical to the effect of exogenously applied hormones. To become able to predict the outcomeof specific plant-attacker interactions, a better understanding of the underlying mechanismsis required. How do the signaling pathways activated during induced plant defense eventuallylead to an effective response directed against the attacker? Particularly relevant is the question


25

how activation of one signaling pathway by a specific attacker is modulated to be effectiveagainst other attackers.

To explore the ways in which plants integrate, prioritize and coordinate induced defenseupon pathogen or insect attack, I studied the response of the model plant Arabidopsis thalianato a range of pathogens and herbivorous insects (Fig. 3). Arabidopsis has been shown to bean excellent model to study diverse plant-pathogen interactions (Meyerowitz and Sommerville,1994; Mitchell-Olds, 2001). Moreover, Arabidopsis can be used for the analysis of direct andindirect defense responses against tissue-chewing, cell-content feeding and phloem-suckinginsects (Reymond et al., 2000; Moran and Thompson, 2001; Moran et al., 2002; Van Poeckeet al., 2003; Reymond et al., 2004; Van Poecke and Dicke, 2004). The recent interest in thecommonalities and differences in pathogen- and insect-induced plant responses has led to amultitude of studies involving different plant-pathogen or plant-insect models and performedunder different experimental conditions. Here, I describe the molecular mechanisms activa-ted, and the signal transduction pathways involved, during attack by a spectrum of microbialand insect attackers in the model plant A. thaliana, integrating an entomological and phyto-pathological approach. In particular, the effectiveness of microbially induced resistance inArabidopsis on both direct and indirect defense against insects was examined.

FIGURE 3. The model plant Arabidopsis thaliana and a wide spectrum of microbial and insect attackers.

26


In chapter two, we started off to determine the degree of commonality in the responsesinduced in Arabidopsis by five different microbial and herbivorous attackers. In close collabo-ration with Martin de Vos, I monitored the quantity, composition and timing of the productionof SA, JA and ET (signal signature) upon attack by a pathogenic leaf bacterium (Pseudomonassyringae pv. tomato), a pathogenic leaf fungus (Alternaria brassicicola), tissue-chewing cater-pillars (Pieris rapae), cell-content feeding thrips (Frankliniella occidentalis) and phloem-suckingaphids (Myzus persicae). In addition, we analyzed the genome-wide transcriptome changesusing Affymetrix GeneChip microarrays.

In chapter three, I investigated the nature of the transcriptional changes induced by thefive different attackers. For this analysis, I used MapMan Software, which allows the displayof microarray data sets onto pictorial diagrams that represent metabolic pathways, cellularprocesses or enzyme families. Moreover, the statistical module facilitates the identification ofpotentially co-responsive groups of genes based on statistical differences. Together, these fea-tures of MapMan allow the identification of genes that are involved in the same pathway, aremembers of the same complex or are otherwise functionally connected.

In the fourth chapter, I asked the question whether microbially induced resistance inArabidopsis is effective against insects.To this end, I investigated the effectiveness of two typesof microbially induced broad-spectrum resistance (SAR and ISR) against the tissue-chewingcaterpillars P. rapae, a specialist on crucifers, and S. exigua, a generalist feeder. Forced feedingexperiments were performed to investigate the effect of the induction of SAR and ISR on thegrowth and development of P. rapae and S. exigua.To correlate the effectiveness of SAR andISR to defense gene expression, I examined the expression of the jasmonic acid- and ethylene-responsive genes PDF1.2 and HEL in uninduced and SAR- and ISR-expressing plants uponfeeding by P. rapae and S. exigua. In a windtunnel set-up, I deter-mined the effect of micro-bially induced SAR and ISR on herbivore-induced attraction of the parasitic wasp C. rubecula.

In chapter five, I went on to test whether genes that are activated in response to feedingby P. rapae or S. exigua were primed in ISR-expressing Arabidopsis plants. To this end, Iinvestigated the transcriptome of uninduced and ISR-expressing Arabidopsis upon P. rapaeand S. exigua feeding using a dedicated microarray enriched in genes responsive to caterpillarfeeding.

Finally, the data presented in this thesis are discussed in chapter six to provide an overviewof the interactions between microbe- and insect-induced defenses in Arabidopsis in the contextof the phytohormone signal transduction pathways involved and the associated transcriptionalalterations.

27

28

Signal signature and transcriptome changes of Arabidopsisduring pathogen and insect attack

Martin De Vos1,a,†, Vivian R. Van Oosten1,2,†, Remco M.P. Van Poecke2,b

, Johan A. Van Pelt1, Maria J.

Pozo1,c

, Martin J. Mueller3, Antony J. Buchala

4, Jean-Pierre Métraux

4, L.C. Van Loon

1, Marcel Dicke

2,

and Corné M.J. Pieterse1

1 Graduate School Experimental Plant Sciences, Section Phytopathology, Institute of Environmental

Biology, Faculty of Science, Utrecht University, P.O. Box 800.84, 3508 TB Utrecht, The Netherlands

2 Graduate School Experimental Plant Sciences, Laboratory of Entomology, Wageningen

University, P.O. Box 8031, 6700 EH Wageningen, The Netherlands

3 Pharmaceutical Biology, Julius-von-Sachs Institute of Biological Sciences, University of Wuerzburg,

D-97082 Wuerzburg, Germany

4 Department of Biology, University of Fribourg, CH-1700 Fribourg, Switzerland

† These authors contributed equally to the paper

a Current address: Boyce Thompson Institute for Plant Research, Tower Rd., Cornell Campus,

Ithaca, NY 15853, USA

b Current address: Department of Plant Biology, University of Minnesota, 220 BioSci Center, 1445

Gortner Avenue, St. Paul, MN 55108, USA

c Current address: Soil Microbiology and Symbiotic Systems, Estación Experimental del Zaidín

(CSIC), Profesor Albereda 1, 18008 Granada, Spain

Molecular Plant-Microbe Interactions 18, 923-937 (2005)

29Signal signature and transcriptomechanges of Arabidopsis during pathogenand insect attack

chapter 2

AB

ST

RA

CT Plant defenses against pathogens and insects are regulated differentially by cross-communicating

signaling pathways in which salicylic acid (SA), jasmonic acid (JA) and ethylene (ET) play key roles.

To understand how plants integrate pathogen- and insect-induced signals into specific defense

responses, we monitored the dynamics of SA, JA, and ET signaling in Arabidopsis after attack by a

set of microbial pathogens and herbivorous insects with different modes of attack. Arabidopsis plants

were exposed to a pathogenic leaf bacterium (Pseudomonas syringae pv. tomato), a pathogenic leaf

fungus (Alternaria brassicicola), tissue-chewing caterpillars (Pieris rapae), cell-content-feeding thrips

(Frankliniella occidentalis), or phloem-feeding aphids (Myzus persicae). Monitoring the signal signature

in each plant-attacker combination showed that the kinetics of SA, JA, and ET production varies greatly

in both quantity and timing. Analysis of global gene expression profiles demonstrated that the signal

signature characteristic of each Arabidopsis-attacker combination is orchestrated into a surprisingly

complex set of transcriptional alterations in which, in all cases, stress-related genes are overrepresented.

Comparison of the transcript profiles revealed that consistent changes induced by pathogens and

insects with very different modes of attack can show considerable overlap. Of all consistent changes

induced by A. brassicicola, P. rapae, and F. occidentalis, more than 50% were also induced consis-

tently by P. syringae. Notably, although these four attackers all stimulated JA biosynthesis, the majority

of the changes in JA-responsive gene expression were attacker-specific. All together our study shows

that SA, JA, and ET play a primary role in the orchestration of the plant’s defense response, but other

regulatory mechanisms, such as pathway cross-talk or additional attacker-induced signals, eventually

shape the highly complex attacker-specific defense response.

30

IntroductionPlants are abundantly present on earth and are at the basis of almost all food webs. Each ofthe approximately 300.000 plant species is attacked by a multitude of other organisms suchas insects and pathogens. The number of insect species is estimated to be in the order of 6million, 50% of which are herbivorous (Schoonhoven et al., 1998).The biodiversity of pathogenicmicroorganisms is less well characterized but it is general knowledge that plant pathogens area common threat to plants. To effectively combat invasion by microbial pathogens and herbi-vorous insects, plants have evolved sophisticated defensive strategies to “perceive” attack bypathogens and insects, and to translate this “perception” into an appropriate defensive response(Dangl and Jones, 2001; Dicke and Hilker, 2003; Pieterse and Van Loon, 2004).These induceddefense responses are regulated by a network of interconnecting signal transduction pathwaysin which salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) play key roles (Dicke andVan Poecke, 2002; Glazebrook, 2001; Pieterse and Van Loon, 1999; Reymond and Farmer,1998; Thomma et al., 2001). SA, JA, and ET accumulate in response to pathogen infection ordamage caused by insect feeding, resulting in the activation of distinct sets of defense-relatedgenes (Glazebrook et al., 2003; Reymond et al., 2004; Schenk et al., 2000). Compelling evi-dence for the significance of SA, JA, and ET in plant defense came from studies using mutantand transgenic plants affected in either SA, JA, or ET signaling (reviewed in Pieterse et al.,2001; Pozo et al., 2005). For instance, SA-defective signaling mutants and transgenics areoften more susceptible to pathogen infection than wild-type plants (Delaney et al., 1994;Nawrath and Métraux, 1999; Wildermuth et al., 2001). Blocking the response to JA generallyrenders plants more susceptible to herbivorous insects (Howe et al., 1996; Kessler et al., 2004;McConn et al., 1997), although enhanced susceptibility towards necrotrophic pathogens hasbeen reported as well (Staswick et al., 1998; Thomma et al., 1998). Furthermore, analysis ofmutants affected in ET signaling demonstrated that ET plays a modulating role in many plantdefense responses (Hoffman et al., 1999; Knoester et al., 1998; Lund et al., 1998).

While the importance of SA, JA, and ET in induced plant defense is clear, evidence isaccumulating that their signaling pathways cross-communicate (Dicke and Van Poecke, 2002;Felton and Korth, 2000; Feys and Parker, 2000; Kunkel and Brooks, 2002; Pieterse and VanLoon, 1999; Reymond and Farmer, 1998; Rojo et al., 2003). For instance, activation of SA-dependent systemic acquired resistance (SAR) has been shown to suppress JA signaling inplants, thereby prioritizing SA-dependent resistance to microbial pathogens over JA-dependentdefense that is, in general, more effective against insect herbivory (Felton and Korth, 2000;Stout et al., 1999; Thaler et al., 2002b; Thaler et al., 1999). Pharmacological and geneticexperiments have indicated that SA-mediated suppression of JA-inducible gene expressionplays an important role in this process (Glazebrook et al., 2003; Peña-Cortés et al., 1993;Van Wees et al., 1999), and can sometimes work in both directions (Glazebrook et al., 2003;Niki et al., 1998). The antagonistic effect of SA on JA signaling was recently shown to becontrolled by a novel function of the defense regulatory protein NPR1 in the cytosol(Pieterse and Van Loon, 2004; Spoel et al., 2003). Cross-talk between defense signaling path-ways is thought to provide the plant with a powerful regulatory potential, which helps theplant to “decide” which defensive strategy to follow, depending on the type of attacker it is

Chapter 02: Signal signature and transcriptome changes of Arabidopsis during pathogen and insect attack

31

encountering. Yet, it may also allow attackers to manipulate plants to their own benefit byshutting down induced defense through influences on the signaling network (Kahl et al., 2000).

In order to study the role of pathway cross-talk in plant innate immunity, it is importantto have insight into the dynamics of SA-, JA-, and ET-signaling during different plant-attackercombinations. The role of SA, JA, and ET in plant defense has been studied for several plant-microbe and plant-insect interactions (Dicke and Van Poecke, 2002; Glazebrook, 2001; Pieterseet al., 2001). However, most of these studies have been performed in different plant species,often using single plant-microbe or plant-insect combinations. Moreover, the large variationin experimental conditions in these studies makes it difficult to integrate the results and drawoverall conclusions. Therefore, we monitored the dynamics of SA-, JA-, and ET-signaling in a single plant species (Arabidopsis thaliana) in response to attack by a range of microbialpathogens and herbivorous insects with very different modes of action. To relate our findingsto those by others, we investigated the response of Arabidopsis to the well-characterized microbialpathogens Pseudomonas syringae pv. tomato and Alternaria brassicicola and the herbivorousinsects Pieris rapae, Myzus persicae, and Frankliniella occidentalis. The production of SA,JA, and ET was monitored during these five Arabidopsis-attacker interactions, and related toglobal gene expression profiles using Affymetrix ATH1 whole-genome GeneChips.

ResultsArabidopsis pathogens and insects

Arabidopsis has been proven to be an excellent model for studying a wide variety of plant-pathogen and plant-insect interactions (Kunkel, 1996; Van Poecke and Dicke, 2004). To studythe dynamics of the response of Arabidopsis to different microbial pathogens and herbivorousinsects simultaneously, we chose two well-characterized Arabidopsis-pathogen interactions andthree Arabidopsis-insect interactions in which the attackers deploy very different modes of attack.

P. syringae is a bacterial leaf pathogen that causes extensive chlorosis and necrotic spots(Whalen et al., 1991).Analyses of Arabidopsis signaling mutants have shown that basal resistanceto this pathogen is predominantly dependent on SA (Delaney et al., 1994; Nawrath and Métraux,1999; Wildermuth et al., 2001), although components of the JA and ET signaling pathwayshave been demonstrated to contribute to resistance against this pathogen as well (Ellis et al.,2002; Pieterse et al., 1998). The transcriptome of Arabidopsis in response to P. syringae pv.maculicola infection has been well-studied (Glazebrook et al., 2003). Recently,Tao et al. (2003)provided evidence that a large part of the differences in transcriptional changes between thecompatible and incompatible interactions is quantitative.Therefore, to induce a strong responsein the plant, we chose to use avirulent P. syringae pv. tomato DC3000, carrying the avirulencegene avrRpt2. Pressure infiltration of whole Arabidopsis leaves with P. syringae pv. tomatoDC3000(avrRpt2) resulted in collapse of the leaf tissue within the first 48 h after inoculation,which is typical for this incompatible interaction (Figure 1).

32

FIGURE 1. Symptom development in Arabidopsis upon pathogen and insect attack

Symptom development on Arabidopsis leaves at different time points after inoculation/infestation with the necrotizingbacterial leaf pathogen P. syringae pv. tomato DC3000(avrRpt2), the necrotrophic fungal leaf pathogen A. brassicicola,tissue-chewing caterpillars of the cabbage white butterfly (P. rapae), cell-content feeding larvae of the Western flowerthrips (F. occidentalis), or phloem-sucking green peach aphids (M. persicae).

A. brassicicola is a necrotrophic fungal pathogen that provokes spreading necrotic lesionson leaves. In contrast to basal resistance against P. syringae, SA is not required for defenseagainst this pathogen, because Arabidopsis genotypes impaired in SA accumulation retain thestrong level of resistance that is characteristic for the wild-type Col-0 plants (Thomma et al.,1998; Van Wees et al., 2003). Basal resistance against A. brassicicola is compromised in thephytoalexin-deficient mutant pad3 and the JA-response mutant coi1, indicating that theArabidopsis phytoalexin camalexin and JA signaling are required for defense against A. brassici-cola (Thomma et al., 1998; Thomma et al., 1999). In our comparative study we used the pad3


33

mutant as the susceptible host for studying a compatible Arabidopsis-A. brassicicola inter-action. After inoculation with A. brassicicola, necrotic lesions developed gradually to a sizethat spanned half the width of the leaf 3 days after inoculation (Figure 1).

Tissue-chewing caterpillars of the cabbage white butterfly (P. rapae) are specialists oncruciferous plant species (Van Loon et al., 2000). Defense against caterpillar feeding in plantshas been suggested to be mainly regulated by JA-dependent defense responses (Kessler andBaldwin, 2002; Van Poecke and Dicke, 2002). In Arabidopsis, P. rapae feeding has been shownto induce expression of JA-responsive genes (Reymond et al., 2000, 2004) and to inducedirect and indirect defenses that involve SA, JA, and ET (Reymond et al., 2004; Stotz et al.,2002; Stotz et al., 2000; Van Poecke and Dicke, 2004; Van Poecke et al., 2001). Moreover,tomato plants affected in JA production or perception are more susceptible to caterpillar feedingthan wild-type plants (Howe et al., 1996; Thaler et al., 2002a). In this study, first-instar larvae of P. rapae immediately started to feed when they were placed onto the leaf tissue. Caterpillarfeeding caused a severe, progressing damage to the leaf tissues (Figure 1).

Western flower thrips (F. occidentalis) causes extensive damage on many plant species,including Arabidopsis (Yudin et al., 1986).Thrips are cell-content feeding insects that penetratesingle cells with a stylet to suck out the contents (Kindt et al., 2003). JA plays an importantrole in defense against cell content-feeding herbivores. Tomato mutant def1, compromised inJA-signaling, shows enhanced susceptibility to thrips feeding. Moreover, overexpression ofJA-inducible prosystemin, a signal peptide involved in the wound-induced expression of proteaseinhibitors (PIs), resulted in plants highly resistant to thrips damage (Li et al., 2002).Arabidopsis leaves infested with F. occidentalis displayed white chlorotic spots, so-called silverscars, which were located mainly at the leaf edges. During the course of the experiments, thesymptoms became more severe (Figure 1).

Green peach aphids (M. persicae) are generalists that feed on the plant’s phloem sapusing a sucking mode of action.The aphids carefully maneuver their stylets around the epider-mal and mesophyll cells before inserting them into the phloem, thereby inflicting minimalwounding to the plant (Tjallingii and Hogen Esch, 1993). M. persicae feeding has been shownto induce the expression of both SA- and JA-responsive genes (Moran and Thompson, 2001),suggesting a role for both signals in defense against aphid feeding. Ellis et al. (2002) demonstra-ted that M. persicae population development is reduced on Arabidopsis mutant cev1, whichconstitutively expresses JA-responsive genes. Moreover, aphid population development wasmuch faster on the JA-insensitive mutant coi1, indicating that JA plays an important role indefense against M. persicae (Ellis et al., 2002; Moran and Thompson, 2001). In our study,M. persicae was allowed to feed for 72 h. During this 72-h time course, the aphids fed pre-dominantly on the main vein at the abaxial side of the Arabidopsis leaves without causingany visible symptoms (Figure 1).

Signal signature

To investigate the dynamics of SA, JA, and ET production during the different Arabidopsis-attacker combinations, we monitored the production of these signals after pathogen and

34

insect attack. Because the progress of disease or damage caused by the pathogens and theinsects differed among the Arabidopsis-attacker combinations (Figure 1), the time points fortissue harvest were selected from early to late stages of infection/infestation and, thus, arenot always identical for each Arabidopsis-attacker combination. For SA and JA measurements,leaf tissue from 20 plants per plant-attacker combination and untreated controls were harvestedat each time point and immediately frozen in liquid nitrogen. For ET determinations, 10 plantsper plant-attacker combination were placed in gas-tight vials immediately after pathogeninoculation or insect infestation. Figure 2 shows the production of SA, JA, and ET during thefirst 72 h after pathogen or insect attack. P. syringae infection induced a strong increase inthe production of all three signal molecules. JA production was detectable as early as 3 hafter inoculation, whereas SA and ET levels were increased significantly from 12 h onwards.Similar to the Arabidopsis-P. syringae interaction, inoculation of Arabidopsis with A. brassici-cola resulted in a strong increase in JA and ET production. Enhanced JA levels weredetectable at 3 h after inoculation, whereas ET levels started to increase between 12 and 24 hpost inoculation. A. brassicicola did not induce an increase in SA levels.

None of the insects induced a detectable increase in SA accumulation (Figure 2).Moreover, the magnitude of JA and ET production was much lower in response to insectinfestation than during pathogen attack. However, this may be due to the fact that the numberof cells contributing to the defense response upon pathogen infection is higher than that uponinsect infestation. Feeding by tissue-chewing caterpillars of P. rapae induced a modest, butsignificant increase in ET production, and a clear increase in JA production. Cell content-feedinglarvae of the Western flower thrips F. occidentalis also induced an increase in JA biosynthesis,whereas ET levels remained unchanged. No changes in the production of JA or ET weredetectable in response to infestation of Arabidopsis with phloem-sucking M. persicae aphids.

Together these results demonstrate that the accumulation patterns of SA, JA, and ETdiffer highly in composition, magnitude, and timing during the different plant-pathogen andplant-insect combinations.The combined patterns of SA, JA, and ET production will subsequentlybe referred to as the signal signature.


35

FIGURE 2. Signal signature of Arabidopsis upon pathogen and insect attack

A. Endogenous levels of free SA in Arabidopsis plants at different time points after inoculation/infestation with P. syringae

pv. tomato DC3000(avrRpt2), A. brassicicola, P. rapae, F. occidentalis, or M. persicae. The values presented are means

(± SE) of five samples, each consisting of four rosettes that received the same treatment.

B. JA levels in Arabidopsis plants at different time points after pathogen inoculation or insect infestation. The values

presented are from 20 pooled rosettes that received the same treatment.

C, D. Cumulative ET production over a 72-h period in leaves of Arabidopsis after inoculation with P. syringae pv.

tomato DC3000(avrRpt2) or A. brassicicola (C), or after infestation with P. rapae, F. occidentalis, or M. persicae

(D). The represented values are means (± SE) for 10 plants that received the same treatment.

Inoculations with A. brassicicola were performed on the Col-0 mutant pad3-1, which is a susceptible host for this

pathogen. All other inoculations/infestations were carried out with Col-0 plants. Depending on the progress of the

symptoms inflicted by the respective pathogens and insects, harvesting of leaf tissue for SA and JA determinations

were omitted at some time points (missing of bars in A and B). FW, fresh weight.

36

Attacker-induced marker gene expression

To investigate in how far the specific patterns of defense signal production during each plant-attacker combination correspond with a coordinate activation of SA-, JA-, and/or ET-responsivegenes, we first analyzed the expression of the well-characterized marker genes PR-1 (SA-responsive), VSP2 (JA-responsive), PDF1.2 (JA- and ET-responsive), and HEL (ET-responsive).To be able to correlate the signal signatures with the gene expression patterns, RNA was iso-lated from the same leaf samples as those used for the SA and JA determinations. Figure 3shows that P. syringae pv. tomato DC3000(avrRpt2) induced the expression of all the SA-,JA-, and ET-responsive marker genes, whereas A. brassicicola only triggered the JA- and ET-responsive marker genes PDF1.2 and HEL. Furthermore, P. rapae and F. occidentalis inducedthe JA-responsive marker genesVSP2 and PDF1.2, respectively. No clear accumulation ofany marker gene transcripts could be detected in M. persicae-infested plants.

Because aphids damage only a small number of cells while probing for feeding sites, wemade use of the transgenic Arabidopsis Col-0 lines PDF1.2::GUS and PR-1::GUS to examinelocal aphid-induced marker gene expression in more detail.The PDF1.2::GUS and PR-1::GUSlines contain a translational fusion of the uidA reporter gene with the JA/ET-responsive pro-moter of the PDF1.2 gene, and the SA-responsive promoter of the PR-1 gene, respectively.No ß-glucuronidase (GUS) activity was detected in PDF1.2::GUS plants in response to M.persicae feeding. In contrast, aphid feeding strongly induced expression of the SA-responsivePR-1 promoter in the cells surrounding the feeding sites on the main vein (Figure 4).

To similarly investigate local effects of thrips and caterpillar feeding on PR-1 andPDF1.2 marker gene expression, GUS activity was also assessed in F. occidentalis- and P.rapae-infested PR-1::GUS and PDF1.2::GUS plants.Thrips feeding locally activated the PR-1promoter to a moderate level (Figure 4), which was apparently too low to be detected in theRNA isolated from whole rosettes (Figure 3). Damage caused by caterpillar feeding had noeffect on GUS activity in PR-1::GUS plants. Both F. occidentalis and P. rapae induced theexpression of the PDF1.2 promoter around the feeding site. The latter was not detected inthe RNA from whole rosettes of P. rapae-infested plants (Figure 3).

These results indicate that the expression patterns of the marker genes correlate only to a limited extent with the accumulation patterns of the signaling compounds themselves.For instance, JA production in P. syringae-infected plants was detectable earlier and to a 5-foldhigher level than in P. rapae-infested plants. Nevertheless, VSP2 transcript levels accumulatedfaster and to a higher level after caterpillar feeding. Furthermore, the timing and magnitudeof JA biosynthesis during P. rapae and F. occidentalis feeding was comparable. However, theexpression patterns of JA-responsive genes PDF1.2 and VSP2 were clearly different.


37

FIGURE 3. Northern blot analysis of SA-, JA-, and ET-responsive marker genes in Arabidopsis upon pathogen andinsect attack

Transcript levels of SA-responsive (PR-1), JA-responsive (VSP2 and PDF1.2), and ET-responsive (PDF1.2 and HEL)marker genes in Arabidopsis leaves at different time points after inoculation/infestation with P. syringae pv. tomatoDC3000(avrRpt2), A. brassicicola, P. rapae, F. occidentalis, or M. persicae. Equal loading of RNA samples waschecked using a probe for 18S rRNA.

FIGURE 4. Histochemical staining of ß-glucuronidase (GUS) activity in leaves of transgenic Arabidopsis PR-1::GUSand PDF1.2::GUS lines after insect feeding

Photographs were taken from representative leaves that were fed on for 24 h by P. rapae, or for 72 h by F. occidentalisor M. persicae. Silver scars inflicted by F. occidentalis feeding appear as a clear white zone at the edge of the leaf.

38

Global expression profiles of Arabidopsis upon pathogen andinsect attack

To explore the complexity of the transcriptional changes of Arabidopsis in response to pathogenor insect attack, we analyzed the transcriptome of Arabidopsis at two time points after pathogeninfection or insect infestation using Affymetrix ATH1 whole-genome GeneChips. Because adetailed qualitative analysis of the transcript profiles of each Arabidopsis-attacker combinationis beyond the scope of this study, we will focus on the comparison of the transcript profilesbetween the different Arabidopsis-attacker combinations.The time points used for the microarrayanalysis were selected on the basis of the signal signature (Figure 2) and the marker-geneexpression (Figure 3), and are listed in Table 1.To be able to relate gene expression to relativeSA, JA, and ET levels, RNA was prepared from the same plant material as was used for thedetermination of the signal signature (Figure 2). RNA was prepared from four biologicalreplicates, each consisting of 5 plants. These replicates were pooled to reduce noise arisingfrom biological variation. The transcript profile of each pool was obtained by hybridizationof an Affymetrix ATH1 GeneChip representing approximately 23,750 Arabidopsis genes(Redman et al., 2004). After hybridization, expressed genes were identified using GeneChipOperating Software (GCOS), which uses statistical criteria to generate a ‘present’ or ‘absent’call for genes represented by each probe set on the array. The average number of detectablegenes (with ‘present’ call) was 13,729 (60,2%), which is in good agreement with the 60%previously reported by Redman et al. (2004).

Expression values from each pooled sample were normalized globally using GCOS. Tovalidate the global normalization, the fold change in expression level of a set of nine genespreviously identified as representative, constitutively expressed controls (Kreps et al., 2002),was calculated. As expected, the fold-change ratio in attacker- over mock-treated leaves wasclose to 1 for most of these genes for all interactions and time points tested (Table 1).

To identify attacker-responsive genes, the transcript profile of each selected time pointof each Arabidopsis-attacker combination was compared to the transcript profile of theirrespective mock-treated control plants that were grown under identical conditions and wereharvested at the same two time points as the attacker-induced plants. To identify a robust set of pathogen- and insect-responsive genes, we chose an experimental set-up in which weselected for genes of which changes in expression level were evident during the whole time-frame monitored for each of the Arabidopsis-attacker combinations.The following conservativeselection criteria were applied. First, per Arabidopsis-attacker combination the expressionlevel had to be detectable (P-flag generated by GCOS) and the hybridization intensity had to be >40 units in at least two out of four data sets. Second, the change in expression level in attacker-treated leaves compared to that in mock-treated control leaves had to be at least2-fold. To avoid false positives we required the changes to occur at both time points and tobe in the same direction. Only those probe sets were selected that met these stringent selectioncriteria at both time points tested. The attacker-induced genes corresponding to the selectedprobe sets are listed in Supplements 1 to 5.


39

Validation of microarray data

To validate the GeneChip results we compared the relative expression values of the markergenes PR-1, PDF1.2, and HEL with the relative mRNA levels on the northern blots. VSP2was left out of this analysis because it is not represented on the ATH1 GeneChip.Hybridization signals on the northern blots were quantified using a Phosphor Imager and the fold-change relative to the respective controls calculated. Table 2 shows that out of 30combinations tested (3 marker genes x 5 Arabidopsis-attacker combinations x 2 time points)29 matched with the microarray data, indicating that the relative expression levels of themarker genes correlated well between GeneChip and northern blot hybridization. In addition,we determined the transcript levels of five attacker-specific genes (At1g30700, At4g26150,At4g15210, At1g72260, At5g62360) in each of the five Arabidopsis-attacker combinationsand their respective mock-treated controls, using quantitative real-time PCR (Q-RT-PCR).Figure 5 shows the fold-change induction of the selected genes in the different Arabidopsis-attacker combinations as determined by microarray analysis (left panel) and Q-RT-PCR (right


41

panel). Although fold induction in gene expression, especially for low abundant mRNAs,has been shown to differ between the two methods (Czechowski et al., 2004), the relativeexpression patterns of the five attacker-specific genes were highly similar, indicating that therelative expression levels of the genes tested correlated well between GeneChip and Q-RT-PCR analysis.

To further validate the GeneChip data obtained, we compared the selected pathogen-and insect-responsive genes with those identified in other transcript profiling studies in whichthe same or similar Arabidopsis-attacker combinations were used (Glazebrook et al., 2003;Moran et al., 2002; Reymond et al., 2004; Reymond et al., 2000; Tao et al., 2003; Van Weeset al., 2003; Verhagen et al., 2004). Although the experimental set-up, such as age of theplant material upon harvest, time points after inoculation and the type of microarray used,often differed in these studies, a large number of genes behaved similarly (data not shown).For instance, 65% of all the P. syringae-responsive genes identified in our study (Supplement 1)that are also represented on the Arabidopsis Genome 8K array of Affymetrix, were also identi-fied as being P. syringae-responsive by Tao et al. (2003). Moreover, 79% of all the A. brassi-cicola-responsive genes identified in our study that are also present on the ArabidopsisGenome 8K array, were also identified as being A. brassicicola-responsive by Van Wees et al.(2003).All together, these results indicate that our experimental set-up and stringent selectioncriteria resulted in the selection of a robust set of pathogen- and insect-responsive genes.

42

FIGURE 5. Comparison of microarray and Q-RT-PCR analysis of five attacker-specific genes in the differentArabidopsis-attacker combinations

Fold induction of five attacker-specific genes (At1g30700, At4g26150, At4g15210, At1g72260, and At5g62360) after infection/infestation of Arabidopsis by P. syringae, A. brassicicola, P. rapae, F. occidentalis, or M. persicae. On the left, the fold-change patterns from the microarray analysis. On the right, the fold-change patterns from the Q-RT-PCR analysis.


43

FIGURE 6. Functional analysis of differentially expressed gene sets

Distribution of the differentially expressed genes identified in the Arabidopsis-attacker combinations over the functional categories. The number of up- or down-regulated genes is given in the center of the respective pies.Classification in functional categories was performed essentially according to the Gene Ontology tool of TAIR.Genes belonging to the functional category ‘response to abiotic and biotic factors’ and ‘response to stress’ weregrouped in a single functional category designated ‘response to abiotic and biotic stress’.

44

FIGURE 7. Functional analysis of differentially expressed gene sets

Degree of overrepresentation of the differentially expressed genes in the functional categories. The distribution of the differentially expressed genes over the functional categories is presented relative to the distribution of all geneson the Affymetrix ATH1 array (set at 100% for each functional category).


45

Functional analysis of differentially expressed genes

All differentially expressed genes identified in the five Arabidopsis-attacker combinationswere classified according to their functional categories derived from the Gene Ontology toolat The Arabidopsis Information Resource (TAIR) (http://www.Arabidopsis.org) (Rhee et al.,2003). The distribution of the identified probe sets over the different functional categories isshown in Figure 6. To evaluate the importance of a given functional category, the percentageof differentially expressed genes belonging to each functional category was compared to thedegree of representation of the respective functional category in the genome. Figure 7 showsthe results of this comparison for the up- and down-regulated genes. The predominant func-tional category that is overrepresented in the up-regulated gene sets of four out of fiveArabidopsis-attacker combinations represent genes involved in the response to abiotic andbiotic stress. In the Arabidopsis-M. persicae interaction, genes from this category are over-represented as well, although the predominant overrepresented category represents genesinvolved in so-far unspecified biological processes (“other biological processes”).

Of the differentially expressed genes that are down-regulated during the Arabidopsis-A. brassicicola interaction, genes involved in the response to abiotic and biotic stress are clearlyoverrepresented. This indicates that, besides differential activation, also repression of stress-related genes occurs during the response of Arabidopsis to this pathogen. In the Arabidopsis-P. rapae and Arabidopsis-F. occidentalis interactions, genes involved in so-far unspecifiedbiological processes (“other biological processes”) are clearly overrepresented in the down-regulated gene sets. However, the specific biological gene functions are diverse, impeding anyspeculation as to their biological relevance. In the interactions of Arabidopsis with P. syringaeand M. persicae none of the functional categories are clearly overrepresented among thedown-regulated genes.

Comparison of transcriptome changes induced by pathogenand insect attack

Table 3 shows the number of genes that are consistently up- or down-regulated in the differentArabidopsis-attacker combinations (complete lists in Supplements 1 to 5). Of all the attackersinvestigated, M. persicae induced the largest number of changes (2181). This is remarkablebecause aphid feeding caused virtually no visual symptoms in comparison to the extensivedamage caused by the other attackers. P. syringae infection resulted in a similar number ofconsistent changes (2034), whereas the number of consistent changes in the other Arabidopsis-attacker combinations was much lower (151 to 199). It must be noted that in all Arabidopsis-attacker combinations, many more genes showed a more than 2-fold change in expression ata single point in time. Because these changes are not as robust as the consistent changes, theywere not analyzed further.

To evaluate the complexity of the transcriptional changes induced during the five differentArabidopsis-attacker combinations, we made a pair-wise comparison of the overlap betweenthe selected probe sets. Table 3 shows that in the majority of the comparisons the overlap is

46

relatively small, indicating that most of the differentially expressed genes are specific for therespective Arabidopsis-attacker combinations. However, more than 50% of all consistent changeselicited by A. brassicicola (68%), P. rapae (52%), and F. occidentalis (72%), are also consis-tently triggered by P. syringae, suggesting that these genes are commonly activated/repressedduring these Arabidopsis-attacker interactions. Interestingly, these four attackers all induced a considerable increase in JA levels (Figure 2), suggesting that JA may be the common regulatorof the overlapping gene sets.

To investigate the role of JA in the regulation of the overlapping gene sets, we identifiedprobe sets representing JA-responsive genes among the selected attacker-responsive genes. Tothis end, 5-week-old Col-0 plants were treated with 0.05 mM MeJA and harvested 0, 1, 3, and6 h later. RNA from these plants was used to prepare probes for the hybridization of AffymetrixATH1 GeneChips. Probe sets showing a >2-fold change (up or down) on at least two of thetime points tested were selected as described above. The resulting 2,209 probe sets were con-sidered to represent JA-responsive genes (Supplement 6). Comparison of these JA-responsivegenes among the selected attacker-responsive probe sets revealed that 32% of the P. syringae-responsive genes are responsive to MeJA (Table 4). The percentages of JA-responsive genesamong the A. brassicicola-, P. rapae-, and F. occidentalis-induced changes were even higher(44%, 55%, and 69%, respectively), indicating that JA plays a dominant role in the transcrip-tional reprogramming of Arabidopsis in response to these attackers. Pair-wise comparisons ofthe overlap between JA-responsive genes in the four Arabidopsis-attacker combinations, revealedthat of all JA-responsive, P. rapae-induced changes, 66% is also induced by P. syringae (Table 4).In the Arabidopsis-F. occidentalis and the Arabidopsis-A. brassicicola interactions, this percen-tage is even higher (80% and 85%, respectively), indicating that the JA-induced defenseresponses triggered by these attackers show considerable overlap. However, this does not holdfor all Arabidopsis-attacker combinations. For instance, when the JA-responsive genes amongthe P. rapae- and F. occidentalis-induced changes were compared with the JA-responsive genesamong the A. brassicicola-induced ones, the overlap was relatively low (6 to 17%). Theseresults indicate that although attackers with very different modes of action (e.g. F. occidentalisand P. syringae) may induce similar sets of JA-responsive genes, the majority of the JA-responsivegenes are affected in an attacker-specific manner, indicating that other factors besides JA shapethe final outcome of the defense response.


47

DiscussionPlants require a broad range of defense mechanisms to effectively combat invasion by microbialpathogens or attack by herbivorous insects. These mechanisms include pre-existing physicaland chemical barriers, as well as inducible defense responses that become activated uponpathogen infection or insect herbivory. A concerted action of these defensive activities helpsthe plant to minimize damage caused by the attacker. The signal molecules SA, JA, and EThave been implicated in many plant-pathogen and plant-insect interactions (Dicke and Hilker,2003; Pieterse and Van Loon, 1999). Despite the evident overlap in signaling that is triggeredupon pathogen or insect attack, the plant response is highly dependent on the plant-attackercombination. Little is known about how plants co-ordinate attacker-induced signals into specificdefense responses. A well-accepted hypothesis is that modulation of the different defense signaling pathways involved plays an important role in this process (Reymond and Farmer,1998). Although ample information is available on the role of SA, JA, and ET in the responseof plants to certain pathogens and insects, the information is often highly specific for a givenplant-pathogen or plant-insect interaction. Moreover, the different studies are often character-ized by unique experimental conditions. Here, we attempted to gain insight into the dynamicsof the response of a single plant species (Arabidopsis thaliana) to a variety of microbial andherbivorous attackers under identical conditions. This approach allowed us to compare thedynamics of signal production and the transcriptional reprogramming of Arabidopsis uponattack by pathogens and insects with very different modes of attack.

Correlation between signal signature and marker gene expression

Gene expression profiles and SA, JA, and ET production were examined simultaneously duringthe entire period between inoculation/infestation and the occurrence of the resulting severesymptoms or damage (Figure 1). Because aphids did not cause any visible symptoms, theresponse of Arabidopsis to this attacker was monitored over a 72-h time course. All otherattackers caused a significant increase in the production of one or more of the signals tested(Figure 2).The accumulation patterns of SA, JA, and ET during the different Arabidopsis-attackerinteractions clearly differed in composition, magnitude, and timing.This so-called signal signa-ture was reflected in the expression patterns of the well-characterized marker genes PR-1,VSP2, PDF1.2, and HEL (Figure 3). For instance, P. syringae infection caused a considerableincrease in SA, JA, and ET production, and was associated with the subsequent activation ofall the SA-, JA-, and ET-responsive marker genes tested. Furthermore, A. brassicicola infectioncaused a significant increase in both JA and ET levels, resulting in the activation of the JA-and ET-responsive marker genes PDF1.2 and HEL. However, in some Arabidopsis-attackercombinations the signal signature correlated only to a limited extent with the expression patterns of the marker genes. The high levels of JA produced by Arabidopsis in response toinfection by A. brassicicola resulted in the activation of the JA-responsive gene PDF1.2, butnot in that of the JA-responsive gene VSP2. Moreover, although P. rapae and F. occidentalisinduced comparable levels of JA in Arabidopsis, VSP2 was activated in the Arabidopsis-P. rapae


49

interaction, whereas PDF1.2 was not. Conversely, F. occidentalis triggered the expression ofPDF1.2 but not that of VSP2. Hence, it must be concluded that the signal signature of a givenplant-attacker combination plays a primary role in the orchestration of the plant’s defenseresponse, but additional layers of regulation lead to differential marker gene expression.

Attacker-induced transcriptional changes

The goal of the microarray analysis was to explore the complexity of the transcriptionalreprogramming initiated by the different pathogens and insects in relation to the observedArabidopsis-attacker signal signatures, and to identify robust sets of attacker-responsivegenes. To this end, we applied stringent selection criteria to identify genes that show a con-sistent change in expression during pathogenesis and herbivore feeding. Depending on theArabidopsis-attacker combination, 151 to 2181 genes showed a consistent change in expres-sion over time. Surprisingly, aphid feeding triggered the largest number of consistent changesin gene expression, even though these insects caused the least symptoms of all attackers testedand did not induce detectable changes in SA, JA, and ET levels (Figures 1 and 2). In contrastto the other four Arabidopsis-attacker combinations, a large proportion of the differentiallyexpressed genes in the Arabidopsis-aphid interaction was down-regulated (62% versus 14-36%in the other combinations).A relatively large fraction of the down-regulated genes is involvedin plant metabolism, confirming previous findings that demonstrate that aphids are majormanipulators of plant physiology and nutrition status (Davies et al., 2004). Previously, Moranand co-workers (Moran et al., 2002; Moran and Thompson, 2001) identified 19 M. persicae-responsive genes in Arabidopsis by northern blot and small-scale microarray analysis.Thirteen of these genes (68%) were among the 2181 identified as being consistently responsiveto M. persicae in our GeneChip analysis, including the SA-responsive genes PR-1 (At2g14160)and PR-2 (ß-1,3-glucanase; At3g57260).Although PR-1 transcript levels were barely detectableon the northern blots (Figure 3), they were clearly expressed in the cells surrounding thefeeding sites on the main veins of the PR-1::GUS reporter line (Figure 4).These results indicatethat significant local changes in gene expression can be identified by microarray analysis, whileescaping from identification by northern blot analysis.

A large proportion of the gene sets identified in our study as being attacker-responsive,has also been identified in comparable studies (Glazebrook et al., 2003; Reymond et al., 2000;Tao et al., 2003; Van Wees et al., 2003; Verhagen et al., 2004). For instance, Reymond et al.(2000) identified 17 genes showing a >2-fold increase in expression level in response to P. rapaefeeding using a small dedicated microarray with probes for 150 Arabidopsis genes. Of thegenes also represented on the ATH1 chip, 59% showed a consistent >2-fold increase in ourArabidopsis-P. rapae data sets, even though different time points after infestation (3 h in thestudy by Reymond et al. versus 12 and 24 h in our study) and different larval stages (L4/L5in the study of Reymond et al. versus L1/L2 in our study) were tested. Furthermore, 65% of the P. syringae-responsive genes that were identified in our study (and were present onboth the ATH1 and the Affymetrix 8k array), were also identified by Tao et al. (2003).Similarly, 79% of the A. brassicicola-responsive genes were also identified by Van Wees et al.

50

(2003), who also used the susceptible phytoalexin-deficient mutant pad3 to study theArabidopsis-A. brassicicola interaction. Together, these data indicate that the gene sets thatwere selected in this study are to a large extent representative for the different Arabidopsis-attacker combinations used. It must, however, be noted that to achieve a maximal responseof Arabidopsis to P. syringae infection, we made use of an avirulent strain of the pathogen.Although it has been suggested that the difference in the transcriptional response of Arabidopsisto virulent and avirulent strains of P. syringae is predominantly quantitative (Tao et al.,2003), it can not be excluded that a small proportion of the selected genes are specific forthe incompatible interaction.

Genes showing a >2-fold change at a single time point are either part of a transientresponse or false positives and, thus, are unlikely to be identified consistently when bioassaysare performed under different experimental conditions. Although some of these genes mayplay an important role in the response of Arabidopsis to the attacker involved, the scope ofthis study was not to provide a qualitative in depth analysis of individual gene sets that aredifferentially expressed in the different Arabidopsis-attacker combinations, but to explore thecomplexity of the transcriptional changes in the response of Arabidopsis to attack by differentpathogens and insects. Therefore, we limited our analysis to those genes that showed a robustchange in expression and disregarded all others. The selected robust gene sets obtained withthe whole-genome ATH1 arrays can be related to actual SA, JA, and ET levels and will be ofvalue for more detailed analyses of individual Arabidopsis-attacker interactions.The data setsof all ATH1 array hybridisations used in this study are available via the Nottingham ArabidopsisStock Centre International Affymetrix service (NASCarrays): affymetrix.Arabidopsis.info.

Stress-related genes are overrepresented in all Arabidopsis-attacker combinations

To gain insight into the function of the differentially expressed genes, we categorized theirbiological function essentially according to the Gene Ontology tool of TAIR. Some of thesefunctional categories cover a relatively large proportion of the Arabidopsis genome, e.g. genes in the functional category ‘metabolism’ represent 21.7% of all annotated genes, while genesin the category ‘response to abiotic and biotic stress’ represent only 5.6% of the genome.Thus, information on the percentage of selected genes in a given functional category is biasedby the degree of representation of this category in the genome.To identify functional categoriesin which a relatively large proportion of the genes show a consistent change in expression in response to pathogen or insect attack, we compared the number of identified genes in agiven functional category with the degree of representation of this category in the wholegenome. In this way, functional categories that are overrepresented in the selected differentiallyexpressed genes sets were readily identified (Figure 7). In all Arabidopsis-attacker combinationstested, the number of up-regulated genes predicted to be involved in the response to bioticand abiotic stress was 2- to 4-fold higher than expected on the basis of representation of thiscategory in the genome. Evidently, differential expression of a large proportion of genes fromthis category plays an important role in the response of Arabidopsis to pathogen and insect


51

attack. However, when looking at the absolute percentages of representation of the genes inthe different functional categories (Figure 6), the contribution of stress-related genes in theinvestigated interactions is not immediately clear. For instance, of all consistently up-regulatedgenes in the different Arabidopsis-attacker combinations 10.6 to 21.7% belongs to the functionalcategory ‘response to abiotic and biotic stress’, while a considerably larger proportion of thegenes (20.8 to 33.8%) fall into the functional category ‘metabolism’ (Figure 5A).Thus, assess-ment of the distribution of the identified gene sets over the different functional classes as afunction of the degree of representation of these functional categories in the genome, makesit is possible to better weigh the importance of a given functional category in the plantresponse studied.

Complexity of transcriptional reprogramming upon pathogenand insect attack

To explore the complexity of transcriptional changes induced by the different Arabidopsisattackers used, we compared the overlap between gene sets. Because both P. syringae and M. persicae induced, by far, the largest number of consistent changes (10- to 14-fold moregenes than A. brassicicola, P. rapae, and F. occidentalis), it is evident that the transcriptionalresponse of Arabidopsis to these very different attackers is highly complex. In the case of P. syringae, this may be related to the fact that infection of Arabidopsis by this pathogenresults in the production of high levels of SA, JA, and ET, each of which may activate differentsets of genes. In the case of M. persicae feeding, however, none of these signals tested wasdetectable. Evidently, the onset of the large transcriptional reprogramming elicited by thesephloem-feeding insects is not based on the production of high overall levels of SA, JA, or ET,suggesting that the responses of Arabidopsis to P. syringae and M. persicae is highly unrelated.Indeed, most of the transcriptional changes induced by P. syringae or M. persicae were unique.Nonetheless, 253 genes (141 up-regulated genes and 112 down-regulated genes; data notshown) of all consistently induced changes in the Arabidopsis-P. syringae and the Arabidopsis-M. persicae interaction overlapped. Thus, although both attackers have very different modesof action and trigger a highly dissimilar signal signature, a large number of Arabidopsis genesare recruited in response to both attackers. However, these overlapping genes only represent12% of the total number of consistent changes identified in both interactions and, thus, mayonly contribute to a limited extent to the overall defense reaction.

Compared to P. syringae and M. persicae, A. brassicicola, P. rapae, and F. occidentalisinduced only a relatively low number of consistent changes in gene expression (151 to 199up- or down-regulated genes). A small number of these genes (6) showed a consistent changein all three Arabidopsis-attacker combinations (data not shown). Pair-wise comparison of thedifferentially expressed gene sets revealed an overlap of 4% (P. rapae versus A. brassicicola),17% (F. occidentalis versus A. brassicicola), and 39% (P. rapae versus F. occidentalis). In thesethree Arabidopsis-attacker interactions, JA is a dominant component of the signal signatureproduced. Indeed, 44 to 69% of all differentially expressed genes identified in these threeArabidopsis-attacker combinations were also found to be responsive to exogenous application

52

of MeJA (Table 4), indicating that JA-responsive gene expression plays a central role in theresponse of Arabidopsis to infection/infestation by all three attackers. However, the majority(94 to 46%) of these MeJA-responsive genes showed an attacker-specific expression patternin pair-wise comparisons between the differentially expressed gene sets. This may be partlyexplained by differences in sampling time points, but on all time points tested JA levels wereclearly elevated. Hence, the sets of JA-responsive genes that are differentially activated orrepressed in the different Arabidopsis-attacker combinations are highly divergent, suggestingthat so far unidentified regulatory processes play an important role in modulating the finaloutcome of the defense response. Figure 8 shows a model of how invasion by JA-inducingattackers may result in the activation of differential sets of JA-responsive genes. Similar modelscan be drawn for genes that are regulated by other defense-related signals such as SA andET, resulting in a network of interconnecting signaling pathways that provides the plant with a powerful regulatory potential to fine-tune it’s defense response.

In conclusion, we demonstrated that Arabidopsis is highly adapted in its response topathogens and herbivorous insects with very different modes of attack. Depending on theArabidopsis-attacker combination, the signal molecules SA, JA, and ET are produced withlarge differences in both quantity and timing.We identified differentially expressed gene setsthat over time show a consistent change in expression for each of the Arabidopsis-attackercombinations. In all cases, stress-related genes are clearly overrepresented in the gene setsidentified. In four of the five Arabidopsis-attacker combinations tested, JA plays an importantrole in the differential regulation of a large proportion of the activated/repressed genes.Nevertheless, the vast majority of the JA-responsive changes are specific for each plant-attackercombination. Evidently, signal molecules such as JA play an important role in the primaryresponse of the plant to pathogen and insect attack. However, additional layers of regulationobviously shape the outcome of the defense reaction. Pathway cross-talk or effects of so farunidentified regulatory factors may play an important role in the fine-tuning of the plant’sresponse to pathogens and insects. The nature and importance of these regulatory processeswill be a challenging topic for future research.


53

FIGURE 8. Differential expression of JA-responsive genes upon attack by JA-inducing pathogens and insects

Attack of Arabidopsis by P. syringae, A. brassicicola, P. rapae, or F. occidentalis, results in a strong increase in theproduction of JA, and a concomitant change in the expression of a large number of JA-responsive genes (numbersare given between parentheses). Nevertheless, the overlap among the JA-responsive genes between the differentArabidopsis-attacker combinations is relatively low (number of overlapping genes between the indicated Arabidopsis-attacker combinations are given in the Venn diagrams). SA and ET have been demonstrated to cross-communicatewith the JA pathway. Hence, depending on the amount and timing of their production, SA and ET may have positiveor negative effects on the expression of specific sets of JA-responsive genes. In addition, so far unidentified plant- orattacker-derived signals, or physiological conditions that are inflicted by the attacker, may be involved in modulatingJA-responsive gene expression.

54

Materials and methodsCultivation of plants

Seeds of Arabidopsis thaliana accession Col-0 and the phytoalexin-deficient Col-0 mutantpad3-1 (Glazebrook and Ausubel, 1994) were sown in quartz sand. Two-week-old seedlingswere transferred to 60-mL pots containing a sand/potting soil mixture that was autoclavedtwice for 20 min. Plants were cultivated in a growth chamber with a 8-h day (200 µE.m-2.s-1

at 24°C) and 16-h night (20°C) cycle at 70% relative humidity for another 3 weeks. Plantswere watered every other day and received half-strength Hoagland nutrient solution (Hoaglandand Arnon, 1938) containing 10 µM Sequestreen (CIBA-Geigy, Basel, Switzerland) once a week.

Pathogen bioassays

Inoculations with the bacterial leaf pathogen Pseudomonas syringae pv. tomato DC3000 was performed as described previously (Van Wees et al., 1999). Briefly, P. syringae pv. tomatoDC3000 with the plasmid pV288 carrying avirulence gene avrRpt2 (Kunkel et al., 1993) wascultured overnight at 28°C in liquid King’s medium B (King et al., 1954), supplemented with25 mg.L-1 kanamycin to select for the plasmid. Subsequently, bacterial cells were collected bycentrifugation and resuspended in 10 mM MgSO4 to a final density of 107 cfu.mL-1.Wild-typeCol-0 plants were inoculated by pressure infiltrating a suspension of P. syringae pv. tomatoDC3000(avrRpt2) at 107 cfu.mL-1 into all fully expanded leaves of 5-week-old plants.

Bioassays with the fungal leaf pathogen Alternaria brassicicola strain MUCL 20297 werecarried out as described by Ton et al. (2002). Briefly, A. brassicicola was grown on potatodextrose agar plates for 2 weeks at 22°C. Subsequently, conidia were collected as describedby Broekaert et al. (1990). Five-week-old susceptible pad3-1 plants were challenge inoculatedby applying 3-µL drops of 10 mM MgSO4 containing 10 6 spores per mL onto all fully expandedleaves of 5-week-old plants.

Insect bioassays

Tissue-chewing larvae of the small cabbage white butterfly Pieris rapae were reared on Brusselssprout plants (Brassica oleracea gemmifera cv. Cyrus) in a growth chamber with a 16-h dayand 8-h night cycle (21°C; 50–70% relative humidity), as described previously (Van Poeckeet al., 2001). Infestation of Arabidopsis Col-0 plants was carried out by transferring five first-instar larvae of P. rapae to each plant using a fine paintbrush.

The population of the Western flower thrips Frankliniella occidentalis originated froma greenhouse infestation on chrysanthemum.This virus-free population was reared on Phaseolusvulgaris cv. Prelude pods, supplied with Pinus pollen, in glass jars that were placed at 25°Cin a growth chamber with a 16-h day and 8-h night cycle as described (Kindt et al., 2003).Thrips infestations were performed by transferring 20 larvae of F. occidentalis to each Arabidopsis


55

Col-0 plant.Phloem-feeding green peach aphids (Myzus persicae) were maintained on Brassica chinensis

L. cv. Granaat under greenhouse conditions (25°C; 50-70% relative humidity). The 16-h lightperiod prevented sexual reproduction, keeping the population clonal.Arabidopsis Col-0 plantswere infested with M. persicae by transferring 40 nymphs and apterous adults to each plant(Van Poecke et al., 2003).

All insect populations used consisted of fairly immobile stages such that individualsremained on the plants to which they were transferred.

MeJA treatment

Induction treatment with methyl jasmonate (MeJA) was performed by dipping 5-week-oldCol-0 plants in an aqueous solution containing 0.05 mM MeJA (Serva, Brunschwig Chemie,Amsterdam, the Netherlands) and 0.01% of the surfactant Silwet L-77 (Van Meeuwen ChemicalsB.V., Weesp, the Netherlands) as described previously (Pieterse et al., 1998). Plants were harvested at 0, 1, 3, and 6 h after induction treatment and immediately frozen in liquid nitrogen.

ET quantification

Immediately after pathogen inoculation or transfer of insect populations to the shoots, rosetteswere detached from the roots, weighed and placed individually in 35-mL gas-tight serumflasks (n=10) that were subsequently incubated under climate chamber conditions.At differenttime intervals, 1-mL gas samples were withdrawn through the rubber seal. The concentrationof ET was measured by gas chromatography as described by De Laat and Van Loon (1982).

JA and SA quantification

All leaves from 20 plants per treatment were frozen in liquid nitrogen and pulverized withmortar and pestle. For each JA extraction, a sample of 1 g was taken from the frozen leafmaterial and transferred to a 50-mL centrifuge tube. To the frozen samples were added 100ng of the internal standard 9,10-dihydrojasmonic acid, 10 mL of saturated NaCl solution,0.5 mL of 1 M citric acid, and 25 mL of diethylether containing 0.005% (w/v) butylatedhydroxytoluene as antioxidant. Subsequently, extraction and GC-MS quantification of JAwas carried out as described by Mueller and Brodschelm (1994).

For each SA extraction, a sample of 0.5 g of ground leaf tissue was transferred to a 1.5-mLmicrofuge tube and 100 µL of the internal standard ortho-anisic acid (1 µg.mL-1) and 0.5 mLof 70% ethanol were added. Subsequently, extraction and quantification of SA were carriedout as described by Meuwly and Métraux (1993).

56

Northern blot analysis

Total RNA was extracted as described previously (Van Wees et al., 1999). For northern blotanalysis, 15 µg RNA was denatured using glyoxal and DMSO (Sambrook et al., 1989), electro-phoretically separated on a 1.5% agarose gel, and blotted onto Hybond-N+ membranes(Amersham, `s-Hertogenbosch, the Netherlands) by capillary transfer.The electrophoresis andblotting buffer consisted of 10 and 25 mM sodium phosphate (pH 7.0), respectively.Northern blots were hybridized with gene-specific probes for PR-1, PDF1.2, VSP2, and HELas described previously (Pieterse et al., 1998).To check for equal loading, the blots were strippedand hybridized with a probe for 18S rRNA. The AGI numbers for the genes studied areAt2g14610 (PR-1), At5g24770 (VSP2), At5g44420 (PDF1.2), and At3g04720 (HEL). Probefor 18S was derived from an Arabidopsis cDNA clone (Pruitt and Meyerowitz, 1986).

Quantitative real-time PCR

Q-RT-PCR analysis was basically performed as described previously (Czechowski et al., 2004).RNA (2 µg) was digested with Turbo DNA-freeTM (Ambion, Huntingdon, United Kingdom)according to the manufacturer’s instructions. To check for genomic DNA contamination, aPCR with primers designed on intron sequences of ACT7 (At5g09810; ACT7-FOR; 5’-GACATG GAA AAG ATA TGG CAT CAC AC-3’; ACT7-REV; 5’-AGA TCC TTC CTG ATA TCGACA TCA C-3’) was carried out. Subsequently, DNA-free total RNA was converted into cDNAusing oligo-dT20 primers (Invitrogen, Breda, the Netherlands), 10 mM dNTPs, and SuperScriptTM

III Reverse Transcriptase (Invitrogen, Breda, the Netherlands) according to the manufacturer’sinstructions. Efficiency of cDNA synthesis was assessed by Q-RT-PCR using primers of theconstitutively expressed gene UBI10 (At4g05320; UBI10-FOR; 5’ AAA GAG ATA ACAGGA ACG GAA ACA TAG T-3’; UBI10-REV; 5’-GGC CTT GTA TAA TCC CTG ATG AATAAG-3’). Gene-specific primers were designed for five Arabidopsis genes, each of whichshowed an attacker-specific expression pattern in one of the five Arabidopsis-attacker interac-tions studied. The corresponding AGI numbers and primers are At1g30700, FOR 5’- TCCGTA ACC TCC GCT TCA AC-3’, REV 5’-CGT GGC CTC CAC TTC TGA TT-3’ (Arabidopsis-P. syringae); At4g26150; FOR 5’ GGA TTT GGA GAC CCAGAG CA-3’, REV 5’-TGG CAGCCT CCT TCT CAT CT-3’ (Arabidopsis-A. brassicicola); At4g15210, FOR 5’-GAC GGC CTACAA AAC GCT GT-3’, REV 5’-CCA TTG TGG GAT CGG GAT AG-3’ (Arabidopsis-P. rapae);At1g72260, FOR 5’-CTG CCC TTC CAA CCA AGC TA-3’, REV 5’-TGG CAT CCA CTCACT TGC AT-3’ (Arabidopsis-F. occidentalis); and At5g62360, FOR 5’-CAA ACA AGC CCCAAG CTC AT-3’, REV 5’- CGC ACC ATC ATT GCT GAA GT-3’ (Arabidopsis-M. persicae).Q-RT-PCR analysis was done in optical 96-well plates with an MyIQTM Single Color Real-TimePCR Detection System (Bio-Rad,Veenendaal, the Netherlands), using SYBR® Green to monitordsDNA synthesis. Each reaction contained 1 µL of cDNA, 0.5 µL of each of the two gene-specific primers (10 pmol/µL), and 10 µL of 2x IQ SYBR® Green Supermix reagent (Bio-Rad,Veenendaal, the Netherlands) in a final volume of 20 µl. The following PCR program wasused for all PCR reactions: 95 °C for 3 min; 40 cycles of 95 °C for 30 sec, 59.5 °C for 30 sec,


57

and 72 °C for 30 sec.Threshold cycle (CT) values were calculated using Optical System Software,version 1.0 for MyIQTM (Bio-Rad,Veenendaal, the Netherlands). Subsequently, CT values werenormalized for differences in dsDNA synthesis using the UBI10 CT values. Normalized transcriptlevels of the five genes in each of the five Arabidopsis-attacker combinations were comparedto those of the respective mock-treated controls and the fold change in expression level wascalculated.

GUS assays

Transgenic Arabidopsis PDF1.2::GUS and PR-1::GUS lines, containing a translational fusionof the PDF1.2 or the PR-1 promoter with the uidA reporter gene in the Col-0 background(kindly provided by Yulia Plotnikova, Massachusetts General Hospital, Boston, USA), weregrown in soil as described above. Insects were transferred to 5-week-old plants as describedabove. After 24 h of caterpillar feeding or 72 h of thrips or aphid feeding, leaf tissues wereharvested and GUS activity was assessed by transferring the seedlings to GUS staining solu-tion (1 mM X-Gluc, 100 mM NaPi buffer, pH 7.0, 10 mM EDTA, and 0.1% (v/v) Triton X-100)as described previously (Spoel et al., 2003).After overnight incubation at 37°C, the leaf tissueswere destained by repeated washes in 70% ethanol and evaluated for staining intensity.

Sample preparation and microarray data collection

For isolation of RNA, whole rosettes were harvested at different time intervals during eachArabidopsis-attacker interaction or at several time points after MeJA treatment, and imme-diately frozen in liquid nitrogen. For all time points, every Arabidopsis-attacker combination,and the MeJA treatment, appropriate mock-treated plants were harvested. RNA was preparedfrom four biological replicates, each consisting of 5 plants, as described above and cleanedusing RNeasy Plant Mini Kit columns (Qiagen Benelux BV,Venlo, the Netherlands). Thesereplicates were pooled to reduce noise arising from biological variation. In retrospect it isnow recognized that pooling RNA samples of biological replicates is not optimal. If theexperiments would have been done today each biological replicate would have been used for hybridization of a GeneChip. Synthesis of cRNA probes, hybridization to GeneChips,and collection of data from the hybridized GeneChips were performed as described previously(Verhagen et al., 2004; Zhu et al., 2001). Hybridizations with labeled cRNAs were conductedwith Arabidopsis ATH1 full-genome GeneChips (Affymetrix, Santa Clara, USA), containing a total number of 22,810 probe sets representing approximately 23,750 Arabidopsis genes(Redman et al., 2004). On this GeneChip, each gene is represented by at least one ‘probe set’consisting of 11 25-mer oligonucleotides. Probe preparations and GeneChip hybridizationswere carried out by ServiceXS (Leiden, the Netherlands) and the Affymetrix service stationof Leiden Univerisity Medical Centre (LUMC) where they passed all internal quality checks.

58

Expression profiling

GeneChip Operating Software (GCOS) (Affymetrix, Santa Clara, USA) was used to globallynormalize the expression data on each GeneChip to an average value of 200 so that hybridizationintensity of all chips was equivalent. In addition, expressed genes were identified by GCOS,which uses statistical criteria to generate a ‘present’ or ‘absent’ call for genes represented byeach probe set on the array. Microarray data files were then analyzed using GeneSpring 6.1(Silicon Genetics, Redwood, CA, USA). The default settings ‘Per Chip: Normalize to 50th per-centile’ and ‘Per Gene: Normalize to specific samples’ were used during the data analyses.The P-values from the Pearson correlation tests run for GeneChips that were hybridized withprobes from four biological replicates of non-treated control plants ranged between 0.92 and0.97. This is in good agreement with the high correlation coefficients previously reported forindependent biological samples (Redman et al., 2004), indicating that the GeneChip hybridi-zations and microarray data collections were performed in a technically sound manner.

AcknowledgementsThe authors acknowledge Dr. Bas van Breukelen for help with the statistical analysis of themicroarray data and Leo Koopman, Frans van Aggelen and André Gidding for insect rearing.This research was supported, in part, by grant 811.36.004 of the Earth and Life SciencesFoundation (ALW), which is subsidised by the Netherlands Organisation of Scientific Research(NWO), and grant QLK5-CT-2001-52136 of the Marie Curie Individual Fellowship programmeof the European Commission.


59

Supplementary MaterialsSupplement 1: MS Excel file with normalized expression levels, fold-change information,AGI numbers and TIGR annotation of the selected probe sets that showed a consistent >2-foldchange (up or down) 12 and 24 h after inoculation of Arabidopsis Col-0 plants with the bac-terial leaf pathogen P. syringae pv. tomato DC3000(avrRpt2).

Supplement 2: MS Excel file with normalized expression levels, fold-change information,AGI numbers and TIGR annotation of the selected probe sets that showed a consistent >2-foldchange (up or down) 24 and 48 h after inoculation of Arabidopsis pad3-1 plants with thefungal leaf pathogen A. brassicicola.

Supplement 3: MS Excel file with normalized expression levels, fold-change information,AGI numbers and TIGR annotation of the selected probe sets that showed a consistent >2-foldchange (up or down) 12 and 24 h after infestation of Arabidopsis Col-0 plants with tissue-chewing caterpillars of the cabbage white butterfly P. rapae.

Supplement 4: MS Excel file with normalized expression levels, fold-change information,AGI numbers and TIGR annotation of the selected probe sets that showed a consistent >2-foldchange (up or down) 12 and 24 h after infestation of Arabidopsis Col-0 plants with larvaeof the cell-content feeding Western flower thrips F. occidentalis.

Supplement 5: MS Excel file with normalized expression levels, fold-change information,AGI numbers and TIGR annotation of the selected probe sets that showed a consistent >2-foldchange (up or down) 48 and 72 h after infestation of Arabidopsis Col-0 plants with phloem-feeding M. persicae aphids.

Supplement 6: MS Excel file with normalized expression levels, fold-change information,AGI numbers and TIGR annotation of the selected probe sets that showed on at least twotime points a consistent >2-fold change (up or down) at 1, 3, or 6 h after treatment ofArabidopsis Col-0 plants with 0.05 mM MeJA.

All supplementary materials can be downloaded from:http://www.bio.uu.nl/~fytopath/GeneChip_data.htm

60


61

62

Analysis of genome-wide transcriptome changes of Arabidopsisduring pathogen and insect attack using the visualization toolMapMan

Vivian R. Van Oosten1,2, L.C. Van Loon2, Marcel Dicke1, and Corné M.J. Pieterse2

1 Graduate School Experimental Plant Sciences, Laboratory of Entomology, Wageningen University,

P.O. Box 8031, 6700 EH Wageningen, The Netherlands


Biology, Faculty of Science, Utrecht University, P.O. Box 800.84, 3508 TB Utrecht, The Netherlands

63Analysis of genome-wide transcriptomechanges of Arabidopsis during pathogenand insect attack using the visualizationtool MapMan

chapter 3

AB

ST

RA

CT Plant pathogens and herbivorous insects are major challenges for plants to combat. To defend them-

selves against these highly diverse enemies, plants utilize sophisticated defensive strategies. To

understand the mechanisms underlying the specific defense responses activated by the plant upon

attack, we previously monitored the signal signature and genome-wide transcriptome changes of

Arabidopsis during attack by a pathogenic leaf bacterium (Pseudomonas syringae pv. tomato), a patho-

genic leaf fungus (Alternaria brassicicola), tissue-chewing caterpillars (Pieris rapae), cell-content-feeding

thrips (Frankliniella occidentalis), or phloem-feeding aphids (Myzus persicae). Analysis of global gene

expression profiles demonstrated that the phytohormone signal signature characteristic of each

Arabidopsis-attacker combination is orchestrated into a surprisingly complex set of transcriptional

alterations. Although most of the attackers stimulated JA biosynthesis, the majority of the changes

in JA-responsive gene expression were attacker-specific. To investigate the nature of the attacker-

induced changes, we made use of the software program MapMan. This package allows the display

of microarray data sets onto pictorial diagrams that represent metabolic pathways and other biological

processes. Although the majority of the transcriptome changes were highly attacker-specific, visuali-

zation of the changes in the five different Arabidopsis-attacker combinations revealed that the biological

processes affected were strikingly similar. This indicates that the different attackers induce changes in

similar plant processes through largely non-overlapping transcriptional alterations. Yet, infestation by

M. persicae induced a transcriptional response that was opposite to those due to the other attackers

or exogenous application of MeJA.

64

IntroductionA broad range of organisms, including microbial pathogens and herbivorous insects, dependon plants as important sources of nutrients. In order to protect themselves, plants have developedsophisticated defense mechanisms, whereas their enemies have found ways to overcome them.This evolutionary arms race has provided plants with a large variety of constitutive defensemechanisms, including pre-existing physical and chemical barriers, and inducible defenseresponses that become activated only upon attack (Van Loon, 2000; Dicke and Van Poecke, 2002).

Plant pathogens include viruses, bacteria, fungi and oomycetes, whereas insect herbivoresrange from tissue-chewing caterpillars to phloem-sucking aphids. Each attacker has developedspecific strategies to exploit either one or a few host plants (specialists) or a wide range ofdifferent plant species (generalists). Plant pathogens can be roughly divided into two groupsdepending on their lifestyle: biotrophs and necrotrophs (Thomma et al., 2001; Glazebrook,2005). Biotrophic pathogens exclusively retrieve nutrients from living plant cells, and haveusually developed a highly specialized relationship with their hosts. Effective plant resistanceagainst biotrophs is often associated with the salicylic acid (SA)-dependent hypersensitiveresponse (HR), which kills the infected cells and encloses the invading pathogen. In contrast,necrotrophic pathogens are mostly generalists that require dead plant tissue for their nutrientsupply. For successful invasion, they need to kill host cells by secreting toxins or lytic enzymes.Resistance towards this group of microbes has been shown to depend mainly on jasmonicacid (JA) signaling. Some pathogens start with a biotrophic phase and switch to a necrotrophiclifestyle once infection has been established. Such pathogens are called hemi-biotrophs(Glazebrook, 2005).

Herbivorous insects can be categorized into three groups based on their feeding modes:tissue-chewing, cell-content-feeding and phloem-sucking insects (Walling, 2000).Tissue-chewinginsects, such as caterpillars, wound plants by feeding on the foliage. Plant responses sharecharacteristics with the JA-dependent wounding response and involve healing of the site ofinjury. JA-dependent defenses have been demonstrated to be effective against a wide rangeof tissue-chewing insects in several host plants (Howe et al., 1996; McConn et al., 1997; Thaleret al., 2002; Reymond et al., 2004). The ethylene (ET)-insensitive Arabidopsis mutant ein2was demonstrated to be less resistant to Spodoptera littoralis (Egyptian cotton worm) larvae(Stotz et al., 2000), whereas Arabidopsis mutants and transgenics that are compromised inSA-dependent defense responses have been shown to exhibit enhanced resistance againstfeeding by the cabbage looper Trichoplusia ni (Cui et al., 2002).Thus, ET- and SA-dependentdefenses also contribute to plant resistance against tissue-chewing insects.

Cell-content-feeders, like spider mites and thrips, are herbivores which use stylets topuncture leaf cells and ingest their contents leaving whitish spots of collapsed cells (Helle andSabelis, 1985; Parker et al., 1995). Evidence exists for a role of JA in the defense against cell-content-feeders in tomato and lima bean (Arimura et al., 2000; Li et al., 2002). In addition,arthropod herbivores such as spider mites induce the emission of methyl salicylate (MeSA) in many plant species (Takabayashi and Dicke, 1996; Ament et al., 2004; De Boer and Dicke,2004), which can lead to the activation of SA-inducible defense-related genes (Arimura etal., 2000; Kant et al., 2004).

Chapter 03: Analysis of genome-wide transcriptome changes of Arabidopsis during pathogen and insect attack using the visualization tool MapMan

65

Phloem-feeding insects, particularly aphids, move their stylets between plant cells throughthe middle lamellae until they reach the sieve elements on which they start feeding (Tjallingiiand Hogen Esch, 1993), largely avoiding wounding the host plant. In Arabidopsis and sorghum,aphids have been shown to induce genes associated with JA-, as well as SA-dependent defense(Moran and Thompson, 2001; Moran et al., 2002; Zhu-Salzman et al., 2004). Moreover, com-pared to wild-type Arabidopsis, aphids perform better on Arabidopsis mutants blocked in JAsignaling and worse on plants with reduced SA signaling, indicating a role for both signalingcompounds in defense against aphids (Ellis et al., 2002; Mewis et al., 2005). However, althoughSA-responsive PR-1 expression is induced locally, Arabidopsis does not produce detectableamounts of SA, JA or ET during aphid infestation, (De Vos et al., 2005, Chapter 2). Aphidsseem to elicit an atypical response that remains largely undefined (Thompson and Goggin,2006).Taken together, it can be concluded that insect feeding can result in dramatic woundingof leaves, as well as in a more intimate, long-lasting interaction involving a feeding site in the phloem.

Evidently, for a plant to defend itself against such a wide range of microbial and herbi-vorous attackers, it needs to be able to recognize the invader and respond with an appropriatedefense. The regulation of induced plant defenses is highly complex. Molecular and geneticapproaches have identified JA, ET and SA as major players with cross-communicating signal-ing pathways involved in pathogen and insect defense (Walling, 2000; Pieterse et al., 2001;Dicke and Van Poecke, 2002; Rojo et al., 2003; Pozo et al., 2004; Glazebrook, 2005; Beckersand Spoel, 2006). Plants are thought to be equipped with such sophisticated defense signalingtools in order to respond quickly and specifically to attack at minimal costs (Walling, 2000).Moreover, when a plant is under attack by several pathogens or insects, it needs to be able to prioritize its defense responses.

We are interested in understanding how a plant fine-tunes its defensive capacities. Tothis end, we previously monitored JA, ET and SA production and genome-wide transcriptomechanges in Arabidopsis during pathogen and insect attack (De Vos et al., 2005, Chapter 2).We noticed that the phytohormone signal signature produced upon attack by an avirulentstrain of a hemi-biotrophic leaf bacterium (Pseudomonas syringae pv. tomato), a necrothrophicleaf fungus (Alternaria brassicicola), tissue-chewing caterpillars (Pieris rapae), cell-content-feeding thrips (Frankliniella occidentalis) and phloem-feeding aphids (Myzus persicae) differedhighly in composition, quantity and timing. Analysis of genome-wide transcriptional changesin each of the plant-attacker combinations demonstrated that the characteristic signal signatureswere orchestrated into a surprisingly complex set of transcriptional alterations. Stress-relatedgenes were overrepresented in each Arabidopsis-attacker combination. Moreover, JA wasfound to be the main phytohormone produced upon attack with A. brassicicola, P. rapae or F. occidentalis.These three invaders induced transcriptional changes of which a large proportion(44-69%) was JA-responsive. However, the JA-responsive genes overlapped only partly (6-54%),indicating that the changes in the expression of most of the JA-responsive genes were attacker-specific. This suggests that other regulatory mechanisms eventually shape the outcome of theattacker-specific defense response.

To gain functional insight in the transcriptional changes induced in Arabidopsis uponattack by different pathogens and insects, and to investigate the role of JA therein, we made

66

use of the MapMan visualization tool (Thimm et al., 2004; Usadel et al., 2005). This softwarepackage allows the categorization of Arabidopsis genes with altered transcription into functionalgroups (BINs) that represent a particular cellular process, a biological response or an enzymefamily. Moreover, the statistical module facilitates the identification of BINs containing genesets with a significantly different response from those in other BINs. Each of the significantlydifferent BINs is likely to contain a group of co-responsive genes (Usadel et al., 2005), thatmay encode proteins that are involved in the same pathway, are members of the same complex,or are otherwise functionally connected (Lee et al., 2004). The genomic data sets can be dis-played onto pictorial diagrams that represent a biological function. By visualizing transcriptionalchanges, it becomes possible to discover patterns that are not immediately obvious by studyingindividual genes. We used these features to analyze the nature of the global gene expressionprofiles induced by the five attackers in comparison with the response of Arabidopsis uponexogenous application of methyl JA (MeJA).

ResultsCategorization of selected transcriptome changes of fiveArabidopsis-attacker combinations into main functional groups(BINs) using MapMan software

Previously, we studied genome-wide transcriptional changes in five Arabidopsis-attacker com-binations (De Vos et al., 2005, Chapter 2). To investigate the nature of the attacker-inducedchanges, we made use of MapMan software (version 1.7.0) (Thimm et al., 2004; Usadel et al.,2005). First, we identified the biological processes that are affected by the induced gene changes.To this end, we studied the genes that showed robust changes at two time points after attack(De Vos et al., 2005, Chapter 2, Supplementary Tables 1-5).The expression ratios for both timepoints were averaged, log2 transformed and imported in the TranscriptScavenger module ofthe MapMan software package.This module classifies the genes represented on the ArabidopsisATH1 full-genome GeneChips into hierarchical functional categories (BINs, subBINs, individualenzymes, etc.). Table 1 shows the categorization of the selected genes into BINs for the fiveArabidopsis-attacker combinations. For each BIN, the number of genes with altered expressionis shown per attacker. P. syringae was the only attacker that induced transcript changes in all35 BINs, demonstrating the large impact on cellular processes upon infection with this pathogen.The genes affected by the four other attackers are distributed over fewer BINs, indicating thatnot all functional groups were affected.


67

As outlined in Materials and Methods, statistical analysis was applied to identify clustersof co-responding genes with assigned cellular functions. It was tested whether the expressionvalues of genes belonging to each BIN differed significantly (P < 0.05) from those in other BINs.Statistical analysis identified significantly different BINs during P. syringae or M. persicaeattack (Table 1).The BINs representing the response to P. syringae were put in order of signifi-cance, starting with the lowest P-value (BIN 1, Photosynthesis, 38 genes affected, P = 2.05E-08).For the other attackers, the same order is followed, and BINs are not in the order of increasingP-value. During the transcriptional response of Arabidopsis to P. syringae infection, five BINswere found to be significantly different from the other BINs with four having an assignedfunction: BIN 1 “Photosynthesis”, BIN 26 “Large enzyme families”, BIN 20 “Stress” andBIN 16 “Secondary metabolism” (Table 1). This indicates that the genes in these categorieshave expression ratios that are significantly different from genes in other categories in responseto P. syringae infection. Hence, each of these BINs is likely to contain a set of co-responsivegenes (Usadel et al., 2005). In the transcriptional response to M. persicae infestation, twosignificantly different BINs were identified: BIN 26 “Secondary metabolism” and BIN 29“Protein” (Table 1).The other attackers affected less than 200 genes distributed over all BINs;these were probably too few to reveal significant differences between individual BINs. Thenumber of genes affected per BIN can also be an indication of the relative importance of that functional group. For instance, BIN 26 “Large enzymes families” was relatively highlyaffected by P. rapae (34 genes = 18%), F. occidentalis (37 genes = 19%) and A. brassicicola(41 genes = 27%).

The most highly affected functional categories shared by all attackers include “Largeenzyme families”, “Stress”, and “Secondary metabolism”, highlighting the general impor-tance of these functional groups in the induced defense response of Arabidopsis againstpathogens and insects (Table 1). In addition, some attackers had an effect on specific functionalcategories that were less affected by the other attackers. Examples are the BINs “Photosynthesis”for P. syringae, and “Protein” for M. persicae. Finally, a large proportion of the respondinggenes (19-29%) were in the “Not assigned” category (BIN 35), indicating that the nature of many transcriptional changes is not evident.

Categorization of statistically significant BINs into functionalsubgroups (subBINs)

As BINs containing significantly different transcript changes were identified for the responseof Arabidopsis to P. syringae or M. persicae, these were analyzed in more detail by furtherdividing them into functional subgroups (subBINs).The transcript changes in the subBINs werestatistically analyzed for significant differences among the subBINs within each BIN.The resultsare presented in Table 2, which shows all the BINs, subBINs, subsubBINs, etc. that were signi-ficant in one of the two interactions, and thus may contain a set of co-responsive genes. ThesubBINs that were not significant in either interaction were omitted from the Table. Statisticalanalysis revealed that P. syringae infection caused a significantly different response in thesubsubBIN 27.3.32 “WRKY domain transcription factor family” from all other transcription


69

factor families in subBIN 27.3 “Regulation of transcription” (Table 2), suggesting that thegenes encoding WRKY transcription factors are expressed coordinately.

70

Although the numbers of genes affected in the selected BINs were roughly similar,statistical analysis revealed clear differences in the direction of the transcriptional responses.Genes in the category “Stress” (BIN 20) were likely to be co-responsive during P. syringaeinfection (104 genes, P = 1.22E-04), but not upon aphid feeding (92 genes, P = 0.664).Likewise, each of the subBINs “Cytochrome P450s” and “Glutathione S-transferases” ofBIN 26 were likely to contain genes co-responsive to P. syringae infection (P = 8.21E-03and P = 1.72E-03, respectively), but not to M. persicae infestation (P = 0.328 and P = 0.843,respectively), despite similar numbers of genes with altered transcription.Thus, both attackersaffected similar numbers of genes in these functional categories, indicating that genes assignedto these BINs are responsive to P. syringae as well as M. persicae. However, only upon P. syringaeinfection, the genes belonging to these BINs responded significantly differently from the otherBINs, indicating that these gene sets probably responded in a coordinate matter.

Both attackers induced significantly different changes in BIN 16 “Secondary metabolism”.However, only aphids significantly affected two subBINs within this functional group:“Isoprenoids” (P = 0.040) and “Flavonoids” (P = 0.015), indicating that genes related to themetabolism of these compounds had a similar response to aphid infestation, but not to P. syringaeinfection. Moreover, statistical analysis of BIN 29 “Protein” revealed that each attacker had a distinct, significant effect within this category: P. syringae infection affected genes with arole in protein synthesis (subBIN 29.2; P = 1.02E-03), whereas M. persicae feeding alteredthe transcription of genes involved in protein degradation (subBIN 29.5; P = 7.47E-04).These results indicate that statistical analysis of the direction of transcript changes can identifypotential effects in important biological processes, whereas the mere analysis of numbers ofgenes would not identify these differences. Moreover, statistical analysis of BINs and subBINsrevealed that M. persicae and P. syringae each had a specific, yet distinct effect on the samefunctional categories.

Display of categorized transcript changes onto MapMan diagrams

To study the potentially co-responsive transcript changes in the functional groups in more detail,the MapMan files created with the TranscriptScavenger Module were displayed onto pictorialdiagrams through the ImageAnnotator Module. The diagrams represent cellular functions,biological responses or classes of enzymes. Each gene is individually represented in the dia-gram, while the direction of the change in expression is reflected by the color and intensity(up-regulation blue, down-regulation red).

To illustrate how a biological response can be visualized in MapMan, the response ofArabidopsis during P. syringae infection was displayed onto the “Photosynthesis” diagram(Fig. 1). The diagram pictured all, except two, affected genes in BIN 1 “Photosynthesis” witha red color, indicating that the majority of genes were down-regulated. Moreover, most geneswith altered transcription were categorized in the subBIN 1.1 “Light reaction”, suggestingthat this process was affected most severely.The visual representation of the effect of P. syringaecorresponded perfectly with the statistical analysis shown in Table 2.The statistical test revealedthat P. syringae infection induced the most different transcriptome changes in BIN 1


71

“Photosynthesis” (P = 2.05E-08) when compared to all other BINs. Moreover, subBIN 1.1“Light reaction” (P = 5.67E-07) was significantly different from the other “Photosynthesis”subBINs. Within the subBIN “Light reaction”, the subsubBIN 1.1.1 “Photosystem II”(P = 2.35E-03) was significantly different from the other subsubBINs. These results indicate

that within the category “Photosynthesis”, these subcategories contain genes with a significantlydifferent response from the other subcategories. In Figure 1, it can be observed that 13 of the38 affected genes (34%) were assigned to the subBIN “Photosystem II”, and responded in acoordinate manner: they were all down-regulated.

The following paragraphs describe the MapMan analysis of the genes in the four mostaffected BINs in all of the five Arabidopsis-attacker combinations: BIN 20 “Stress”, BIN 26“Large enzyme families”, BIN 16 “Secondary metabolism” and BIN 29 “Protein”. TheseBINs contained either significantly differently responding gene sets or relative large numbersof responding genes when compared to the other BINs.The responses to A. brassicicola, P. rapaeand F. occidentalis are discussed in comparison with each other, because these attackersinduced a similar phytohormone signal signature, and comparable numbers of genes (150-200genes) (De Vos et al., 2005, Chapter 2). Likewise, the responses to P. syringae and M. persicaeare compared, because these invaders induced a similarly larger number of genes (about 2,000genes).

FIGURE 1. MapMan “Photosynthesis” diagram showing the effect of P. syringae infection on photosynthesis-relatedgenes in Arabidopsis.

Previously selected (De Vos et al., 2005, Chapter 2) Arabidopsis genes with altered transcription levels upon P. syringaeinfection were categorized into functional groups (BINs) using MapMan software. The transcriptional changes withinthe BIN “Photosynthesis” were loaded onto the “Photosynthesis” diagram. This diagram depicts the subBINs “Lightreaction”, “Calvin cycle” and “Photorespiration”. Single genes are represented by a square, of which the color indicatesthe direction of transcriptional change (up-regulation blue, down-regulation red). Color intensity indicates the fold changeratio on a log2 scale.

72

BIN 20 “Stress”

In our previous study, we noticed an overrepresentation of stress-related genes in the transcrip-tome changes of Arabidopsis upon attack by the five different pathogens or insects (De Vos et al., 2005, Chapter 2).We had categorized the affected genes according to The ArabidopsisInformation Resource Gene Ontology Tool (TAIR) (Rhee et al., 2003). Here, using MapMansoftware resulted in the identification of the significantly different BIN 20 “Stress” upon P. syringae infection (104 genes, P = 1.22E-04,Table 2).Aphid feeding induced a similar numberof genes (92) in the same BIN, but statistical analysis did not identify the BIN “Stress” to besignificantly different from the other BINs. F. occidentalis induced 17 genes (9%) and A. brassicicola 18 genes (12%), while P. rapae only affected 4 genes (2%) in this BIN.Hence, all attackers, with the exception of P. rapae, induced a relatively high number of stress-related genes, confirming previous results.

In order to visualize the transcript changes in the BIN “Stress”, the expression ratios of the genes affected by each attacker were displayed onto the MapMan “Cellular response”diagram (Fig. 2). This diagram displays the genes of four different BINs: 20 “Stress” (subBINs“Abiotic stress” and “Biotic stress”), 21 “Redox regulation” or “Redox”, 31 “Cell” (subBINs“Cell cycle” and “Cell division”), and 33 “Development”. Since we had noticed a large numberof JA-responsive genes in the majority of Arabidopsis-attacker combinations (De Vos et al.,2005, Chapter 2), the selection of genes responding to application of MeJA (De Vos et al.,2005, Chapter 2, Supplementary Table 6) was also included in the MapMan analysis.

F. occidentalis and A. brassicicola induced genes involved mainly in biotic stress, where-as P. syringae and M. persicae affected both biotic and abiotic stress genes (Fig. 2). The genesaffected were not identical for the different attackers. However, they largely belonged to thesame functional groups, indicating that a different transcriptional response to each attackeraffected similar cellular processes. Although each attacker induced a specific transcriptionalresponse, the direction of the response was generally similar to the one upon application ofMeJA, except for M. persicae. The direction of the transcriptional responses observed uponinfestation by M. persicae appeared to be opposite to that observed upon P. syringae infectionor MeJA treatment. For instance, upon M. persicae attack, genes in the subcategories “Heat”in the subBIN “Abiotic stress” and “Ascorbate/glutathione” of the BIN “Redox regulation”were down-regulated, whereas P. syringae infection or MeJA application up-regulated genesin these subBINs (Fig. 2).


73

FIGURE 2. MapMan “Cellular response” diagram for Arabidopsis genes that are up- or down-regulated in response to attackby P. rapae, F. occidentalis, A. brassicicola, P. syringae, or M. persicae, or upon application of MeJA.

The MapMan “Cellular response” diagram displays the genes of four different BINs: BIN 20 “Stress” (subBINs “Biotic stress”and “Abiotic stress”), BIN 21 “Redox regulation” or “Redox”, BIN 31 “Cell” (subBINs “Cell division” and “Cell cycle”), and BIN 33“Development”. Single genes are represented by a square, of which the color indicates the direction of transcriptional change(up-regulation blue, down-regulation red, no change white). Color intensity indicates the fold change ratio on a log2 scale.

74

F. occidentalis M. persicae

A. brassicicola MeJA

P. rapae P. syringae

BIN 26 “Large enzyme families”

Previously, we observed that JA is the major phytohormone involved in the response ofArabidopsis to attack by the pathogens P. syringae and A. brassicicola, and by the insect herbivores P. rapae and F. occidentalis (De Vos et al., 2005, Chapter 2). These attackers stimulated JA production in Arabidopsis. In addition, a large part of the transcriptome changesinduced by these invaders (32-69%), was also changed upon MeJA application. Although M. persicae did not induce detectable amounts of JA, 30% of the transcriptionally alteredgenes were also affected by MeJA application. To obtain more insight into the role of JA inthe specific defense responses of Arabidopsis upon attack by these pathogens and insects,we compared the attacker-induced transcriptional changes with the changes that were inducedupon exogenous application of 50 µM MeJA. To this end, two separate MapMan files werecreated per attacker: one file containing all genes that are changed in response to the specificattacker, and another file containing all MeJA-responsive genes from the first file.Subsequently, the transcriptome changes were visualized onto diagrams and compared with the response to application of MeJA.

Since BIN 26 “Large enzyme families” was highly affected in all Arabidopsis-attackercombinations (Table 1), we displayed the transcript changes occurring in this functional grouponto the corresponding diagram in Figure 3.The majority of the effects of attack by P. rapae,F. occidentalis and A. brassicicola were on genes from the subBINs “Cytochrome P450s”,“Glutathione S-transferases” and “UDP glycosyltransferases”. In addition, P. rapae feedinghad an effect on several genes from the subBIN “Oxidases” and A. brassicicola infection ongenes from the subBIN of “Nitrilases” (Fig. 3). Many genes affected in the BIN “Largeenzyme families” upon attack by P. rapae, F. occidentalis, or A. brassicicola were JA-responsive(74, 84 and 68%, respectively). Moreover, these genes were typically up-regulated, resemblingthe majority of the MeJA-induced changes. These results demonstrate that the regulation ofseveral enzyme families of Arabidopsis in these three interactions was highly influenced by JA.

Of all P. syringae-responsive genes in the BIN “Large enzyme families”, 52% was JA-responsive.The subBINs “Cytochrome P450s” (P = 8.21E-03) and “Glutathione S-transferases”(P = 1.72E-03) were significantly different from the other subBINs of BIN 16, suggestingcoordinate gene expression in these subgroups (Table 2). Indeed, 94% of the P. syringae-responsive genes and all JA-responsive, P. syringae-induced genes were up-regulated (Fig. 3).Within the BIN “Large enzyme families”, 36% of the M. persicae-affected genes were JA-responsive. In contrast to the other attackers, both the JA-responsive and the JA-non-responsivegenes were mainly down-regulated. This indicates that aphid feeding had a different effect on the transcriptional alterations in Arabidopsis than the other four attackers. Although themajority of the changes within the BIN “Large enzyme families” occurred within the samesubBINs, the specific genes that were affected were attacker-specific. Thus, although similarenzyme families were affected by the different attackers, the transcriptional responses didnot overlap.


75

76

P. rapae-induced changes (all) P. rapae-induced changes (P. r./MeJA-responsive)

F. occidentalis-induced changes (all) F. occidentalis-induced changes (F. o./MeJA-responsive)

A. brassicicola-induced changes (all) A. brassicicola-induced changes (A. b./MeJA-responsive)

(note: Figure 3 is divided over two pages)


77

P. syringae-induced changes (all) P. syringae-induced changes (P. s./MeJA-responsive)

M. persicae-induced changes (all) M. persicae-induced changes (M. p./MeJA-responsive)

MeJA-induced changes (all)

FIGURE 3. MapMan “Large enzyme families” diagram forArabidopsis genes that are up- or down-regulated in response to attack by P. rapae, F. occidentalis, A. brassicicola, P. syringae,or M. persicae, or upon application of MeJA.

For each Arabidopsis-attacker interaction two diagrams are shown:first, the response of all genes affected by this attacker, and thena selection of these genes that responded also to the applicationof MeJA. Single genes are represen-ted by a square, of whichthe color indicates the direction of transcriptional change (up-regulation blue, down-regulation red, no change white). Colorintensity indicates the fold change ratio on a log2 scale.

BIN 16 “Secondary metabolism”

All attackers induced changes in BIN 16 “Secondary metabolism” (Table 1). Moreover, statis-tical analysis predicted that genes in this BIN responded differently from the genes in the otherBINs upon attack by P. syringae (P = 1.72E-03) or M. persicae (P = 3.39E-06) (Table 2). Tofurther examine the importance of JA in the different Arabidopsis-attacker combinations, thetwo MapMan files containing the transcriptome changes of all Arabidopsis genes or the selectionof JA-responsive genes were used to create diagrams of the genes affected (Fig. 4). In similaritywith the BIN “Large enzyme families”, the transcripts affected by P. rapae, F. occidentalisand A. brassicicola were up-regulated and predominantly JA-responsive (65%, 75% and 50%respectively). However, each attacker influenced specific subgroups of secondary metabolism.P. rapae and F. occidentalis infestation mainly changed transcript levels of genes in the subBINs“Anthocyanins” and “Glucosinolates”. In addition, P. rapae affected genes involved in thesubBINs “Phenylpropanoids”, “Lignins and lignans” and “Dihydroflavonols”, whereas A. brassicicola infection affected genes in the subBIN “Alkaloid-like”-compounds.

Both P. syringae and M. persicae had an effect on many different subgroups of genes witha role in secondary metabolism. The changes upon P. syringae infection were largely JA-responsive (58%), and most of the affected genes were up-regulated. Despite the fact that a largeproportion (39%) of the M. persicae-affected genes in this BIN were also JA-responsive, thosegenes were predominantly down-regulated.The subBINs “Isoprenoids” and “Flavonoids” weresignificantly different from the other subBINs in the response to M. persicae (Table 2). ThesubBIN “Isoprenoids” consists of the subcategories “Mevalonate pathway” (MVA), “Non-MVApathway”, ”Tocopherol”, “Carotenoids” and “Terpenoids”.The subBIN “Flavonoids” is dividedin the categories “Anthocyanins”, “Chalcones”, “Dihydroflavonols”, “Flavonols” and“Isoflavonoids”. Figure 4 shows that whereas P. syringae infection and application of MeJAactivated isoprenoid- and flavonoid-related genes, M. persicae infestation suppressed them.Indeed, aphid infestation resulted in the down-regulation of a large group of genes in thesubBINs “Phenylpropanoid” and “Lignin and lignans”, containing genes related to the bio-synthesis of these compounds. The effect on the genes in these subBINs was predominantlyJA-independent, because these genes are not present in the respective subBINs of the MeJA-responsive genes.

Thus, all attackers, except aphids, induced a transcriptional response in Arabidopsis thatnot only overlapped to a large extent with the response upon application of MeJA in termsof identical genes affected, but also in terms of direction of induction (up- or down-regulation).This highlights the importance of JA in secondary metabolism during Arabidopsis defenseagainst the majority of attackers. Moreover, visualization of the transcriptome changes in the BIN “Secondary metabolism” in the five different Arabidopsis-attacker combinations revealedthat each interaction resulted in specific transcriptional alterations, mainly in the same subBINs.These results indicate again that similar plant responses are affected, but through predominantlynon-overlapping transcriptional responses.

78


79

P. rapae-induced changes (all) P. rapae-induced changes (P. r./MeJA-responsive)

F. occidentalis-induced changes (all) F. occidentalis-induced changes (F. o./MeJA-responsive)

A. brassicicola-induced changes (all) A. brassicicola-induced changes (A. b./MeJA-responsive)

(note: Figure 4 is divided over two pages)

80

P. syringae-induced changes (all) P. syringae-induced changes (P. s./MeJA-responsive)

M. persicae-induced changes (all) M. persicae-induced changes (M. p./MeJA-responsive)

MeJA-induced changes (all)

FIGURE 4. MapMan “Secondary metabolism” diagram forArabidopsis genes that are up- or down-regulated in response to attack by P. rapae, F. occidentalis, A. brassicicola, P. syringae,or M. persicae, or upon application of MeJA.

For each Arabidopsis-attacker interaction two diagrams are shown:first, the response of all genes affected by this attacker, and thena selection of these genes that responded also to the applicationof MeJA. Note that the categories “Mevalonate pathway” (MVA),“Non-MVA pathway”, ”Tocopherol”, “Carotenoids” and “Terpenoids”all belong to the subBIN “Isoprenoids”. Single genes are repre-sented by a square, of which the color indicates the direction oftranscriptional change (up-regulation blue, down-regulation red,no change white). Color intensity indicates the fold change ratioon a log2 scale.


81

BIN 29 “Protein”

For several BINs, we observed that feeding by the aphid M. persicae led to the down-regulationof Arabidopsis genes that were up-regulated upon invasion by the other attackers or applicationof MeJA (Fig. 2-4). Aphid infestation induced a large number of genes involved in “Proteindegradation” (subBIN 29.5, 124 genes; Table 2). Moreover, these genes responded in a signifi-cantly different fashion (P = 7.47E-04) when compared to the other subBINs in BIN 29“Protein”. This notion prompted us to investigate whether the effect on the genes related toprotein degradation was also opposite of the transcriptional effects of MeJA application. Themain proteolytic pathway in eukaryotes is the ubiquitin/26S proteasome system (reviewed in Smalle and Vierstra, 2004), which is depicted in the “Proteasome” diagram. Therefore, thetranscript changes induced by aphid feeding were visualized onto the “Proteasome” diagramand compared with the response to application of MeJA.The “Proteasome” diagram representsall genes in the subsubBIN 29.5.11 “Ubiquitin dependent protein degradation” (71 genes,P = 1.13E-04; Table 2).

Figure 5 shows the transcript changes of genes in this category in response to aphidinfestation and exogenous application of MeJA.The majority of the M. persicae-induced changesconsist of up-regulated genes, whereas the majority of the MeJA-responsive genes are down-regulated. Aphid feeding significantly changed the expression of 19 genes encoding F-boxproteins, 15 of which were up-regulated. Exogenous application of MeJA induced changes inthe expression of 15 F-box protein genes, but the majority of these (13) were down-regulated.Six F-box protein-encoding genes responded to both aphids and MeJA, of which five wereaffected in opposite directions. F-box proteins play an important role in targeting proteins for ubiquitin-dependent E3-ligase protein degradation in the proteasome (reviewed in Smalleand Vierstra, 2004).Apparently, aphid feeding and MeJA treatment affect this process in oppositedirections.

FIGURE 5. MapMan “Proteasome” diagram for Arabidopsis genes that are up- or down-regulated in response to attack byM. persicae, or upon application of MeJA.

The “Proteasome” diagram displays all genes in the subsubBIN 29.5.11 “Ubiquitin dependent protein degradation”. It representsthe previously selected Arabidopsis genes responding to attack by M. persicae (all genes and a MeJA-responsive selection)or to application of MeJA (all genes and a M. persicae-responsive selection). Thus, two diagrams displaying the response to M. persicae are shown, the first representing all genes responding to M. persicae infestation, the second a selection ofthese genes that also responded to the application of MeJA. Conversely, one diagram displays all genes responding to theapplication of MeJA, and the other a selection of those genes that also responded to M. persicae infestation. Single genesare represented by a square, of which the color indicates the direction of transcriptional change (up-regulation blue, down-regulation red, no change white). Color intensity indicates the fold change ratio on a log2 scale.

82

M. persicae-induced changes (all) M. persicae-induced changes (M. persicae/MeJA-responsive)

MeJA-induced changes (all) MeJA-induced changes (M. persicae/MeJA-responsive)

DiscussionPlants can protect themselves against a wide range of microbial and insect invaders by theinduction of various defenses upon attack. The regulation of induced plant defenses involvesa limited number of plant hormones, but is highly complex. In order to understand how a plantresponds to a broad spectrum of microbes and insects, and how this response leads to effec-tiveness against various attackers, we subjected Arabidopsis to attack by a range of pathogensand insects. The plant response was investigated by determining the phytohormone signal signature produced and the genes expressed (De Vos et al., 2005, Chapter 2). Here, we investi-gated the nature of the transcriptional changes in Arabidopsis induced by P. syringae,A. brassicicola, P. rapae, F. occidentalis and M. persicae, using the MapMan visualizationtool and built-in statistical analysis to identify clusters of co-responding genes with assignedcellular functions (Thimm et al., 2004; Usadel et al., 2005). Despite the limited overlap ofidentical genes with altered expression levels during attack, the processes affected by the different attackers were largely similar. Hence, the same cellular processes appear to be affectedthrough non-overlapping transcriptional alterations.

Categorization and statistical analysis of selected transcriptomechanges of five Arabidopsis-attacker combinations into functionalgroups (BINs) using MapMan software

In order to study the nature of the transcriptome changes during the five Arabidopsis-attack-er combinations, we made use of MapMan software. MapMan analysis of previously selectedgenes (De Vos et al., 2005, Chapter 2) revealed that the most affected categories shared by allattackers include the categories “Stress”, “Large enzyme families”, and “Secondary metabolism”,demonstrating the general importance of these functional groups in the induced defense responseof Arabidopsis against pathogens and insects (Table 1).

For the response to P. syringae or M. persicae, statistical analysis in MapMan revealedclear differences in the transcriptional changes of gene sets assigned to certain BINs in com-parison with all other BINs, even though the numbers of genes affected were roughly similar.Examples are the BINs “Photosynthesis” for P. syringae-infected plants, and “Protein” forM. persicae-infested plants (Table 2). Statistical analysis between BINs and subBINs indicatedthat P. syringae and M. persicae each had a specific effect on the same functional categories.The BIN “Protein” nicely illustrates this, because P. syringae infection significantly affected“Protein synthesis”, whereas M. persicae feeding significantly affected the subBIN “Proteindegradation” (Table 2). Significantly different BINs are likely to contain co-responding genes,suggesting that the genes assigned to the BIN “Photosynthesis” co-responded to P. syringaeinfection, whereas the genes assigned to the subBIN “Protein degradation” co-responded toaphid feeding. Indeed, the genes related to photosynthesis were predominantly down-regulatedupon P. syringae infection (Fig. 1), while the genes involved in protein degradation were to a large extent up-regulated after M. persicae infestation (Fig. 4). These results indicate thatstatistical analysis of the transcriptional changes of genes assigned to functional categories


83

can facilitate the identification of important biological processes during response studies.The 16 WRKY transcription factors affected by P. syringae (Table 2) were all up-regulated

(diagram not shown). Conserved WRKY protein binding sites are present in the promoters ofmany plant genes associated with plant defense (Rushton and Somssich, 1998). Moreover, 43of 72 studied Arabidopsis WRKY transcription factors are up-regulated in Arabidopsis by avirulentP. syringae or SA application (Dong et al., 2003). Our results correspond well with theseexperiments, because 13 of the WRKY transcription factors induced by P. syringae in our studywere also up-regulated by P. syringae in the study of Dong and co-workers (Dong et al., 2003).Nine of these proteins belonged to group II WRKY transcription factors (Eulgem et al., 2000).Recently, 3 of these WRKY proteins from our study,WRKY18,WRKY40 and WRKY60, wereshown to interact physically and functionally with each other in the response of Arabidopsisto pathogens (Xu et al., 2006). This confirms that statistical analysis using MapMan can iden-tify co-responsive groups of genes with a potentially important biological effect.

Display of transcript changes onto MapMan diagram revealeddown-regulation of genes related to photosynthesis uponinfection by P. syringae

Together with the categorization and statistical analysis, the visualization onto MapMan diagramsrevealed in detail how specific biological processes were transcriptionally affected as a resultof pathogen or insect attack. During the Arabidopsis response to P. syringae, genes from theBIN “Photosynthesis” were responding significantly differently from those in the other BINs(P = 2.05E-08). Likewise, the gene sets in two of the “Photosynthesis” subBINs, “Light reaction”and “Calvin cycle”, responded significantly differently from the genes in the other subBINs(Table 2), suggesting that these gene sets are co-responsive. Indeed, we showed the down-regulation of 36 genes involved in photosynthesis upon P. syringae infection (Fig. 1). For soy-bean, it has been shown that infection with Pseudomonas syringae pv. glycinea carrying theavirulence gene avrB resulted in the down-regulation of nearly 100 putative chloroplast genes(Zou et al., 2005). Moreover, the down-regulation of these genes was associated with a reductionin photosynthesis that occurs in the chloroplasts.The suppression of photosynthesis was specificfor the R-gene-mediated response, as it did not occur during the compatible interaction withvirulent P. syringae pv. glycinea. Since we infected Arabidopsis with avirulent P. syringae pv.tomato, the subsequent hypersensitive response probably led to a similar down-regulation ofphotosynthesis-related genes.These results demonstrate how statistical analysis and visualizationof changes in gene expression can contribute to the identification of the biological processesaffected.

84

Stress-related genes were not induced upon challenge with P. rapae according to MapMan

The usefulness of MapMan depends in large part on the accuracy and the completeness ofthe assignment of the genes to the proper BINs (Usadel et al., 2005). When MapMan wasdeveloped, the focus was on central metabolism (Thimm et al., 2004), which explains whygenes in metabolic processes are better organized into BINs than genes with a role in otherbiological functions. Here we found that, according to MapMan categorization, only threebiotic stress-related genes and no genes involved in wounding were activated upon feedingby P. rapae (Fig. 2). This contrasts with our previous findings using the Gene Ontology Toolof The Arabidopsis Information Resource that identified an overrepresentation of stress-relatedgenes upon caterpillar feeding (De Vos et al., 2005, Chapter 2). Probably, this is due toincomplete categorization in MapMan, as P. rapae infestation on Arabidopsis resulted in a similar gene expression profile in our experiments as found previously by Reymond et al.(2000; 2004). Moreover, in all Arabidopsis-attacker combinations many genes were categorizedinto the BIN “Not assigned” (Table 1).The improving quality of the annotations and the properassignment into functional categories in MapMan should lead to a more precise categorizationin more diverse biological groups in the near future.

Genes from the BIN “Large enzyme families” were activated ina predominantly JA-responsive fashion by all attackers, exceptaphids

Many genes encoding various types of enzymes were activated in the response of Arabidopsisduring pathogen or insect attack. Upon P. syringae infection, gene expression ratios from thesubBINs “Cytochrome P450s” and “Glutathione S-transferases” were significantly differentfrom those of genes in other subBINs in the BIN “Large enzyme families” (Table 2). Similarnumbers of genes in these subcategories were affected by M. persicae feeding.The other attackersalso affected genes in these subBINs (Fig. 3). Cytochrome P450 monooxygenases are a groupof heme-containing proteins that catalyze a large range of oxidative reactions (Chapple, 1998).Glutathione S-transferases are plant enzymes with a role in detoxification of compounds ofxenobiotic or endogenous origin (Marrs, 1996).They have been implicated in numerous stressresponses, including pathogen attack or oxidative stress. Both cytochrome P450s and glutathioneS-transferases play a role in the metabolism of secondary metabolites, such as phenylpropanoids,which are associated with plant defense against pathogens and insects.The genes encoding theseenzymes were typically up-regulated upon interaction with four attackers or upon applicationof MeJA, indicating that a common outcome of an Arabidopsis-attacker combination is enzymeactivation. However, aphid feeding led to down-regulation of genes from these subBINs,suggesting that aphids actively down-regulate plant defense responses.


85

Specific up-regulation of genes in several subBINs of the BIN“Large enzyme families”

P. rapae, F. occidentalis and A. brassicicola are three highly different attackers that elicit similarresponses in Arabidopsis in terms of phytohormone signal signature, numbers of genes affected(150 to 200), and functional groups of genes involved (“Stress”, “Large enzyme families”, and“Secondary metabolism”). Within the BIN “Large enzyme families”, these invaders mainlyhad an effect on genes from the subBINs “Cytochrome P450s”, “Glutathione S-transferases”and “UDP glycosyltransferases” (Fig. 3), but several differences became apparent upon MapManvisualization. For instance, P. rapae induced several copper- and flavin-containing amine oxi-dases, which produce hydrogen peroxide (H2O2) as a byproduct. H2O2 production is associatedwith cell wall maturation and lignification during normal development, as well as wound healingand cell wall strengthening during pathogen invasion (Cona et al., 2006).Therefore, activationof oxidase-encoding genes by tissue-chewing caterpillars may be related to the wound response.Several phenylpropanoid- and lignin biosynthesis-related genes were also activated as a resultof the wound response.

A. brassicicola infection altered the transcript behavior of the subgroup of “Nitrilases”.Of this subgroup, five berberine bridge-forming enzymes (BBE-like) involved in alkaloid synthesiswere up-regulated. Alkaloids are secondary metabolites many of which have been describedto function in plant defense against pathogens or herbivores (Facchini, 2001). The same BBE-like enzymes, including several others, were up-regulated upon P. syringae infection, suggestingactivation of alkaloid biosynthesis upon infection with these pathogens in Arabidopsis.Infection with A. brassicicola also resulted in up-regulation of a secreted lipase with a GDSL-motif (GLIP1), which has been identified to function in Arabidopsis defense against A. brassicicola(Oh et al., 2005). Mutant glip1 plants are more susceptible to A. brassicicola infection.Moreover, the recombinant GLIP1 protein directly disrupts the fungal spores. In our experiments,GLIP1 was not up-regulated by application of MeJA, confirming the results of Oh et al. (2005).

Genes in the BIN “Secondary metabolism” are predominantlyJA-responsive for all attackers, except aphids

Visualization of the transcriptome changes in the BIN “Secondary metabolism” revealed thateach interaction resulted in specific transcriptional alterations mainly in the same subBINs (Fig. 4).The differential activation of genes related to the metabolism of flavonols and anthocyaninsillustrated this nicely. All attackers, except aphids, induced a transcriptional response inArabidopsis that not only overlapped to a large extent with the response to MeJA in terms of genes affected, but also in terms of direction of induction (up- or down-regulation). Thishighlights the importance of JA in secondary metabolism during Arabidopsis defense againstthe majority of attackers. A good example is the subBIN “Glucosinolates”, because genesinvolved in glucosinolate metabolism were activated upon attack by P. rapae, F. occidentalisor P. syringae, or by application of MeJA. Glucosinolates and their hydrolysis products aredefensive compounds that are toxic to many generalist insects, although they serve as attactants

86

to specialist herbivorous insects (reviewed in Halkier and Gershenzon, 2006).Visualization of aphid-induced transcriptome changes revealed that, unlike the other

attackers, these insects down-regulate many genes involved in secondary metabolism in a JA-independent fashion (Fig. 4). Aphids have developed a highly specialized feeding strategy,which may explain this remarkable difference. In order to reach the phloem while minimallydamaging the plant tissue, aphids maneuver their stylets carefully around the cells throughthe middle lamellae (Tjallingii and Hogen Esch, 1993). Meanwhile, the aphid secretes a gellingsaliva thought to function in cell wall dissolution (Miles, 1999).Analysis of global gene expressiondata in MapMan revealed that genes related to phenylpropanoid and lignin biosynthesis weretypically down-regulated upon M. persicae infestation (Fig. 4). This suggests that in additionto cell wall-degrading enzymes of aphid origin, these phloem-feeding insects are capable ofmanipulating the plant in order to down-regulate lignin synthesis. In this way, they can ensureaccess of the stylets through mesophyll to the phloem. In addition, isoprenoids and flavonoidsare groups of secondary compounds with an antioxidant function. During oxidative stress, theyare typically produced. Since M. persicae infestation repressed genes related to isoprenoid andflavonoid synthesis in Arabidopsis, this further indicates that aphids actively suppress plantdefense responses.

Aphids specifically activate genes in the subBIN “Proteindegradation” of BIN 29 “Protein”

MapMan analysis of aphid-induced gene expression revealed that gene expression in the subBIN“Protein degradation” was significantly different from the expression of genes in the othersubBINs of BIN 29 “Protein” (Fig. 5).The subBIN “Protein degradation” includes genes encodingF-box proteins that are known to play an important regulatory role in ubiquitin-dependent,E3 ligase-dependent protein degradation.This suggests that aphid infestation promotes proteindegradation in Arabidopsis. Previous research showed that phloem sap is rich in sugars, butlow in essential amino acids.This poses a challenge to insects exclusively feeding on it (reviewedin Douglas, 2006). Indeed, aphid performance is correlated with nutrient quality: aphids performbetter on plants with more free amino acids. Since M. persicae feeds solely on phloem sap,it needs to compensate for the lack of essential amino acids. It has been shown that aphidscontain bacterial symbionts, Buchnera sp., that synthesize essential amino acids (reviewed inDouglas, 2006). In addition, aphid feeding can result in a nutritionally enhanced phloem content,benefiting aphid performance (Telang et al., 1999). Upon infestation by the Russian wheataphid, wheat plants increased the production of ten essential amino acids in the phloem sap(Telang et al., 1999). Here, we showed that M. persicae feeding on Arabidopsis led to theactivation of genes associated with protein degradation. Therefore, it is tempting to speculatethat aphids can manipulate their host to degrade proteins, leading to phloem sap enriched in free amino acids. It would be interesting to investigate this hypothesis by determining thechanges in free amino acids in phloem of uninfested versus M. persicae-infested Arabidopsisplants.


87

Aphids and the response to MeJA

Given the fact that aphids did not measurably stimulate the production of JA, ET and SA inArabidopsis (De Vos et al., 2005, Chapter 2), it was quite surprising to note that 30% of theM. persicae-responding genes was JA-responsive. However, the direction of the changes in aphid-induced gene expression (up- or down-regulated) was often opposite to the effect observed uponapplication of MeJA. Previous investigations have shown that M. persicae and Brevicorynebrassicae performance was affected by JA signaling in Arabidopsis (Ellis et al., 2002; Mewiset al., 2005). Both aphid species performed better on the Arabidopsis mutant coi1, which isinsensitive to JA (Ellis et al., 2002; Mewis et al., 2005). Conversely, M. persicae performedless well on cev1, a constitutive JA-signaling mutant (Ellis et al., 2002). Moreover, sprayingwith JA enhanced Arabidopsis defense against M. persicae both in the wildtype Col-0 andthe mutant cev1 (Ellis et al., 2002).Taken together, this suggests that aphids suppress specificJA-dependent defense responses to their own benefit.

Conclusion

The MapMan software greatly enhanced the analysis of genome-wide transcriptome changesin Arabidopsis during pathogen and insect attack. Due to visualization of the affected genesonto pictorial diagrams in combination with statistical analysis of the expression ratios of genesassigned to different BINs, we were able to identify and further analyze the most dominantlyaffected functional groups. Although Arabidopsis responded with unique transcriptional alter-ations towards each attacker, the most dominantly affected functional categories were similar:“Stress”, “Large enzyme families”, and “Secondary metabolism”.Also within these categories,the subcategories of affected genes were largely similar.This indicates that the different attackersinduced changes in similar plant processes through largely non-overlapping transcriptionalresponses. Yet, M. persicae infestation often induced an opposite transcriptional response inthe affected plant processes, when compared to the other Arabidopsis-attacker combinationsor application of MeJA.

Materials and methodsGene expression analysis with 22 K Affymetrix GeneChips

Plant treatments and transcriptome profilingPreviously, RNA was prepared from Arabidopsis plants during microbial or insect attack andused for genome-wide transcription profiling (De Vos et al., 2005, Chapter 2). Briefly,Arabidopsisplants were treated with an avirulent strain of the bacterial leaf pathogen Pseudomonas syringaepv. tomato DC3000 (avrRpt2), the fungal leaf pathogen Alternaria brassicicola MUCL 20297,tissue-chewing larvae of the small cabbage white butterfly Pieris rapae, piercing-sucking larvae

88

of the Western flower thrips Frankliniella occidentalis and phloem-feeding green peach aphidMyzus persicae. Infection with A. brassicicola was performed on the phytoalexin-deficient Col-0mutant pad3-1 (Glazebrook and Ausubel, 1994), whereas all other plant treatments were carriedout on wild type Arabidopsis Col-0.Whole plant rosettes were harvested at intervals duringinfection/infestation, and used for RNA isolation. The harvest times were 12 and 24 h for P. syringae, P. rapae and F. occidentalis, 24 and 48 h for A. brassicicola and 48 and 72 h forM. persicae. To monitor transcriptome changes, 22k ATH1 Affymetrix GeneChip analyses(Affymetrix, Santa Clara, CA, U.S.A.) were performed with RNA from each time point.Robust sets of attacker-responsive genes with a >2-fold change in expression ratio in the samedirection at both time points were selected. In addition, a dipping treatment with 0.05 mM MeJAwas performed. The time points 1, 3 and 6 h after dipping were monitored on GeneChips.Genes were considered to be MeJA-responsive when their transcripts had a >2-fold changein expression at two of the three time points.The selection of genes was published previously(De Vos et al., 2005, Chapter 2, Supplementary Tables 1-6)(http://www.bio.uu.nl/~fytopath/GeneChip_data.htm). Raw data of the GeneChip hybridizationscan be obtained via The Arabidopsis Information Resource (TAIR): www.arabidopsis.orgNottingham Arabidopsis Stock Centre International Affymetrix Service (NASCarray:affymetrix.Arabidopsis.info).

Categorization and visualization of selected transcriptome changesHere, the selected genes were classified into functional categories and the expression ratiosvisualized using MapMan software (version 1.7.0). The program was obtained fromhttp://gabi.rzpd.de/projects/MapMan/ (Thimm et al., 2004; Usadel et al., 2005). All 22 KAffymetrix GeneChip probe set identifiers were loaded for the five attackers. The expressionratios of the selected genes were averaged, log2 transformed, and imported in the Transcript-Scavenger module. This module classifies the genes represented on the Arabidopsis ATH1full-genome GeneChips into hierarchical functional categories (BINs, subBINs, individualenzymes, etc.).The MapMan files created with the TranscriptScavenger Module were displayedonto pictorial diagrams through the ImageAnnotator Module. The genes appear in blue (up-regulation) or red (down-regulation).All other genes on the GeneChip were excluded from thestatistical analysis, and visualized in grey. Some MeJA-responsive genes may appear in white(no change). This is caused by the fact that the expression ratios of all three time points wereaveraged for MapMan analysis, while the genes had been selected for a >2-fold change inexpression ratio at only two out of the three time points.

Statistical analysis of categorized gene setsCoordinate gene expression was analyzed with the built-in Wilcoxon Rank Sum test withBenjamini-Hochberg P-value correction (Usadel et al., 2005). When a BIN is significantly different from the other BINs, this indicates that the combined response of the genes assignedto this BIN is significantly different compared to the response of the genes in all other BINs.The highest probability is that the genes in one BIN respond similarly as the other genes inall other BINs (e.g. as many up- as down-regulated), whereas the lowest probability is that all genes in one BIN respond completely opposite from the genes in the other BINs (e.g. all


89

up- or all down-regulated). Each significantly different BIN (P < 0.05) is likely to contain co-responsive genes (Usadel et al., 2005). Genes responding in a coordinate fashion may encodeproteins that are involved in the same pathway, are members of the same complex or areotherwise functionally connected (Lee et al., 2004). Thus, the statistical module of MapManhelps to identify responses that affect a large proportion of the genes assigned to a particularfunctional category.

90

Supplementary MaterialsSupplementary Table 1: MS Excel file containing log2-transformed average expression valuesof Arabidopsis genes responding to attack by P. syringae, A. brassicicola, P. rapae, F. occidentalisor M. persicae.

Affymetrix probe identifiers are shown for all Arabidopsis genes present on the 22 K AffymetrixGeneChip. The log2-transformed average expression values of consistent changes on two timepoints are shown for previously selected genes (De Vos et al., 2005, Chapter 2).All other genesare marked x.

Genes had been selected for a consistent >2-fold change (up or down) 12 and 24 h afterinoculation of Arabidopsis Col-0 plants with the bacterial leaf pathogen P. syringae pv. tomatoDC3000 (avrRpt2), 24 and 48 h after inoculation of Arabidopsis pad3-1 plants with the fungalleaf pathogen A. brassicicola, 12 and 24 h after infestation of Arabidopsis Col-0 plants withtissue-chewing caterpillars of the small cabbage white butterfly P. rapae, 12 and 24 h afterinfestation of Arabidopsis Col-0 plants with larvae of the cell-content feeding Western flowerthrips F. occidentalis, or 48 and 72 h after infestation of Arabidopsis Col-0 plants with phloem-feeding M. persicae aphids.

Supplementary Table 2: MS Excel file containing log2-transformed average expression valuesof a selection of MeJA-responsive Arabidopsis genes responding to attack by P. syringae,A. brassicicola, P. rapae, F. occidentalis or M. persicae, and all genes responding to applicationof MeJA.

Affymetrix probe identifiers are shown for all Arabidopsis genes present on the 22 K AffymetrixGeneChip. The log2-transformed average expression values of consistent changes on two (fiveattackers) or three (MeJA application) time points are shown for previously selected genes(De Vos et al., 2005, Chapter 2). Only the expression values of genes also responsive to MeJAare shown. All other genes are marked x.

Genes had been selected for a consistent >2-fold change (up or down) 12 and 24 h afterinoculation of Arabidopsis Col-0 plants with the bacterial leaf pathogen P. syringae pv. tomatoDC3000 (avrRpt2), 24 and 48 h after inoculation of Arabidopsis pad3-1 plants with the fungalleaf pathogen A. brassicicola, 12 and 24 h after infestation of Arabidopsis Col-0 plants withtissue-chewing caterpillars of the small cabbage white butterfly P. rapae, 12 and 24 h afterinfestation of Arabidopsis Col-0 plants with larvae of the cell-content feeding Western flowerthrips F. occidentalis, 48 and 72 h after infestation of Arabidopsis Col-0 plants with phloem-feeding M. persicae aphids, or a consistent >2-fold change (up or down) on at least two timepoints after treatment at 1, 3, or 6 h of Arabidopsis Col-0 plants with 0.05 mM MeJA.



91

92

Differential effectiveness of microbially induced resistanceagainst herbivorous insects in Arabidopsis

Vivian R. Van Oosten1,2, Johan A. Van Pelt2, L. C. Van Loon2, Corné M. J. Pieterse2 and Marcel Dicke1




Biology, Faculty of Science, Utrecht University, P.O.Box 800.84, 3508 TB Utrecht, The Netherlands

93Differential effectiveness of microbiallyinduced resistance against herbivorousinsects in Arabidopsis

chapter 4

AB

ST

RA

CT Plants possess inducible defense mechanisms to protect themselves against different types of micro-

bial pathogens and herbivorous insects. Two types of microbially induced resistance have been well

characterized: systemic acquired resistance (SAR), which is induced upon predisposal infection by

necrotizing pathogens, and rhizobacteria-mediated induced systemic resistance (ISR), which is triggered

by selected strains of non-pathogenic, root-colonizing rhizobacteria. Both types of induced resistance

are expressed systemically throughout the plant and are effective against a broad, yet partly distinct,

range of pathogens. Because the signaling pathways that control induced resistance against pathogens

and insects partly overlap, we decided to investigate the effectiveness of ISR and SAR against the tissue-

chewing herbivorous insects Pieris rapae and Spodoptera exigua in Arabidopsis. Insect-induced resistance

consists of two components: direct defense, such as the production of toxins and feeding deterrents,

and indirect defense, such as the production of plant volatiles that attract carnivorous enemies of the

herbivores. Feeding experiments revealed that induction of ISR and SAR significantly reduced growth

and development of the generalist herbivore S. exigua, whereas the performance of the specialist

P. rapae was unaffected. The jasmonic acid- and ethylene-responsive genes PDF1.2 and HEL, which

were activated upon feeding by either of the two herbivores, showed a strong potentiated expression

pattern in ISR- and SAR-expressing plants upon feeding by S. exigua, but not upon feeding by P. rapae.

These results suggest that the effectiveness of microbially induced ISR and SAR against S. exigua

feeding is associated with priming for enhanced defense-related gene expression. In a wind-tunnel

set-up, we tested the effect of ISR and SAR on herbivore-induced attraction of the parasitic wasp

Cotesia rubecula. Upon feeding by either P. rapae or S. exigua, Arabidopsis Col-0 plants were signi-

ficantly more attractive to the parasitic wasps than undamaged control plants. However, induction of

either ISR or SAR did not significantly enhance the attractiveness of insect-damaged plants. Together,

these results indicate that microbially induced resistance differentially affected direct defenses that were

triggered by the specialist P. rapae and the generalist S. exigua, but had no effect on indirect defenses

that were triggered by these insect herbivores.

94

IntroductionPlants are under constant threat of a multitude of pathogenic and herbivorous attackers.Resistance against microbes and insects can be mediated through defenses that are constitutivelypresent, or through defense mechanisms that are induced only upon attack (Van Loon, 2000;Dicke and Van Poecke, 2002). An important question in plant defense signaling research is:how are plants capable of integrating signals induced by pathogenic micro-organisms and herbivorous insects into defenses that are specifically active against the invader encountered?The plant hormones salicylic acid (SA), jasmonic acid (JA), and ethylene (ET) are major playersin the regulation of signaling networks involved in induced defense (Reymond and Farmer,1998; Pieterse and Van Loon, 1999; Feys and Parker, 2000; Glazebrook, 2001; Thomma et al.,2001; Dicke and Van Poecke, 2002; Kessler and Baldwin, 2002; Spoel et al., 2003). The pro-duction of SA, JA, and ET varies greatly depending on the nature of the attacking pathogenor insect.The quantity, composition and timing of the hormonal blend results in the activationof a specific set of genes that eventually determines the nature of the defense response that is triggered by the attacker encountered (De Vos et al., 2005, Chapter 2; Mur et al., 2006).Although there are exceptions (Thaler et al., 2004), in general it can be stated that pathogenswith a biotrophic life style are more sensitive to SA-dependent defense responses, whereasnecrotrophic pathogens and herbivorous insects are primarily resisted by JA/ET-dependentdefenses (Thomma et al., 2001; Dicke and Van Poecke, 2002; Glazebrook, 2005).

Induced defense against insect herbivores is triggered upon feeding and consists of twocomponents: direct defense such as the production of secondary chemicals that act as toxinsor feeding deterrents (Kessler and Baldwin, 2002; Howe, 2004), and indirect defense such asthe production of a blend of volatiles that attract carnivorous enemies of the herbivores (Turlingset al., 1995; Takabayashi and Dicke, 1996). Induced direct defense has been demonstrated inmany plant species (Kessler and Baldwin, 2002). A classic example is the observation thatfollowing herbivore attack, tomato leaves systemically accumulate proteinase inhibitor proteinsthat reduce further insect feeding (Howe, 2004). Both JA and ET emerged as important signalsin this response (Farmer and Ryan, 1992; O'Donnell et al., 1996). JA is also a major phyto-hormone involved in the induced production of plant volatiles that attract carnivorous enemiesof the herbivores (Dicke et al., 1999; Gols et al., 1999; Dicke and Van Poecke, 2002). SA hasbeen implicated in induced indirect defense against herbivory as well. Herbivores such as spidermites induce the emission of methyl salicylate (MeSA) in many plant species (Dicke et al.,1990; Takabayashi and Dicke, 1996; Ament et al., 2004; De Boer and Dicke, 2004), which canlead to the activation of SA-inducible defense-related genes (Arimura et al., 2000; Kant et al.,2004). Moreover, certain combinations of JA and SA treatments induce a blend of volatilesthat is similar to the blend induced by spider mite feeding (Dicke et al., 1999; Arimura et al.,2000; De Boer and Dicke, 2004) and attracts carnivorous enemies that can exterminate theherbivore population (Dicke, 1999; Dicke et al., 1999).

In Arabidopsis, induced defense against insect herbivores is very similar to that shown in other plant species. For instance, caterpillars of the specialist herbivore Pieris rapae (smallcabbage white butterfly) stimulate the production of JA and ET and trigger a systemic defenseresponse that affects insect performance (De Vos et al., 2005; 2006). Moreover, Arabidopsis

Chapter 04: Differential effectiveness of microbially induced resistance against herbivorous insects in Arabidopsis

95

mutants affected in the JA response are more susceptible to insect feeding (McConn et al., 1997;Stintzi et al., 2001; Ellis et al., 2002; Stotz et al., 2002; Reymond et al., 2004; Van Poeckeand Dicke, 2004), indicating that JA plays an important role in induced direct defense.Furthermore,Arabidopsis plants infested by caterpillars of P. rapae emit volatiles from severalmajor biosynthetic pathways, including terpenoids, MeSA and green leaf volatiles. This blendof volatiles attracts adult females of Cotesia rubecula, a specialist parasitic wasp of P. rapae,resulting in a fitness increase of the Arabidopsis plant (Van Loon et al., 2000; Van Poecke et al.,2001; Van Poecke and Dicke, 2004). Transgenic Arabidopsis S-12 plants with severely reducedwound-inducible JA levels (Bell et al., 1995), and transgenic Arabidopsis NahG plants, whichcannot accumulate SA (Delaney et al., 1994), are significantly less attractive to the C. rubeculawasps upon herbivory by P. rapae than wild-type Col-0 plants (Van Poecke and Dicke, 2002).This indicates that both JA and SA are involved in indirect defense in Arabidopsis.

Against microbial pathogens, two forms of induced resistance have been well-characterized:systemic acquired resistance (SAR), which is typically activated upon primary, limited infectionby a necrotizing pathogen (Durrant and Dong, 2004), and induced systemic resistance (ISR),which is typically elicited by specific strains of nonpathogenic, root-colonizing rhizobacteria(Van Loon et al., 1998). SAR is controlled by a signaling pathway that depends on endogenousaccumulation of SA and the defense regulatory protein NPR1 (Durrant and Dong, 2004), andis associated with the activation of pathogenesis-related (PR) genes, some of which encodeproteins with anti-microbial activity (Van Loon et al., 2006). ISR triggered by the nonpathogenicrhizobacterium Pseudomonas fluorescensWCS417r functions independently of SA but requiresresponsiveness to JA and ET (Pieterse et al., 1996; 1998). In contrast to SAR, the onset ofISR is not associated with enhanced defense-related gene expression (Verhagen et al., 2004).However, after challenge inoculation, a large set of predominantly JA-responsive genes show a potentiated expression pattern in ISR-expressing leaves, indicating that these genes are sensitized to respond faster and/or more strongly upon pathogen attack. This phenomenonis called “priming” and has been demonstrated to play a role in different types of inducedresistance, including SAR (Conrath et al., 2002; 2006). Both ISR and SAR are effective againsta broad range of pathogens (Van Loon et al., 1998; Durrant and Dong, 2004). However, theirranges of effectiveness overlap only partly. Pathogen-induced SAR is predominantly effectiveagainst biotrophic pathogens, whereas rhizobacteria-mediated ISR is predominantly effectiveagainst pathogens that are sensitive to JA- and ET-dependent defenses (Ton et al., 2002; Verhagenet al., 2006).

The common involvement of SA, JA and ET in the regulation of the induced defenseresponses that are triggered by microbes and insects suggests that the effectiveness of therespective induced resistance responses will overlap, at least partly. However, several studieshave shown that activation of SA-dependent SAR negatively affects resistance against specificinsects (reviewed in Felton and Korth, 2000; Pieterse et al., 2001; Bostock, 2005), which isprobably caused by the antagonistic effect of SA on JA signaling (Spoel et al., 2003). By contrast, rhizobacteria-mediated ISR in cucumber has been demonstrated to simultaneous-ly reduce bacterial wilt disease, caused by Erwinia tracheiphila, and feeding of the cucumberbeetle vector that transmits this disease (Zehnder et al., 2001). To investigate the extent ofoverlap in effectiveness of induced resistance that is triggered by micro-organisms and insects,

96

we previously studied the effectiveness of herbivore-induced resistance in Arabidopsis againstthe microbial pathogens Alternaria brassicicola, Pseudomonas syringae pv. tomato,Xanthomonas campestris pv. armoraciae, and turnip crinkle virus (TCV) (De Vos et al., 2006).Here we report on a reciprocal study in which we investigated the effect of microbially inducedresistance (ISR and SAR) in Arabidopsis on direct and indirect defense against two differentchewing insects: P. rapae, a specialist on crucifers, and the generalist Spodoptera exigua (beetarmyworm).

ResultsPerformance of P. rapae and S. exigua on Arabidopsis genotypes affected in JA or SA signaling

In Arabidopsis, ISR and SAR are regulated by JA- and SA-dependent signaling pathways,respectively (Pieterse et al., 2002; Durrant and Dong, 2004). JA- and SA-dependent defenseshave been demonstrated to antagonistically affect insect performance (Pieterse et al., 2001;Dicke and Van Poecke, 2002). To verify whether the performance of the specialist herbivoreP. rapae and the generalist herbivore S. exigua are affected by JA and/or SA signaling, wemonitored the growth of their larvae on Arabidopsis genotypes S-12 and NahG, which areimpaired in the production of JA or SA, respectively. To assess the effect of JA on directdefenses against P. rapae and S. exigua, first-instar (L1) larvae were placed onto individuallyconfined S-12 plants, after which their weight was monitored until approximately two daysbefore the first larva started to pupate. Figure 1A shows that the fresh weights of the larvaewere significantly higher on S-12 plants than on wild-type Col-0 plants. However, S-12 plantslack trichomes because the transgene is in the background of glabrous Col-5, which carries a mutation in the GLABROUS1 (GL1) gene (Oppenheimer et al., 1991). Comparison of growthof S. exigua larvae on Col-0 and glabrous Col-5 (gl1) showed that the gl1 mutation had noeffect on S. exigua performance, indicating that compromised accumulation of JA in S-12 (gl1)plants accounted for the increased growth of S. exigua on this genotype. By contrast, P. rapaelarvae gained significantly more weight on glabrous Col-5 than on Col-0, indicating that tri-chomes had a clear negative effect on the growth of P. rapae larvae, confirming previous findings(Reymond et al., 2004). To further verify the role of JA in direct defense against P. rapae, wemonitored growth of P. rapae larvae on glabrous Col-5 and the glabrous JA signaling mutantcoi1-16 (Ellis and Turner, 2002) (Fig. 1B). In agreement with previous findings (Reymond et al.,2004), P. rapae larvae gained significantly more weight on coi1-16, when compared to glabrousCol-5 plants. Together, these results indicate that JA-dependent defenses negatively affect thegrowth of both P. rapae and S. exigua larvae.

To assess the effect of SA on direct defenses against P. rapae and S. exigua, larvalperformance was monitored on Col-0 and SA-non-accumulating NahG plants. Figure 1Cshows that growth of P. rapae larvae was not affected by the nahG transgene. By contrast,growth of S. exigua larvae was significantly reduced on NahG plants, indicating that the


97

absence of SA in NahG negatively affected growth of the generalist, but not the specialistherbivore.

FIGURE 1. Performance of P. rapae and S. exigua on Arabidopsis genotypes affected in JA or SA signaling.

First-instar larvae were allowed to feed on single plants confined in Magenta vessels with an insect-proof mesh lid.A. Relative weight of P. rapae larvae after 7 days (1 = 101 mg) and S. exigua larvae after 9 days (1 = 22.3 mg) offeeding on Col-0, Col-5 (gl1) and S-12 (gl1) plants. B. Fold change in weight of P. rapae larvae after 48 h of feedingon Col-5 (gl1) and coi1-16 (gl1) (1 = 128 mg). The starting weight of the larvae was about 40 mg. C. Relative weightof P. rapae larvae after 7 days (1 = 101 mg) and S. exigua larvae after 10 days (1 = 34.0 mg) of feeding on Col-0 andNahG plants. The data shown are means ±SE (n = 14-25) of the weights of the caterpillars 2 d before the first larvastarted to pupate. Asterisks indicate statistically significant differences (P < 0.05) according to Kruskal-Wallis followedby Mann-Whitney U tests in SPSS 11.5 for Windows. The number of caterpillars weighed (n) is given in the bar.Experiments were performed at least twice with similar results.

Effect of ISR and SAR on the performance of P. rapae and S. exigua

ISR and SAR are effective against a wide range of pathogens. In Arabidopsis, it has been shownthat ISR is predominantly effective against pathogens that are sensitive to JA/ET-dependentdefenses, whereas SAR is mainly effective against pathogens that are sensitive to SA-dependentdefenses (Ton et al., 2002). Because both P. rapae and S. exigua can be affected by JA-dependentdefense responses, we hypothesized that induction of rhizobacteria-mediated ISR would enhancethe level of resistance against both herbivores. Because SA has been shown to antagonize JAsignaling (Spoel et al., 2003), we hypothesized that induction of SAR would decrease the basallevel of insect resistance.

To investigate the effect of ISR and SAR on P. rapae and S. exigua performance, L1 larvaewere placed on individually confined ISR- and SAR-expressing plants and monitored untilpupation. ISR was induced by growing plants in soil with ISR-inducing Pseudomonas fluorescensWCS417r bacteria. SAR was induced by infiltrating three lower leaves with avirulentPseudomonas syringae pv. tomato DC3000 (avrRpt2) 3 days prior to transfer of the caterpillarsto the leaves. Figure 2A shows that feeding by P. rapae caterpillars on either uninduced, ISR-

98

or SAR-expressing plants did not result in significant differences in the average weight of thelarvae. Moreover, neither ISR nor SAR had an effect on the average time to pupation (Fig. 2B).By contrast, both ISR and SAR significantly reduced the weight gain of S. exigua larvae, resultingin a 28-32% lower weight of the larvae two days before the first larvae started to pupate.Reduced growth of S. exigua larvae on ISR- and SAR-expressing plants coincided with a prolonged time until pupation, although this effect was only significant for the larvae feedingon ISR-expressing Arabidopsis (Fig. 2B; ISR, P = 0.018; SAR, P = 0.096).These results indicatethat microbially induced ISR and SAR are associated with enhanced resistance against feedingby the generalist herbivore S. exigua, but not against feeding by the specialist herbivore P. rapae.

FIGURE 2. Performance of P. rapae and S. exigua on Arabidopsis Col-0 plants expressing microbially induced ISRor SAR.

First-instar larvae were placed on individually confined plants that were either uninduced (Ctrl) or expressing ISR orSAR, after which they were monitored until pupation. ISR was induced by growing the plants in soil containing ISR-inducing P. fluorescens WCS417r bacteria. SAR was induced by pre-infecting three leaves per plant with avirulent P. syringae pv. tomato DC3000 (avrRpt2). A. Relative weight of P. rapae larvae at day 7 (1 = 64.2 mg) and S. exiguaat day 12 (1 = 96.0 mg). The data shown are means ±SE of the weights of the caterpillars 2 d before the first larvastarted to pupate. B. Average number of days by which the larvae started to pupate (±SE). Asterisks indicate statis-tically significant differences (P < 0.05) based on Kruskal-Wallis followed by Mann-Whitney U tests (A) and One-wayANOVA followed by LSD post hoc test (B) in SPSS 11.5 for Windows. The number of caterpillars weighed (n) is givenin the bar. Each experiment was performed at least twice with similar results.

ISR and SAR are associated with potentiated expression ofPDF1.2 and HEL upon herbivory

As mentioned above, both ISR and SAR are associated with priming for enhanced defense-relatedgene expression, resulting in a potentiated activation of specific gene sets upon pathogenchallenge (Verhagen et al., 2004; Conrath et al., 2006). Previously, herbivory by the specialistP. rapae and the generalist Spodoptera littoralis (Egyptian cotton worm) has been demonstratedto activate similar sets of predominantly JA-responsive genes in Arabidopsis (Reymond et al.,2004).To investigate whether priming for enhanced herbivore-induced gene expression is asso-ciated with the reduced performance of S. exigua on ISR- and SAR-expressing plants, we analyzed


99

the expression of the JA/ET-responsive genes PDF1.2 and HEL in uninduced control plants,and in ISR- and SAR-expressing plants at different time points after infestation with P. rapaeor S. exigua (Fig. 3). In uninduced control plants, herbivory by P. rapae or S. exigua resultedin a low, but detectable increase in PDF1.2 transcript levels 72 h after infestation, whereasHEL mRNA levels remained virtually unchanged. In comparison to uninduced control plants,expression of either ISR or SAR had no clear effect on the P. rapae-induced levels of PDF1.2and HEL mRNA. In contrast, in response to S. exigua feeding, expression of ISR and SARresulted in a strong potentiated level of expression of PDF1.2 and HEL.These results indicatethat ISR- and SAR-expressing plants are primed to express PDF1.2 and HEL to a much higherlevel upon herbivory by the generalist S. exigua, but not upon feeding by the specialist P. rapae.

FIGURE 3. ISR- and SAR-mediated priming of herbivore-induced PDF1.2 and HEL gene expression.

RNA-blot analysis of herbivore-induced PDF1.2 and HEL transcript levels in Arabidopsis Col-0 plants that were uninduced(Ctrl) or expressing microbially induced ISR or SAR. Leaf tissues were harvested from uninfested plants and frominfested plants after 0, 24, or 72 h of feeding by first-instar larvae of P. rapae or S. exigua. Equal loading of RNAsamples was checked using a probe for 18S rRNA. h.p.i., hours post infestation. The experiment was repeated withsimilar results.

100

Effect of ISR and SAR on herbivore-induced indirect defense

Volatiles that are produced upon insect feeding play an important role in the indirect defenseof plants through the attraction of carnivorous enemies that attack the herbivores.The tritrophicinteraction between Arabidopsis, the insect herbivores P. rapae and S. exigua, and the parasiticwasp C. rubecula is an established model to study herbivore-induced indirect defense (VanPoecke et al., 2001; 2003). Using transgenic Arabidopsis S-12 and NahG plants, both JAand SA have been shown to play a role in the herbivore-induced attraction of the parasiticwasp C. rubecula (Van Poecke and Dicke, 2002). Interestingly, herbivore-induced volatiles havebeen implicated in priming for enhanced defense (Engelberth et al., 2004; Kessler et al.,2006; Paschold et al., 2006; Ton et al., 2007). Because of the role of priming in ISR andSAR, we hypothesized that both types of microbially induced defense would affect the indirectdefense response that is triggered upon insect herbivory.

To investigate the effect of ISR and SAR on the indirect defense response of Arabidopsis,we monitored the attraction of C. rubecula to ISR- and SAR-expressing plants upon feedingby P. rapae or S. exigua. To this end, L1 larvae of P. rapae or S. exigua were allowed to feedfor 24 h (P. rapae) or 72 h (S. exigua) on control, ISR-, or SAR-expressing plants.Quantification of the damage caused by either P. rapae or S. exigua during the short feedingperiod, revealed that the extent of damage did not significantly differ between uninduced, ISR-,and SAR-expressing plants (data not shown). Naive female C. rubecula wasps were individuallyreleased in a wind-tunnel set-up (Van Poecke et al., 2001) and given the choice between twoodor sources in different combinations. Odor sources comprised control, ISR-, or SAR-expressingplants, with or without caterpillar damage. Subsequently, the flight behavior of the wasps wasobserved and the number of first landings on one of the odor sources recorded. First landings ofthe wasps on other parts of the wind-tunnel were recorded as “no choice”. The number offirst landings on a given odor source reflects the degree of attractiveness to the parasitic wasp.Control, ISR- and SAR-expressing Col-0 plants without caterpillar damage were equally un-attractive to C. rubecula, as there were relatively few wasps (<30%) that responded to theoffered plants (data not shown).The two-choice tests in which herbivore-damaged plants wereoffered resulted in a much greater response of the C. rubecula wasps (57-79%). Uninducedcontrol plants infested with either P. rapae or S. exigua attracted significantly more parasiticwasps than undamaged control plants (Fig. 4, Ctrl vs. Ctrl + P.r. and Ctrl vs. Ctrl + S.e.), indicatingthat herbivory by P. rapae or S. exigua larvae resulted in enhanced attractiveness of the plantto the wasps, confirming previous findings (Van Poecke et al., 2003).When caterpillar-damagedcontrol plants were compared with equally damaged ISR- or SAR-expressing plants, no statis-tically significant differences were observed in any of the combinations. Similarly, the totalnumber of parasitoid wasps responding to any of the two odor combinations was not significantlydifferent among experiments.These results indicate that neither ISR, nor SAR significantly affectthe indirect defense response of Arabidopsis that is triggered upon feeding by either the specialistP. rapae or the generalist S. exigua.


101

FIGURE 4. Responsiveness of the parasitic wasp C. rubecula to herbivore-infested ISR- and SAR-expressingArabidopsis plants in a two-choice flight set-up.

Individual female C. rubecula wasps were offered two odor sources in a wind-tunnel set-up. Each odor source consistedof 15 Arabidopsis plants, that were either uninduced (Ctrl), or expressing microbially induced ISR or SAR, with or withoutfeeding damage caused by P. rapae (P.r.) or S. exigua (S.e.). The number of landings on each odor source was recordedas “choice”. When the parasitoid wasp did not land on an odor source, this was noted as “no-choice”. The figure showsthe percentage of responding and non-responding wasps. Differences in non-responsiveness between the odor combi-nations were tested using a contingency table. Similar letters indicate that there were no statistically significant differences(P > 0.05). The differences in responsiveness per odor combination were analyzed with a Chi-square test in SPSS 11.5for Windows. Asterisks indicate statistically significant differences (P < 0.001); NS, non-significant differences (P > 0.05);n, number of wasps tested.

DiscussionThe signaling events that occur between perception of microbial or insect-derived elicitors, andthe subsequent activation of defense responses are relatively well-studied, especially for well-defined types of induced resistance such as pathogen-induced SAR, rhizobacteria-mediatedISR, the wound response and the response as triggered upon insect herbivory (Dicke and VanPoecke, 2002; Pieterse et al., 2002; Durrant and Dong, 2004; Howe, 2004). However, little isknown about the spectrum of effectiveness of the different types of induced resistance that issubsequently expressed. Is herbivore-induced resistance specifically effective against insectherbivores, or does it also affect pathogens? Is microbially induced resistance specifically directedagainst microbial pathogens, or are insect herbivores affected as well? In the past, a numberof studies investigated the effect of microbially induced defense on insects and vice versa(reviewed in Stout et al., 2006). However, due to the large variety in the choice of plant, insectand pathogen species, the results varied and it was difficult to draw general conclusions aboutthe response of one plant against multiple attackers.Therefore, we decided to address the above-mentioned questions in a single plant species, Arabidopsis thaliana. To gain insight in thecomplexity of the defense signaling network of Arabidopsis, we previously monitored the

102

signal signature and whole-genome transcript profile of Arabidopsis upon attack by pathogensand insects with different modes of action (De Vos et al., 2005, Chapter 2). In addition, weinvestigated the effectiveness of herbivore-induced resistance in Arabidopsis against the microbialpathogens A. brassicicola, P. syringae pv. tomato, X. campestris pv. armoraciae, and TCV(De Vos et al., 2006). Here, we performed a reciprocal study in which we investigated the effectof two types of microbially induced resistance (ISR and SAR) on direct and indirect defenseagainst the insect herbivores P. rapae and S. exigua.We demonstrate that ISR and SAR haveno effect on direct defense against feeding by the specialist P. rapae. However, both ISR andSAR negatively affect growth and development of S. exigua, and this effect on direct defenseagainst this generalist herbivore is associated with priming for enhanced expression of the S. exigua-induced genes PDF1.2 and HEL. In addition, we show that ISR and SAR haveno effect on the indirect defense response that is triggered upon herbivory by either P. rapaeor S. exigua.

Direct insect defense

Analysis of insect performance on Arabidopsis mutants and transgenics that are affected inJA signaling revealed an important role for JA in direct insect defense in this plant species.Herbivorous insects such as P. rapae, S. exigua, S. littoralis, Myzus persicae (green peach aphid),and Bradysia impatiens (common fungus gnat) were shown to be negatively affected by JA-dependent defenses (McConn et al., 1997; Stintzi et al., 2001; Ellis et al., 2002; Stotz et al.,2002; Reymond et al., 2004; Mewis et al., 2005). In this study, we demonstrated that P. rapaeand S. exigua larvae perform better on S-12 or coi1-16 plants, confirming that JA-dependentdefense responses contribute to direct defense against these caterpillars.

Arabidopsis mutants and transgenics that are compromised in SA-dependent defenseresponses have been shown to exhibit enhanced resistance against feeding by the cabbage looperTrichoplusia ni (Cui et al., 2002), the Egyptian cotton worm S. littoralis (Stotz et al., 2002)and the beet armyworm S. exigua (Cipollini et al., 2004; Mewis et al., 2005). Similarly, wedemonstrated here that S. exigua performance is significantly reduced on SA-non-accumulatingNahG plants (Fig. 1), indicating that in wild-type Arabidopsis plants SA signaling promotessusceptibility to insect feeding. In many studies SA signaling has been demonstrated to actantagonistically on JA-dependent signaling (Pieterse et al., 2001; Dicke and Van Poecke, 2002;Spoel et al., 2003).Therefore, it is plausible that the observed enhanced resistance to S. exiguafeeding in NahG plants is caused by the fact that the antagonistic effect of SA on JA signalingis relieved in this genotype.

In contrast to S. exigua, P. rapae larvae performed similarly on wild-type Col-0 and NahGplants. Apparently, the enhanced JA response in NahG plants had no effect on the level ofresistance against this herbivore.This is supported by the fact that glabrous S-12 plants displayeda similar level of resistance to P. rapae feeding as glabrous Col-5 plants. Only when JA signalingwas completely blocked, as in coi1-16, an effect on P. rapae performance could be detected(Fig. 1). These results indicate that P. rapae is less sensitive to JA-dependent defenses than S. exigua. P. rapae is a specialist herbivore which has been reported to feed on many


103

Brassicaceous species in the field, including Arabidopsis (Yano and Ohsaki, 1993), and to bewell adapted to the induced defenses of crucifers. In wild radish for instance, growth of P. rapaelarvae was not affected by defenses induced by insect-damage and JA application, while thegeneralist S. exigua was negatively affected (Agrawal, 1999). In line with this observation,P. rapae was demonstrated to detoxify deleterious Arabidopsis-derived glucosinolate compoundsin its gut (Wittstock et al., 2004), further illustrating that specialist insects have adapted to theresistance mechanisms of their host plants.Although it was shown that P. rapae is not insensitiveto all induced defenses in Arabidopsis (Agrawal and Kurashige, 2003; De Vos et al., 2006), thedifference in sensitivity to JA-mediated defenses between P. rapae and S. exigua seems to bedue to a better adaptation of P. rapae to Brassicaceous plants.

ISR, SAR and direct insect defense

In view of the effects of JA and SA signaling on direct defense against herbivory, we expectedthat both ISR and SAR would affect the level of resistance against insect herbivores. Only alimited number of studies investigated the effect of rhizobacteria-mediated ISR against insects.For example, Zehnder and coworkers (2001) demonstrated that ISR triggered by non-pathogenicrhizobacterial strains resulted not only in resistance against bacterial wilt caused by the pathogenE. tracheiphila, but also reduced feeding by the cucumber beetles Diabrotica undecimpunctataand Acalymma vittata that are vectors of this pathogen. ISR against cucumber beetle feedingwas associated with reduced concentrations of cucurbitacin, a secondary plant metabolite andpowerful beetle-feeding stimulant.A number of studies investigated the effect of SAR triggeredby avirulent pathogens on the performance of herbivorous insects. Depending on the plant-insect combination, the effect of SAR on insect performance was either positive, negative orneutral, indicating that the effect of SAR is not easy to predict (reviewed in Stout et al., 2006).

In this study, we demonstrated that neither ISR nor SAR had a significant effect on thelevel of resistance of Arabidopsis against P. rapae feeding (Fig. 2).These findings are in line withour observation that specialist P. rapae larvae are relatively insensitive to JA-dependent defensesand, therefore, also to the effect that SA exerts on JA signaling. Surprisingly, both ISR and SARsignificantly reduced growth of S. exigua (Fig. 2A). Moreover, ISR also significantly delayedthe onset of pupation (Fig. 2B). For ISR, the negative effect on S. exigua larval developmentwas to be expected, because this type of induced resistance is effective against attackers thatare sensitive to JA-dependent defenses (Ton et al., 2002; Verhagen et al., 2006). However,because of the antagonistic effect of SA on JA signaling, we expected SAR-expressing plantsto have an enhanced susceptibility to S. exigua feeding, which was clearly not the case.

Interestingly, the negative effect of ISR and SAR on S. exigua performance (and the lackof it on P. rapae) correlated with a potentiated expression pattern of the defense-related genesPDF1.2 and HEL in microbially induced, herbivore-challenged plants (Fig. 3). Again, for ISRthe augmented expression of PDF1.2 and HEL was not unexpected because this type of inducedresistance is primarily based on priming for JA/ET-dependent defenses (Verhagen et al., 2004),but for SAR this result could not be readily explained. In the past, the role of priming duringSAR was mainly focused on SA-responsive genes, such as PAL and the SAR marker gene PR-1

104

(Conrath et al., 2002; 2006). Both genes show a potentiated expression pattern in SAR-expressing plants upon challenge inoculation with SA-inducing pathogens, or upon challengewith chemical elicitors of the SA response (e.g. BTH and INA). However, little is known aboutthe expression pattern of JA/ET-responsive genes in SAR-expressing plants upon challengewith biological or chemical elicitors of the JA response. Our data clearly show that inductionof SAR by avirulent P. syringae pv. tomato DC3000 (avrRpt2) is associated with primingfor enhanced expression of PDF1.2 and HEL upon feeding by S. exigua. Because inductionof SAR by predisposal infection with avirulent P. syringae pv. tomato DC3000 (avrRpt2)not only results in enhanced SA levels, but also in increased production of both JA and ET(De Vos et al., 2005, Chapter 2), the potentiated expression pattern of PDF1.2 and HELmay be associated with the pathogen-induced increase in the level of JA and ET.Previously, it was demonstrated that JA- and ET-impaired signaling mutants (e.g. jar1, etr1,and ein2) express normal levels of SAR against microbial pathogens, such as P. syringae pv.tomato DC3000 and Hyaloperonospora parasitica (Lawton et al., 1996; Pieterse et al., 1998).Hence, these signaling mutants will be instrumental in elucidating the role of priming for JA/ET-dependent defense-related genes during pathogen-induced SAR against S. exigua.

ISR, SAR and indirect insect defense

Previously, it was shown that the Arabidopsis genotypes S-12 and NahG are both affected inthe ability to attract C. rubecula (Van Poecke and Dicke, 2002), suggesting a role for both JAand SA in this induced indirect defense response. Because ISR and SAR are associated withpriming for enhanced JA/ET- or SA-dependent defense responses (Conrath et al., 2006), weexpected herbivore-damaged ISR- and SAR-expressing plants to be more attractive for C. rubecula wasps than herbivore-damaged control plants. However, neither ISR nor SAR hada significant effect on the attraction of C. rubecula wasps to Arabidopsis plants that were infestedby either P. rapae or S. exigua (Fig. 4). It must, therefore, be concluded that neither of the twotypes of microbially induced resistance have a significant effect on the herbivore-induced indirectdefense response in Arabidopsis.

ISR- and SAR-expressing plants are primed for enhanced expression of S. exigua-responsivegenes.Although this priming phenomenon was associated with enhanced direct defense againstS. exigua, it did not result in an increased attractiveness to the parasitic wasps, suggesting thatthe augmented defense response does not affect volatile synthesis.This observation is in agree-ment with the hypothesis that there may be a negative interaction between direct and indirectdefense (Dicke and Van Poecke, 2002).Alternatively, the differences in volatile production mighthave been too small to be detected in our experimental set-up.

Priming for enhanced defense

Over the past decades, there has been increasing evidence demonstrating that plants can beprimed for more efficient activation of cellular defense responses upon challenge, resulting in


105

enhanced resistance to various biotic or abiotic stresses (Conrath et al., 2006).The primed statecan be activated in various ways, such as upon infection by necrotizing pathogens, colonizationof plant roots by beneficial microbes (e.g. rhizobacteria and mycorrhizal fungi), insect herbivory,or specific natural or synthetic compounds, such as ß-aminobutyric acid (BABA), SA and BTH(Cameron et al., 1999; Van Wees et al., 1999; Zimmerli et al.., 2000; Kohler et al., 2002;Engelberth et al., 2004; Pozo et al., 2004; Verhagen et al., 2004; Ton et al., 2005; De Vos et al.,2006; Kessler et al., 2006; Paschold et al., 2006; Ton et al., 2007). Each inducer of theprimed state primes the plant for a specific set of genes. For instance, rhizobacteria-mediatedISR is associated with priming for enhanced JA/ET-responsive gene expression, whereas lowdoses of the SA-mimic BTH prime the plant for enhanced SA-responsive gene expression.Hence, the outcome of the defense reaction depends on the set of genes that is activated bythe attacker encountered. Priming of the SA response is not expected to affect pathogens andinsects that solely activate JA/ET-dependent defenses, whereas priming of the JA response isnot expected to affect attackers that solely activate SA-dependent defenses.The potential impactof the priming phenomenon in induced pathogen and insect resistance highlights the complexityof the defense signaling interactions between plants, pathogens and insect herbivores.

Materials and methodsPlant growth conditions

Arabidopsis thaliana accession Col-0, glabrous Col-5 plants (gl1-1 mutant of Col-0; carryinga mutation in GL1 resulting in a trichomeless phenotype), transgenic LOX2 co-suppressed S-12(gl1) plants harboring a sense construct of LOX2 cDNA behind the constitutive 35S-CaMVpromoter (Bell et al., 1995), coi1-16 originating from an ethylmethyl sulfonate (EMS) mutagenesisscreen for mutants impaired in their response to MeJA (Ellis and Turner, 2002), and transgenicNahG plants harboring the bacterial nahG gene (Delaney et al., 1994) were grown from seedsin quartz sand. Two-week-old seedlings were transplanted into 60-ml pots containing a sand-potting soil mixture that had been autoclaved twice for 20 min with a 24 h interval. Plants werecultivated in a growth chamber with an 8-h day (200 µE.m-2.s-1 at 24±1˚C) and a 16 h night(20±1˚C) cycle (L8:D16) at 70% relative humidity (RH) for 3 to 5 more weeks, dependingon the experiment. During the whole experiment, all plants remained in the vegetative state.Plants were watered on alternate days and once a week supplied with a modified half-strengthHoagland's nutrient solution, as described (Hoagland and Arnon, 1938).

Cultivation of microorganisms

The non-pathogenic, rifampicin-resistant Pseudomonas fluorescens strain WCS417r was usedfor induction of ISR (Pieterse et al., 1996).The strain was grown for 24 h at 28±1°C on King’smedium B agar plates (King et al., 1954) containing the appropriate antibiotics as described

106

previously (Pieterse et al., 1996). An avirulent strain of Pseudomonas syringae pv. tomatoDC3000 with the plasmid pV288 carrying the avirulence gene avrRpt2 (Kunkel et al., 1993)was used for SAR induction. P. syringae pv. tomato DC3000 (avrRpt2) bacteria were grownovernight at 28±1°C in liquid KB medium supplemented with 25 mg.ml-1 kanamycin to selectfor the plasmid.After centrifugation for 10 min at 5,000 x g, the bacterial cells were resuspendedin 10 mM MgSO4 to a final density of 107 cfu.ml-1.

Insect rearing

Pieris rapae (small cabbage white butterfly) was reared on Brussels sprouts plants (Brassicaoleracea gemmifera cv. Icarus) in a climate chamber (21±1°C, 50-70% RH, L16:D8), as describedpreviously (Van Poecke et al., 2001). First-instar (L1) larvae, hatched on Brussels sprouts plants,were used in all experiments.

Spodoptera exigua (beet armyworm) was reared on an artificial diet in a climate chamber(27±2°C, 70-80% RH, L16:D8) according to Smits et al. (1986). In all experiments, L1 larvaewere used, hatched in a climate room (23±1°C, 50-70% RH, L16:D8) and supplied with Brusselssprouts plants.

The parasitoid Cotesia rubecula was reared on P. rapae larvae on Brussels sprouts plantsin a glasshouse (25±5°C, 50-70% RH, L16:D8). The pupae were collected and transferred toa climate room (23±1°C, 50-70% RH, L16:D8).The emerging parasitoid wasps were providedwith water and honey. Adult female wasps without oviposition experience were selected forthe experiments, as described previously (Van Poecke et al., 2001).

Microbially induced resistance

Induction of ISR and SAR was performed as described (Pieterse et al., 1996). Briefly, to induceISR, a suspension of P. fluorescens WCS417r rhizobacteria was mixed thoroughly through thesoil to a final density of 5x107 cfu.g-1 of soil. Subsequently, 2-week-old seedlings were transferredto the soil and allowed to grow for another 3-5 weeks. For induction of SAR, three lower leavesof 5- to 7-week-old plants were pressure-infiltrated with a suspension of P. syringae pv. tomatoDC3000 (avrRpt2) at 107 cfu.ml-1.When used for parasitoid two-choice flight experiments,the infiltrated leaves of SAR-induced plants were removed after three days, as were the threecorresponding leaves of uninduced or ISR-induced plants.

Herbivore development bioassays

Using a fine paint brush, single L1 larvae of P. rapae or S. exigua were transferred to individual6- to 7-week-old Arabidopsis plants (n = 20-25). Each plant was confined in a Magenta GA-7vessel (SIGMA-Aldrich, Zwijndrecht, the Netherlands) with an insect proof mesh lid. TheMagenta boxes were kept in a climate room (23±1°C, 50% RH, L16:D8).The caterpillars were


107

weighed to the nearest 0.1 mg on a micro balance (Sartorius AC211P).The time until pupationwas determined by scoring the day on which the caterpillars pupated (P. rapae) or the daythe larvae moved into the soil in order to pupate (S. exigua). Once the first plant was eaten,a second plant was supplied to all caterpillars on the same day. Usually, two plants were enoughfor the caterpillars to reach pupation. The performance of P. rapae on coi1-16 plants wasdetermined by placing a single L3/L4 larva onto individually confined plants for 48 h. Thelarvae were weighed before and after the feeding period. Subsequently, the relative increasein weight was determined. Non-parametric tests were performed to analyze weights of larvaethat survived until pupation (Kruskal-Wallis test followed by Mann-Whitney U test for pair-wisecomparisons). One-way ANOVA with an LSD post hoc test was performed to analyze the timeto pupation. The data were analyzed in SPSS 11.5 for Windows.

RNA extraction and northern blot analysis

Total RNA was extracted as described (Van Wees et al., 1999) from shoots of 5-week-old plantsthat were either uninduced or expressing ISR or SAR. Leaf material was harvested 24 or 72h after challenge with 5 L1 P. rapae or 10 L1 S. exigua larvae. For RNA-blot analysis, 15 µgof RNA were denatured using glyoxal and dimethyl sulfoxide (Sambrook et al., 1989).Subsequently, the denatured RNA was separated by electrophoresis on a 1.5% agarose gel,and blotted onto a Hybond-N+ membrane (Amersham Biosciences, Roosendaal, the Netherlands)by capillary transfer. The electrophoresis and blotting buffer consisted of 10 mM and 25 mMsodium phosphate (pH 7.0), respectively. Gene-specific probes for PDF1.2 and HEL were usedfor hybridization as described previously (Van Wees et al., 1999). The Arabidopsis GenomeInitiative numbers for the genes studied are At5g44420 (PDF1.2) and At3g04720 (HEL).A probe for 18 S ribosomal RNA, derived from an Arabidopsis cDNA clone (Pruitt andMeyerowitz, 1986), was used to check for equal loading.

Parasitoid two-choice flight experiments

Parasitoid two-choice flight experiments were performed in a wind-tunnel set-up (25±5°C,50-70% RH, 13 mE.m-2.s-1) with an air flow of 0.2 m.s-1 as described by Geervliet et al. (1994).The set-up was used as described (Van Poecke et al., 2001) with 15 5- to 7-week-old Arabidopsisplants per odor source. Uninduced, ISR- or SAR-expressing plants were infested with L1 larvaeusing a fine paint brush. Five L1 P. rapae larvae per plant were allowed to feed for 24 h. S. exigualarvae were smaller and caused less damage than P. rapae larvae. Therefore, 10 L1 S. exigualarvae per plant were allowed to feed for 72 h. The flight behavior of the parasitoid wasp C. rubecula was observed. First landings on one of the odor sources were recorded as “choice”,whereas first landings elsewhere in the experimental set-up, or no flight within 10 min, werenoted as “no-choice”. Per experimental day, 3 (no damage test) or 4 (caterpillar-damage test)odor sources were compared in pair-wise combinations. Per odor combination, an equal numberof wasps, either 6 or 8, were tested. Moreover, the position of the two odor sources was

108

exchanged after testing half of the number of parasitoids per combination to avoid positionaleffects on the behavior of the wasps. Chi-square tests were used to analyze pair-wise differencesin the number of choices between the two odor sources. Differences in the total number ofresponsive wasps between the different odor combinations were determined using a contingencytable.

Damage analysis

For each two-choice flight experiment, total damage caused by caterpillar feeding was determinedfor five plants per treatment per day.Total leaves were photocopied and subsequently scannedwith a Sony b/w CCD camera type XC-77CE (frame size 752 x 574; 256 grey levels). Imageanalysis was performed using the KS400 version 3.0 software service pack 9 (Carl Zeiss Vision,Oberkochen, Germany) to quantify the total size of the damaged leaf area.

AcknowledgementsThis work was supported, in part, by grants 811-36-004, 865-04-002 and 813-06-002 of theEarth and Life Sciences Foundation (ALW), which is subsidized by the Netherlands Organisationof Scientific Research (NWO).We gratefully acknowledge Wendy van Zaanen, JoffreyMandersloot,Antonio Leon-Reyes, Henk Brouwer and Maarten Terlou for technical assistanceand Leo Koopman, Frans van Aggelen and André Gidding for insect rearing.


109

110

Microarray analysis of insect herbivore-induced genes inArabidopsis plants expressing rhizobacteria-induced systemicresistance

Vivian R. Van Oosten1,2, Natacha Bodenhausen3, Philippe Reymond3, L.C. Van Loon1,

Corné M.J. Pieterse1 and Marcel Dicke 2


Biology, Faculty of Science, Utrecht University, P.O. Box 800.84, 3508 TB Utrecht, The Netherlands.



3 Department of Plant Molecular Biology, University of Lausanne, CH-1015 Lausanne, Switzerland.

111Microarray analysis of insect herbivore-inducedgenes in Arabidopsis plants expressing rhizo-bacteria-induced systemic resistance

chapter 5

AB

ST

RA

CT In nature, mutually beneficial interactions between plants and microbes can result in an enhanced

capacity of the plant to ward off a broad spectrum of pathogens. In Arabidopsis, non-pathogenic rhizo-

bacteria can trigger an induced systemic resistance (ISR) that is effective against a large range of

pathogens. The onset of ISR in Arabidopsis is not associated with major systemic alterations in gene

expression. However, upon pathogen challenge a large number of jasmonic acid (JA)-responsive defense

genes show augmented expression, indicating that ISR expressing plants are primed to respond faster

or stronger to pathogen attack. Previously, we investigated the effectiveness of ISR in Arabidopsis against

the tissue-chewing insect herbivores Pieris rapae and Spodoptera exigua and demonstrated that micro-

bially induced ISR is effective against S. exigua, but not against P. rapae. RNA-blot analysis of the

defense-related genes PDF1.2 and HEL indicated that the effectiveness against S. exigua was associated

with priming for enhanced expression of these genes. To identify additional ISR-primed, S. exigua-

induced genes, we surveyed the transcriptome of uninduced and ISR-expressing Arabidopsis upon

challenge with P. rapae and S. exigua, using a dedicated microarray containing 111 Arabidopsis genes

that are responsive to caterpillar feeding. We obtained further evidence that the transcriptome of ISR-

expressing Arabidopsis plants responds differentially to feeding by P. rapae and S. exigua. A set of five

genes showed increased expression, only upon S. exigua feeding, whereas six genes were repressed

specifically during P. rapae infestation.

112

IntroductionPlants can be protected from disease through interactions with beneficial members of themicrobial community in the soil.An important group of beneficial microorganisms are the non-pathogenic, plant growth-promoting rhizobacteria (PGPR). Not only can PGPR antagonizedeleterious microorganisms in the soil directly (Schippers et al., 1987), selected strains of rhizobacteria are also capable of triggering a plant-mediated defense response, resulting in anenhanced defensive capacity (Van Loon et al., 1998). Rhizobacteria-mediated induced systemicresistance (ISR) is effective against a broad spectrum of pathogens, and has been demonstratedto occur in different plant species (Van Loon and Bakker, 2006). In Arabidopsis, ISR triggeredby the non-pathogenic rhizobacterium Pseudomonas fluorescens WCS417r is effective pre-dominantly against pathogens that are sensitive to jasmonic acid (JA)- or ethylene (ET)-dependentbasal defenses (Ton et al., 2002; Verhagen et al., 2006). Unlike pathogen-induced systemicacquired resistance (SAR),WCS417r-mediated ISR is not associated with the activation of PRgenes (Pieterse et al., 1996). Mutant analysis showed that ISR functions independently of salicylic acid (SA), but requires responsiveness to JA and ET (Pieterse et al., 1996; 1998).

Analysis of the Arabidopsis transcriptome revealed that locally in the roots ISR-inducingWCS417r bacteria elicited a substantial change in the expression of almost 100 genes (Verhagenet al., 2004; Léon-Kloosterziel et al., 2005). In contrast, systemically in the leaves no consistentalterations in gene expression were observed.Thus, the onset of ISR in leaves is not associatedwith obvious changes in gene expression (Verhagen et al., 2004). No alterations in the productionof either JA or ET were detected in plants expressing ISR, suggesting that the induced resistancestate is based on an enhanced sensitivity to these plant hormones rather than on an increasein their production (Pieterse et al., 2000). Further analysis after challenge inoculation of ISR-expressing Arabidopsis leaves with Pseudomonas syringae pv. tomato DC3000 revealed a largeset of genes with augmented expression, indicating that the plants were primed to respondfaster and/or more strongly to pathogen attack (Verhagen et al., 2004). The majority of thesegenes were predicted to be regulated by JA and/or ET, indicating that colonization of the rootsby WCS417r primes Arabidopsis plants for augmented expression of JA- and/or ET-responsivegenes. Other ISR-inducing PGPR have also been demonstrated to enhance the plant’s defensivecapacity by priming for potentiated expression of defense genes upon challenge (e.g. De Meyeret al., 1999; Ahn et al., 2002; Kim et al., 2004; Tjamos et al., 2005), suggesting that primingis a common feature of PGPR-mediated ISR.

Priming has been demonstrated in various plant species (Conrath et al., 2006).The primedstate of the plant refers to the physiological condition in which a plant is capable of activatingdefense responses in a faster and stronger manner upon attack by pathogens, insects or inresponse to abiotic stress (Conrath et al., 2002). It can be achieved by application of low dosesof elicitors, primary infection with necrotizing pathogens, or colonization of the roots by ISR-inducing rhizobacteria. Initial studies on priming in plants were performed using parsley cellcultures.At low doses, an oomycete cell-wall elicitor or SA did not activate the defense-relatedgene PAL encoding phenylalanine ammonia-lyase. In contrast, when the cells had been pre-treated and were then challenged with the elicitor, PAL was expressed to high levels (Thulkeand Conrath, 1998), indicating that the cells had been primed for potentiated PAL expression.

Chapter 05: Microarray analysis of insect herbivore-induced genes in Arabidopsisplants expressing rhizobacteria-induced systemic resistance

113

Although research on priming has been focused mainly on enhanced defense againstpathogens, an increasing number of studies report on the role of priming in plant-insectinteractions. For instance, feeding by larvae of the specialist insect herbivore Pieris rapae(small cabbage white butterfly) primes Arabidopsis plants for potentiated expression of SA-dependent defenses, such as those triggered upon challenge with turnip crinkle virus (TCV)(De Vos et al., 2006). Thus, feeding by this herbivore primes the plant for enhanced defenseagainst a viral pathogen.The emission of volatiles that are commonly released upon herbivoryleads to the attraction of carnivorous enemies of the attacking insect (Dicke, 1999).Volatileshave also been implicated in priming. Corn plants that were exposed to specific green leafvolatiles rapidly started to produce JA and sesquiterpene volatile compounds to much higherlevels upon wounding plus application of caterpillar regurgitant than corn seedlings withoutthe pretreatment (Engelberth et al., 2004). This indicates that exposure to green leaf volatilesprimed the plants for potentiated JA production and volatile emission. Indeed, plant volatilesproduced upon insect herbivory have been shown to prime neighboring plants for augmentedJA-dependent defense responses (Choh et al., 2004; Heil and Kost, 2006; Kessler et al., 2006;Paschold et al., 2006; Ton et al., 2007).Thus, communication between plants through herbivore-induced volatile emission can prime other members of the plant community to better defendthemselves against subsequent herbivore attack.

Phytophagous insects are typically resisted through JA-dependent defense mechanisms.In tomato, attack by tissue-chewing or cell-piercing insects results in enhanced synthesis ofJA and related oxylipins, and subsequent activation of JA-responsive genes (Farmer and Ryan,1992; Li et al., 2002). A range of mutant and transgenic plants with decreased production orperception of JA have been shown to become more susceptible to various herbivorous insects(Orozco-Cardenas et al., 1993; Howe et al., 1996; McConn et al., 1997; Ellis et al., 2002; Li et al., 2002; Thaler et al., 2002; Kessler et al., 2004), indicating the importance of JA forplant resistance against insects with different feeding modes. Moreover, JA mediates the herbivore-induced emission of plant volatiles that attract carnivorous insects as an indirectdefense (Dicke et al., 1999; Lou et al., 2005).

In Arabidopsis, grazing by P. rapae or Spodoptera littoralis (Egyptian cotton worm) cater-pillars stimulates the production of JA and the expression of a large set of predominantly JA-responsive genes (Reymond et al., 2004; De Vos et al., 2005, Chapter 2). Larvae of either P. rapaeor Spodoptera exigua (beet armyworm) perform better on Arabidopsis mutants with decreasedJA production or JA signaling (Reymond et al., 2004, Chapter 4), indicating that JA-dependentdefense responses are effective against these insects. JA is also involved in P. rapae-induced indirect defense mediated by parasitoid-attracting volatiles (Van Poecke and Dicke, 2002).Previously, we showed that ISR-expressing Arabidopsis reduced the growth and development ofthe generalist caterpillar S. exigua, whereas the specialist insect herbivore P. rapae remained un-affected (Chapter 4).The reduced development of S. exigua was correlated with an augmentedexpression of the JA/ET-responsive genes PDF1.2 and HEL, indicating that the effectiveness ofISR against S. exigua was associated with priming.These genes were not differentially expressedupon feeding by P. rapae.To gain insight into the nature of the genes that show a potentiated ex-pression pattern in ISR-expressing plants upon feeding by S. exigua, we analyzed the transcriptomeof Arabidopsis, using a dedicated Arabidopsis microarray containing 111 herbivore-induced genes.

114

ResultsValidation of potentiated PDF1.2 expression in ISR-expressingplants upon challenge with S. exigua, but not P. rapae

Previously, we observed that effectiveness of ISR against insect herbivores was correlated withpotentiated expression of the defense-related genes PDF1.2 and HEL (Chapter 4). To furthervalidate the differentially primed expression pattern of S. exigua- and P. rapae-induced genesin ISR-expressing plants, and to identify additional insect-responsive genes that show a primedexpression pattern in microbially induced plants, uninduced and ISR-expressing Arabidopsisplants were infested with first-instar P. rapae or S. exigua larvae. Plant material was harvestedafter 24, 48 or 72 h of feeding in three independent experiments.The RNA was extracted andPDF1.2 expression was verified by northern blot analysis. The result of one representativeexperiment is shown in Figure 1.As observed previously (Chapter 4), both P. rapae and S. exiguafeeding induced PDF1.2 expression in uninduced plants. Moreover, PDF1.2 showed an augmented expression pattern in ISR-expressing plants upon S. exigua challenge (3.4-fold),whereas there was no obvious difference during P. rapae infestation (1.3-fold).The augmentedexpression of PDF1.2 in S. exigua-infested, ISR-expressing Arabidopsis plants was most obvious48 h after S. exigua infestation.Therefore, the RNA from Arabidopsis leaves harvested at 48 hwas selected for microarray analysis. Despite variation in expression level and timing betweenindependent experiments, these results confirmed the potentiated PDF1.2 expression in S. exigua–infested, ISR-expressing Arabidopsis, and its lack in P. rapae-infested plants.


115

FIGURE 1. Primed PDF1.2 gene expression in ISR-expressing Arabidopsis plants upon challenge with S. exigua, butnot P. rapae.

RNA-blot analysis showing PDF1.2 transcript levels in untreated control (Ctrl) and ISR-expressing (ISR) Arabidopsisplants that were either uninfested or infested with P. rapae or S. exigua for 24, 48 or 72 h. Signal intensities of PDF1.2mRNA on the RNA blot were quantified using a phosphor imager and normalized for equal levels of 18S ribosomalRNA (rRNA). The fold differences in PDF1.2 transcript levels in ISR-expressing over control plants (ISR/Ctrl) are givenand compared to the fold differences as observed in the microarray analysis. The experiment was performed threetimes with similar results. h.p.i., hours post infestation.

Microarray analysis of herbivore-induced gene expression

Experimental set-upTo identify additional primed genes, we performed microarray experiments with a dedicatedcDNA microarray containing probes for 111 Arabidopsis genes that were previously shown to be induced by P. rapae (Reymond et al., 2004). In addition, the microarray contained cDNAsof 70 pathogen- or drought-responsive genes, and 17 housekeeping genes (Bodenhausen andReymond, in preparation). Microarray hybridizations were performed with RNA probes fromthree biological replicates of leaf material that was harvested from uninduced or ISR-expressingplants 48 h after challenge with S. exigua or P. rapae. RNA probes from unchallenged plantswere used as references on all microarrays.The following treatments were compared on twelvemicroarrays (Supplementary Table 1): 1) P. rapae on uninduced control vs. uninfested, uninducedcontrol plants; 2) P. rapae on ISR-expressing vs. uninfested, ISR-expressing plants; 3) S. exiguaon uninduced control vs. uninfested, uninduced control plants; 4) S. exigua on ISR-expressingvs. uninfested, ISR-expressing plants. To identify insect-responsive and ISR-primed genes, weapplied the selection criteria described in the Materials and Methods section.

116

Microarray analysis of herbivore-induced gene expression under different laboratoryconditionsThe dedicated microarray was comprised of genes selected for their induction after 3-5 h offeeding by fourth- and fifth-instar (L4-L5) P. rapae larvae (Reymond et al., 2004).The majorityof these genes were also responsive to L4-L5 S. littoralis feeding for 3-5 h (Reymond et al., 2004),and to 24 h feeding by first-instar (L1) P. rapae or S. littoralis larvae (N. Bodenhausen and P. Reymond, unpublished results). Since we challenged Arabidopsis for 48 h with L1 larvae of P. rapae or S. exigua using slightly different experimental conditions, we compared geneexpression ratios of experiments performed at the University of Utrecht, the Netherlands(Utrecht experiment, in triplo: 48 h L1 P. rapae on control vs. uninfested, control plants; 48 hL1 S. exigua on control vs. uninfested control plants) with the material prepared at the Universityof Lausanne, Switzerland (Lausanne experiment, in duplo: 24 h L1 P. rapae on control vs.uninfested control plants; 24 h L1 S. littoralis on control vs. uninfested control plants)(Supplementary Table 2). First, we compared the gene expression of uninduced control plantsupon feeding by L1 P. rapae for 24 h (Lausanne experiment) with feeding for 48 h (Utrechtexperiment) (Fig. 2A). In the Lausanne experiment, challenge with P. rapae for 24 h resultedin the identification of 94 induced genes. In the Utrecht experiment, challenge for 48 h yielded68 induced genes.Although the expression ratios differed between these experiments, 56 geneswere induced in both experiments, indicating that 82% of the P. rapae-induced changes in theUtrecht experiment were also found in the Lausanne experiment.We also compared the geneexpression ratios upon challenge of Arabidopsis with S. littoralis for 24 h (Lausanne experiment)with feeding for 48 h by S. exigua (Utrecht experiment) (Fig. 2B). S. littoralis induced theexpression of 73 genes in the Lausanne experiment, whereas 41 genes were induced by S. exiguain the Utrecht experiment. Of the 41 S. exigua-induced genes, 38 genes (93%) were also inducedby S. littoralis.Thus, the similarity of the Utrecht and Lausanne experiments with two differentSpodoptera species was comparable to that of the experiments in Lausanne and Utrecht thatboth involved P. rapae caterpillars. Collectively, these results demonstrate that there is a highcorrelation between the experiments conducted in Lausanne and Utrecht, despite the differentlaboratory conditions and the different durations of caterpillar infestation. Hence, the dedicatedmicroarray is a valid tool for further identification of primed, insect-responsive genes.


117

FIGURE 2. Overlap of insect-induced genes from experiments performed in Lausanne and Utrecht with Arabidopsisplants challenged with larvae of Pieris rapae, Spodoptera littoralis or Spodoptera exigua.

The experiments were carried out in similar, but not identical laboratory conditions in Lausanne and Utrecht. Plantchallenge with P. rapae or S. littoralis (24 h) was performed in Lausanne. Plant challenge with P. rapae or S. exigua(48 h) was carried out in Utrecht. Indicated are numbers of induced genes from plants challenged with Pieris rapae(24 h-Lausanne vs. 48 h-Utrecht) (A), or Spodoptera littoralis (24 h-Lausanne) vs. Spodoptera exigua (48 h-Utrecht) (B).

Differential gene induction in ISR-expressing plants upon feeding by P. rapae and S. exigua

The onset of rhizobacteria-mediated ISR was previously demonstrated to have no obviouseffect on gene expression in the shoots (Verhagen et al., 2004). To verify this conclusion, micro-array analysis was performed with probes derived from uninfested control, and uninfested,ISR-expressing plants from a single experiment.As expected, none of the genes on the dedicatedmicroarray were induced in plants treated with the P. fluorescens strain WCS417r (SupplementaryTable 3). Having confirmed the lack of transcriptional differences between the uninduced andISR-expressing plants, we continued to study differences in the response of uninduced and ISR-expressing plants to insect challenge. In total, 12 microarray analyses were performed in whichwe compared the effect of caterpillar feeding with the corresponding no-damage control(Supplementary Table 1). P. rapae feeding induced 68 genes in uninduced control plants, and 59in ISR-expressing plants (Table 1).The majority of the P. rapae-induced genes in control plantswere also induced in the ISR-expressing plants (52 genes, 76%). Challenge with S. exiguainduced 41 genes in uninduced control plants and 53 in ISR-expressing plants (Table 2). All 41S. exigua-induced genes in control plants were also induced in ISR-expressing plants. Moreover,all S. exigua-responsive genes were also responsive to P. rapae feeding.

118


119

The average gene expression ratios of three replicates of ISR-expressing plants were log2-transformed and plotted against the log2-transformed data from the control experiments withP. rapae (Fig. 3A) or S. exigua (Fig. 3B). Only genes that were significantly (P-value <0.05),and >2-fold induced in at least one of the treatments were included in the analysis (representedby grey dots and open triangles, Fig. 3). The correlation coefficient was close to 1, indicatingthat gene expression ratios were very similar between ISR-expressing and uninduced Arabidopsisupon challenge with P. rapae (R2 = 0.93) or S. exigua (R2 = 0.96).

From the herbivore-induced genes, we selected those with a potentiated expression patternin ISR-expressing plants when compared to uninduced plants. To this end, we identified theISR-primed genes with statistically significantly different expression ratios (represented by triangles,Fig. 3; Table 3). Previously, ISR-primed genes were selected based on >1.5-fold difference inexpression level (Verhagen et al., 2004). Therefore, we also selected the genes with >1.5-folddifference in expression level in at least two out of the three biological replicates (representedby triangles, Fig. 3; Table 3). PDF1.2 transcript levels were 1.6-fold higher in S. exigua-challenged,ISR-expressing plants than in S. exigua-challenged control plants, confirming the primedexpression pattern of PDF1.2 (Fig. 1; Fig. 3; Table 3). As in the RNA blot analysis, the primedexpression pattern of PDF1.2 was not observed upon P. rapae feeding (1.3-fold in RNA blotanalysis, 0.6-fold in microarray analysis).Although the expression ratios differed both betweenexperiments and according to the techniques used, the augmented expression in rhizobacteria-treated Arabidopsis upon S. exigua infestation was consistent between the two methods in threeindependent replicates.

In the S. exigua-challenged plants, we identified four additional ISR-primed genes (Table 3).These genes encode a cysteine proteinase and three lipid transfer proteins (LTPs). Like PDF1.2,these genes showed a potentiated expression pattern in S. exigua-infested plants, but not inP. rapae-infested plants. Seven genes showed a significant repression in ISR-expressingArabidopsis upon P. rapae challenge (Table 3).These genes encode a glutathione S-transferase(GSTU5, naming convention after (Wagner et al., 2002)), two putative myrosinase-bindingproteins, and genes involved in the metabolism of the phytohormones JA (hydroxyjasmonatesulfotransferase) and auxin (IAR3, coding for IAA-ala hydrolase). The senescence-associatedgene SAG21 was also repressed in P. rapae-challenged, ISR-expressing plants. An expressedprotein (At1g74950) with a ZIM-domain that occurs in specific types of zinc-finger transcriptionfactors was significantly repressed in ISR-expressing plants upon challenge by both P. rapaeand S. exigua.

Thus, ISR-expressing plants showed a potentiated induction of a set of five genes uponS. exigua challenge, whereas P. rapae challenge resulted in the repression of seven genes onthe dedicated microarray. These results strengthen our notion that ISR-expressing Arabidopsisis differentially primed for augmented gene expression upon challenge by the generalist herbivoreS. exigua and the specialist P. rapae.


121


123

124

FIGURE 3. Comparison of changes in gene expression in uninduced control and in ISR-expressing Arabidopsis,upon challenge with Pieris rapae or Spodoptera exigua.

Pairwise comparisons of log2-transformed expression ratios of uninduced control and ISR-expressing Arabidopsisafter challenge for 48 h with P. rapae (A.) or S. exigua (B.). Mean expression ratios calculated from experimentscomparing herbivore-challenged and unchallenged, uninduced control plants (three biologically independent replicates)are plotted against mean expression ratios between challenged and unchallenged ISR-expressing plants (three bio-logically independent replicates). Genes expressed < 2-fold are represented by black dots (not induced), genes induced> 2-fold in both treatments, but not differentially expressed, are represented in grey. The genes represented by trianglesare ISR-primed: they are significantly different between uninduced control and ISR-expressing plants (Student’s t test, P < 0.05), or 1.5-fold different in at least two out of three replicates. The 1.5-fold difference (increased or repressed)in gene expression ratio between control and ISR-expressing plants is indicated by the dotted lines.

ISR-primed genes contain AtMYC2 binding motif in their promoter

Previously, we identified a set of 132 JA-responsive genes that show a potentiated expressionpattern in ISR-expressing plants upon treatment with methyl JA (MeJA) (Pozo et al., inpreparation, M. Pozo and S.Van der Ent, unpublished results). Analysis of the promoters ofthese ISR-primed genes revealed that these promoters are significantly enriched for the elementCACATG, which is a binding site for the transcription factor AtMYC2 (Abe et al., 1997).Mutant Arabidopsis plants lacking a functional MYC2 protein are unable to mount ISR againstseveral pathogens upon rhizobacteria treatment, demonstrating a function for AtMYC2 in ISR(S.Van der Ent and M. Pozo, unpublished results). To check if the promoters of the S. exigua-induced, ISR-primed genes also contain the AtMYC2 binding motif, we analyzed the first 1000base pairs (bp) upstream of the start codon of the 12 selected genes (Table 3) using the MotifAnalysis Tool of The Arabidopsis Information Resource (TAIR) (Rhee et al., 2003). Of allannotated genes in the Arabidopsis genome, 36% contain the CACATG motif in the first 1000bp of their promoters. Of all 111 herbivore-induced genes spotted onto the dedicated microarray,45 (41%) contained the CACATG motif in the first 1000 bp of their promoters.All five S. exigua-


125

induced, ISR-primed genes contained the AtMYC2 binding motif. In contrast, only two of theseven repressed genes contained an AtMYC2 binding site.Taken together, these results suggestthat AtMYC2 might be involved in the augmented expression of defense genes in ISR-expressingArabidopsis during S. exigua infestation.

DiscussionPriming for enhanced defense has mainly been studied during defense against pathogens.Only recently, the involvement of priming in plant-insect interactions has become apparent(Conrath et al., 2006, and references therein). Plant volatiles produced upon wounding or insectherbivory prime neighboring plants for augmented JA-dependent defense responses, therebyenhancing their induced protection mechanisms against subsequent attack by insect herbivores.Rhizobacteria-mediated ISR in Arabidopsis is associated with priming for augmented JA-dependent defense responses against pathogens (Verhagen et al., 2004) and insects (Chapter4). Using a dedicated microarray containing 111 insect-responsive genes, we identified anadditional set of Arabidopsis genes with a primed expression pattern during insect challenge.These genes were differentially primed for augmented expression upon feeding by the specialistherbivore P. rapae or the generalist S. exigua.

Microarray analysis of herbivore-induced gene expression

Previously, we found a correlation between the augmented expression of the JA/ET-responsivegene PDF1.2 in ISR-expressing, S. exigua-challenged plants, and enhanced resistance againstthis insect (Chapter 4). RNA blot analysis of three independent additional experiments confirmedthe primed expression pattern of PDF1.2 in ISR-expressing, S. exigua-damaged plants, andits lack in P. rapae-damaged plants (Fig. 1).Additionally, we found that P. rapae-feeding induced75 genes (Table 1), and S. exigua-feeding 53 genes in uninduced or ISR-expressing plants(Table 2). In ISR-expressing plants, several genes showed potentiated expression levels: six uponP. rapae infestation, five upon S. exigua infestation, and one upon infestation by either insect.The ratio of primed genes over all attacker-induced genes corresponds to previous findings(Verhagen et al., 2004; Pozo et al., in preparation) and ranges between 7% and 12% of theattacker-responsive genes. Microarray analysis revealed four genes, in addition to PDF1.2, thatshow a primed expression pattern in ISR-expressing plants upon herbivory by S. exigua, andnot upon feeding by P. rapae (Table 3). Seven other genes were negatively primed, meaningthat they were expressed at a significantly lower level in ISR-expressing plants than in uninducedplants upon herbivory. Six of the negatively primed genes had a lower expression level uponP. rapae challenge, and not upon S. exigua challenge.The seventh gene was expressed to a lowerlevel upon infestation with P. rapae and with S. exigua in ISR-expressing plants.

Of the positively primed genes, three (At3g22600, At3g22620 and At5g59310) encodelipid transfer proteins (LTP). The Arabidopsis genome contains more than 15 LTP-encoding

126

genes (Arondel et al., 2000), which are expressed during a wide range of biotic or abiotic stresses.The precise function of LTPs in plants is still unknown, although some of them have beenimplicated in plant defense (Segura et al., 1993; García-Olmedo et al., 1995; Kader, 1997).Interestingly, the LTP-encoding gene DIR1 (At5g48485) was previously shown to be requiredfor SAR signal transduction (Maldonado et al., 2002). Therefore, it is tempting to speculatethat the identified LTPs play a role in induced defense signaling. The fourth ISR-primed gene(At4g11320) encodes a cysteine protease-like protein with similarity to cysteine protease RD21(Yamada et al., 2001), which is a member of the papain family of cysteine proteases that havebeen shown to possess anti-insect properties (Konno et al., 2004).

Of the seven negatively primed genes, six were specifically repressed in P. rapae-damaged,ISR-expressing plants (Table 3). One gene (At2g29450) encodes a glutathione S-transferase(GSTU5) that is induced by P. rapae in a coi1-dependent fashion (Reymond et al., 2004).Glutathione S-transferases are plant enzymes that can inactivate toxic compounds and havebeen implicated in numerous stress responses, including pathogen attack and oxidative stress(Marrs, 1996). The stress-induced, senescence-associated gene SAG21 (At4g02380) was alsosignificantly less induced in ISR-expressing plants.Two genes encode putative myrosinase-bindingproteins (At3g16410 and At3g16420). Myrosinases are ß-glucosidases that are specific to theBrassicaceae. They are capable of hydrolyzing glucosinolates, which are defensive compoundsthat are toxic to many generalist insects, but serve as feeding and oviposition stimulants tospecialist herbivores (reviewed in Halkier and Gershenzon, 2006). At3g16420 encodes PBP1(PYK10-binding protein 1) that assists the ß-glucosidase PYK10 in its activity (Nagano et al.,2005).

Two genes involved in hormone metabolism were negatively primed.The first (At5g07010)is a hydroxyjasmonate sulfotransferase involved in the modification of the phytohormone JA.The second gene (At1g51760) codes for the IAA-ala hydrolase IAR3. This gene is implicatedin the production and release of auxin (Davies et al., 1999). Both genes were shown to beinduced by P. rapae in a coi1-dependent manner (Reymond et al., 2004). The gene that wasrepressed upon P. rapae and upon S. exigua challenge (At1g74950) encodes a putative trans-cription factor protein with a ZIM (Zinc-finger protein expressed in Inflorescence Meristem)domain (Nishii et al., 2000). The same gene has been demonstrated to be down-regulated in Arabidopsis plants in which ISR was induced by the rhizobacterial strain P. fluorescensFPT9601-T5 (Wang et al., 2005).

Because of their putative role in plant defense, one can speculate that repression of thesegenes promotes growth of the P. rapae larvae. However, in our previous experiments, we didnot observe a significant difference in growth and development of P. rapae larvae on ISR-expressing, compared to uninduced control plants (Chapter 4).

In conclusion, the microarray analysis confirms previous findings that the specialist P. rapaeand the generalist insect herbivores S. exigua or S. littoralis induce highly similar changes inthe transcriptome of Arabidopsis.We extend our analysis of differentially primed gene expressionin Arabidopsis plants expressing P. fluorescens WCS417r-mediated ISR upon challenge withP. rapae or S. exigua. These results highlight the complexity of the plant’s capacity to adaptto different environmental cues. Once a defense signaling pathway is activated, such as uponcolonization of the roots by ISR-inducing rhizobacteria, the response to a second stimulus is


127

affected, resulting in a different outcome of the resistance reaction. In the past years, the primingphenomenon emerged as an important mechanism in different types of induced resistance.Hence, future research on this topic will yield exciting new insights in the complexity of theplant’s defense signaling network.

Materials and methodsPlant growth conditions

Arabidopsis thaliana accession Col-0 was grown from seeds in quartz sand. Two-week-oldseedlings were transplanted into 60-ml pots containing a sand-potting soil mixture that hadbeen autoclaved twice for 20 min with a 24-h interval. Plants were cultivated in a growthchamber with an 8-h day (200 µE.m-2.s-1 at 24±1˚C) and a 16-h night (20±1˚C) cycle (L8:D16)at 70% relative humidity (RH) for 5 more weeks. During the whole experiment, all plantsremained in the vegetative state. Plants were watered on alternate days and once a weeksupplied with a modified half-strength Hoagland's nutrient solution, as described (Hoaglandand Arnon, 1938).

Cultivation of microorganisms

The non-pathogenic, rifampicin-resistant Pseudomonas fluorescens strain WCS417r was usedfor induction of ISR (Pieterse et al., 1996).The strain was grown for 24 h at 28±1°C on King’smedium B agar plates (King et al., 1954) containing the appropriate antibiotics as describedpreviously (Pieterse et al., 1996).

Microbially induced resistance

Induction of ISR was performed as described (Pieterse et al., 1996). Briefly, to induce ISR,a suspension of P. fluorescens WCS417r rhizobacteria was mixed thoroughly through the soilto a final density of 5x107 cfu.g-1 of soil. Subsequently, 2-week-old seedlings were transferredto the soil and allowed to grow for another 5 weeks.

Insect rearing

Pieris rapae (small cabbage white butterfly) was reared on Brussels sprouts plants (Brassicaoleracea gemmifera cv. Icarus) in a climate chamber (21±1°C, 50-70% RH, L16:D8), as describedpreviously (Van Poecke et al., 2001). First-instar (L1) larvae, hatched on Brussels sprouts plants,were used in all experiments.

128

Spodoptera exigua (beet armyworm) was reared on an artificial diet in a climate chamber(27±2°C, 70-80% RH, L16:D8) according to Smits et al. (1986). In all experiments, L1 larvaewere used, hatched in a climate room (23±1°C, 50-70% RH, L16:D8) and supplied with Brusselssprouts plants.

RNA extraction and RNA blot analysis

For the Utrecht experiment, total RNA was extracted as described (Reymond et al., 2004) fromshoots of 7-week-old plants that were either uninduced or expressing ISR. Leaf material washarvested 24, 48 or 72 h after challenge with 5 L1 P. rapae or 10 L1 S. exigua larvae. Fornorthern blot analysis, 15 µg of RNA was denatured using glyoxal and dimethyl sulfoxide(Sambrook et al., 1989). Subsequently, the denatured RNA was separated by electrophoresison a 1.5% agarose gel, and blotted onto a Hybond-N+ membrane (Amersham Biosciences,Roosendaal, the Netherlands) by capillary transfer. The electrophoresis and blotting bufferconsisted of 10 mM and 25 mM sodium phosphate (pH 7.0), respectively. A gene-specificprobe for PDF1.2 (At5g44420) was used for hybridization as described previously (Van Weeset al., 1999).A probe for 18S ribosomal RNA, derived from an Arabidopsis cDNA clone (Pruittand Meyerowitz, 1986), was used to check for equal loading. For microarray analysis, RNAwas further purified using an RNeasy Mini Kit (Qiagen, Basel, Switzerland).

For the Lausanne experiment, 5-7 P. rapae or 10 Spodoptera littoralis L1 larvae wereallowed to feed for 24 h on 6- to 7-week-old Arabidopsis Col-0 plants.The cultivation of plantsand insects has been described previously (Reymond et al., 2000, 2004). The damaged leaveswere harvested for RNA isolation, according to previously published protocols (Reymond et al.,2004).

Microarray experiments and data analysis

Microarray experiments were performed with probes from three independent biological replicasand a custom-made microarray with 292 Arabidopsis cDNAs representing 111 insect-responsivegenes, 17 housekeeping genes, and 70 genes previously demonstrated to be associated withplant defense (Bodenhausen and Reymond, in preparation). Preparation of DNA, printing ofcDNA microarrays, labeling of RNA probes and hybridization of microarrays was carried outas described previously (Reymond et al., 2004).After hybridization, microarrays were scannedwith a ScanArray 4000 (Packard BioScience SA, Zurich, Switzerland). Photomultiplier andlaser power settings were adjusted so that the expression ratio of housekeeping genes was asclose to 1.0 as possible and signal intensities were below saturation of the scanner.The averagefluorescence intensity for each fluor and for each gene was determined using the ImaGeneprogram (BioDiscovery Inc., Los Angeles, CA). Median background fluorescence around eachgene spot was subtracted from each spot. To adjust signal intensities between Cy3 and Cy5channels, a normalization factor was calculated from the expression ratios of 17 housekeepinggenes so that the median expression ratio of these genes was equal to 1.0. This normalization


129

factor was then applied to all background-corrected Cy5 intensities. Signal values <1000 (twoto three times the average background intensity) were raised to 1000 to avoid extreme expressionratios. Normalized signal intensities were used to calculate expression ratios. The individualvariability of each cDNA on the microarray was determined using microarray data of controlvs. control hybridizations of five independent experiments as described (Bodenhausen andReymond, in preparation).

To identify insect-responsive genes, 12 microarray analyses were performed.The followingtreatments were compared on the microarrays: 1) uninfested control plants vs. P. rapae-infestedcontrol plants, 2) uninfested ISR plants vs. P. rapae-infested ISR plants, 3) uninfested controlplants vs. S. exigua-infested control plants, and 4) uninfested ISR plants vs. S. exigua-infestedISR plants. All four comparisons were performed with RNA probes that were derived fromplant material of three independent biological replicates. Uninfested and infested plant materialwas harvested simultaneously 48 h after challenge with the insect herbivores. One additionalmicroarray analysis was performed with RNA probes from uninfested, ISR-expressing vs.uninfested, uninduced control plants.

To identify genes of which the expression was substantially affected by insect herbivoryin either ISR-expressing or uninduced control plants, we applied two selection criteria: 1) thechange in gene expression should be statistically significant according to the Student’s t teston the log2-transformed expression ratios (two sample hypothesis, equal variance; P < 0.05),and 2) the change should be at least 2-fold. This 2-fold cut-off value was previously shownfor this microarray to be well above the variability encountered in biological and technicalreplicates (Reymond et al., 2004).Although statistically significant changes in gene expressionbelow the threshold level of 2-fold may be biologically relevant, we have chosen to disregardall changes below this threshold level to limit the number of non-robust positives.

To identify ISR-primed genes, we selected the genes that showed a statistically significantdifference in expression between herbivore-infested ISR-expressing plants vs. herbivore-infestedcontrol plants (Student’s t test; P < 0.05). Previously, ISR-primed genes were selected basedon >1.5-fold difference in expression level (Verhagen et al., 2004).Therefore, we also selectedthe insect-responsive genes (that already passed selection criteria 1 and 2) that showed a >1.5-folddifference in expression between herbivore-infested ISR-expressing plants vs. herbivore-infestedcontrol plants in at least two out of the three biological replicates.

Normalized raw data for all genes and all experiments are presented as supplementarydata (Supplementary Tables 1 to 3).

130

Supplementary MaterialsSupplementary Table 1: MS Excel file containing micro array data with AGI numbers for eachgene, gene description and NCBI Numbers for each EST, normalized Cy3 and Cy5 values,and Cy5/Cy3 expression ratios. The treatments are uninfested, uninduced control (Cy3) vs.uninduced control plants after challenge with P. rapae or S. exigua (Cy5), and uninfested, ISR-expressing (Cy3) vs. ISR-expressing Arabidopsis after challenge with P. rapae or S. exigua (Cy5).Normalized raw signal intensities are presented for three biological replicates per treatment.

Supplementary Table 2: MS Excel file containing micro array data with AGI numbers for eachgene, gene description and NCBI Numbers for each EST, normalized Cy3 and Cy5 values, andCy5/Cy3 expression ratios. The Lausanne treatments are uninfested control (Cy3) vs. controlplants after challenge for 24 h with P. rapae or S. littoralis (Cy5). The Utrecht treatments areuninfested control (Cy3) vs. control plants after challenge for 48 h with P. rapae or S. exigua(Cy5). Normalized raw signal intensities are presented for two (Lausanne experiment) or three(Utrecht experiment) biological replicates per treatment.

Supplementary Table 3: MS Excel file containing micro array data with AGI numbers for eachgene, gene description and NCBI Numbers for each EST, normalized Cy3 and Cy5 values,and Cy5/Cy3 expression ratios. . The treatments are uninfested, uninduced control (Cy3) vs.uninduced ISR-expressing Arabidopisis plants (Cy5). Normalized raw signal intensities arepresented for one biological experiment per treatment.



131

133

chapter 6

GENERAL DISCUSSION

General discussion Plants are under constant threat of a multitude of pathogenic and herbivorous attackers, andhave developed sophisticated strategies to defend themselves. Resistance against microbes andinsects can be mediated through defenses that are constitutively present, or through defensemechanisms that are induced only upon attack (Van Loon, 2000; Dicke and Van Poecke, 2002).In the regulation of signaling networks involved in induced defense, the plant hormones salicylicacid (SA), jasmonic acid (JA), and ethylene (ET) are major players (Reymond and Farmer, 1998;Pieterse and Van Loon, 1999; Feys and Parker, 2000; Glazebrook, 2001; Thomma et al., 2001;Dicke and Van Poecke, 2002; Kessler and Baldwin, 2002; Spoel et al., 2003).

An important question in plant defense signaling research is: how are plants capable ofintegrating signals induced by pathogenic micro-organisms and herbivorous insects into defensesthat are specifically active against the attacker encountered? To explore the ways in which plantsintegrate, prioritize and coordinate induced defense upon pathogen or insect attack, we studiedthe response of the model plant Arabidopsis thaliana to a range of pathogens and herbivorousinsects, integrating a phytopathological and entomological approach. We monitored the production of the phytohormones SA, JA and ET and the genome-wide transcriptome changesupon attack (De Vos et al., 2005, Chapter 2 and 3).

To fend off insects, plants have evolved two distinct inducible strategies: induced defensedirected against the attacker, referred to as direct defense, and induced defense that exploitsnatural enemies of the attacker, referred to as indirect defense. Both types of defense can betriggered upon insect feeding. Direct defense includes induced responses such as the productionof secondary chemicals or enzymes that act as toxins or feeding deterrents (Kessler and Baldwin,2002; Howe, 2004), whereas indirect defense can involve production of a blend of volatilesthat attracts predatory or parasitic enemies of the herbivores (Turlings et al., 1995; Dicke, 1999).

After pathogen infection, plants can activate the strengthening of cell walls, synthesis ofphytoalexins and the production of pathogenesis-related (PR)-proteins. In addition, upon primaryinfection, plants can induce resistance mechanisms that are effective against subsequentlyinvading pathogens. Two types of systemic induced resistance have been well investigated:pathogen-induced systemic acquired resistance (SAR) (Ross, 1961) and non-pathogenic rhizo-bacteria-mediated induced systemic resistance (ISR) (Pieterse et al., 1996; Pieterse et al., 1998).

To investigate the effectiveness of microbially induced resistance in Arabidopsis on bothdirect and indirect defense against insects, I examined the effect of SAR and ISR on the per-formance of the specialist insect herbivore Pieris rapae and the generalist insect herbivore Spodoptera exigua, and on the attractiveness of the parasitic wasp Cotesia rubecula to volatilesreleased upon herbivory by larvae of these insects (Chapter 4 and 5).

134

Cross-talk between SA, JA and ET in plant defense againstmicrobes and insects

During plant defense, extensive cross-talk and signal integration takes place among theSA-, JA- and ET-dependent signal transduction pathways (Felton and Korth, 2000; Feys andParker, 2000; Pieterse et al., 2001; Kunkel and Brooks, 2002; Rojo et al., 2003; Bostock, 2005).It has been suggested that the interacting signaling pathways enable plants to mount responsesthat are specifically tailored to the attacker encountered, and to optimize the defense responsewhen multiple attackers are present. However, microbial or insect attackers can also manipulatethese signaling pathways to their own benefit (Kloek et al., 2001; Musser et al., 2002). Theregulation of defense-related genes was expected to be affected differentially depending onthe outcome of the signal interactions between SA, JA and ET.

In Chapter 2, we established the degree of commonality in the responses induced inArabidopsis by five different microbial and herbivorous attackers (De Vos et al., 2005). Wemonitored the quantity, composition and timing of the production of SA, JA and ET (signalsignature) upon attack by a pathogenic leaf bacterium (Pseudomonas syringae pv. tomato),a pathogenic leaf fungus (Alternaria brassicicola), tissue-chewing caterpillars (Pieris rapae),cell-content feeding thrips (Frankliniella occidentalis) and phloem-sucking aphids (Myzuspersicae). We found that the signal signature differed strongly, depending on the invader.Local expression of marker genes suggested that insect feeding results in a localized productionof a blend of phytohormones. The specificity of each interaction appears to result from thislimited number of hormones, because they are produced at different sites, in different amountsand combinations, and at different speeds depending on the Arabidopsis-attacker combination.This demonstrates that inclusion of spatial and temporal components is important for under-standing the expression of induced plant defenses.

Using Affymetrix GeneChip microarrays, we analyzed the genome-wide transcriptomechanges upon pathogen or insect attack, as well as upon application of MeJA. Infection withP. syringae induced consistent expression changes of 2,034 genes. The majority of the geneselicited upon attack by A. brassicicola (68%), P. rapae (52%) and F. occidentalis (72%) over-lapped with the P. syringae-induced genes, suggesting that these genes are co-regulated.However, comparison of the transcriptome changes induced by the attackers, except P. syringae,revealed a limited overlap (4-39%), demonstrating that these transcript changes were pre-dominantly specific for each of the three attackers. Thus, the cross-communicating signalingpathways converge only partially on a common set of transcriptional targets related to defense.

In Chapter 3, I investigated the nature of the transcriptional changes induced by all fivedifferent attackers, using MapMan Software.Visualization revealed that, although the overlapof the transcriptome changes between most attacker-combinations was limited, the biologicalprocesses affected were strikingly similar. All five attackers induced transcriptional alterationsmainly in the functional categories “Stress”, “Large enzyme families” and “Secondary meta-bolism”.They all affected similar subgroups, such as the enzyme families “Cytochrome P450s”and “Glutathione S-transferases”. P. syringae and M. persicae both affected many, but non-overlapping genes in the subcategories “Phenylpropanoid” and “Lignin and lignans”, suggestinga differential effect on phenylpropanoid metabolism and cell wall strengthening.Taken together,

Chapter 06: General discussion

135

this indicated that different attackers activate or repress the same enzyme families, metabolicpathways or plant processes by independent transcriptional programs. However, within eachfamily, pathway or process, the genes affected are different, leading to an attacker-specific changein metabolic or physiological activity of the plant. Hence, the cross-talk between SA-, JA- andET-dependent pathways results in the fine-tuned regulation of similar plant processes throughnon-overlapping transcriptional changes.

JA was produced upon attack by P. syringae, A. brassicicola, P. rapae and F. occidentalis,and many of the responding genes (32-69%) were JA-responsive (De Vos et al., 2005, Chapter2). However, analysis of the transcriptome changes revealed that many of the JA-responsivegenes (6-54%) did not overlap and, thus, were attacker-specific. This suggests that attacker-derived factors shape the final outcome of the defense response in Arabidopsis.Visualizationin MapMan revealed that the direction (activation or repression) of these changes was similarto that upon application of MeJA.Thus, despite the limited overlap of the JA-responsive genes,similar plant processes were affected in a similar manner. M. persicae did not induce the pro-duction of JA. Yet, 30% of the induced transcriptome changes involved JA-responsive genes.Infestation with M. persicae induced a predominantly opposite transcriptional response in theaffected plant processes, demonstrating that phloem-sucking aphids induce an atypical responsein comparison with the other attackers.

Recently, a transcriptome analysis was performed with the aim of identifying a coretranscriptional growth-regulatory module in Arabidopsis (Nemhauser et al., 2006). Groups of young seedlings were each treated with one of seven different hormones known to affectgrowth. Genome-wide transcriptome analysis revealed that these hormones did not regulatea common set of transcriptional targets. Instead, distinct members of protein families wereregulated in a hormone-specific manner, demonstrating that these hormones regulate similarprocesses through largely non-overlapping transcriptional responses. The authors argue thatthe observed hormone specificity may be due to the accumulation of changes in regulatoryregions among genes with identical functions, providing the plant with a means to fine-tunegrowth in response to multiple input pathways. This interesting idea might be applied to ourwork on the integration of signals from the cross-communicating SA-, JA- and ET-dependentsignaling pathways during Arabidopsis defense. Even upon challenge with attackers that inducesimilar signal signatures, the transcriptional changes were predominantly attacker-specific (De Vos et al., 2005, Chapter 2). Yet, they affect surprisingly similar processes (Chapter 3),suggesting that Arabidopsis has evolved highly sophisticated mechanisms to regulate defense.

Cross-resistance between pathogens and insects

Little is known about the spectrum of effectiveness of the different types of induced resistancethat are expressed upon attack by pathogens or insects. Is herbivore-induced resistance specificallyeffective against insect herbivores, or does it also affect pathogens? Is microbially inducedresistance specifically directed against microbial pathogens, or are insect herbivores affectedas well? During the collaboration between the Laboratory of Entomology,Wageningen Universityand the Section Phytopathology, Utrecht University, Martin de Vos (De Vos, 2006) focused on

136

the first, while I tried to answer the latter question (this thesis). The effects of microbiallyinduced defense on insects and of herbivore-induced resistance against pathogens are difficultto predict (reviewed in Stout et al., 2006). Due to variation in the choice of plant, pathogenand insect species, the results of the different studies varied, making it difficult to draw generalconclusions about the response of one plant to multiple attackers.

We hypothesized that a thorough analysis of a single model plant and of the signalingpathways induced after attack by various pathogens and insects would enable us to betterpredict whether or not cross-resistance will occur. Attackers that induced a similar signal sig-nature were expected to induce a defense response with a comparable spectrum of effective-ness. Therefore, grazing by P. rapae larvae, found to mainly stimulate the production of JA(Chapter 2), was hypothesized to induce protection against A. brassicicola, a fungus that issensitive to JA-dependent defenses (Thomma et al., 1998). However, P. rapae-induced resistancewas not effective against this fungus (De Vos et al., 2006). Instead, Arabidopsis previouslyinfested with P. rapae is more resistant to the biotrophic pathogen Turnip crinkle virus (TCV).This is unexpected, because resistance against TCV requires SA, but not JA (Kachroo et al.,2000).Analysis of SA-induced PR-1 expression revealed that infestation with P. rapae sensitizedArabidopsis leaves for enhanced expression of SA-dependent defenses (De Vos et al., 2006).This phenomenon is called “priming” and had been demonstrated to underlie different typesof induced resistance (Conrath et al., 2002; 2006). Primed plants are capable of activatingdefense responses in a faster and stronger manner upon attack by pathogens, insects or inresponse to abiotic stress. The finding that P. rapae feeding primes Arabidopsis for enhancedSA-dependent defenses demonstrates that the insect-induced defense response is surprisinglycomplex.

Microbially induced defense: SAR and ISROnce plants have been attacked and have effectively resisted the invader, they develop anenhanced defensive capacity against further attack.Two inducible resistance mechanisms againstpathogens have been well investigated: pathogen-induced systemic acquired resistance (SAR)(Ross, 1961) and non-pathogenic rhizobacteria-mediated induced systemic resistance (ISR)(Pieterse et al., 1996; 1998). Both types of induced resistance are expressed systemically through-out the plant and are effective against a broad, yet partly distinct, range of pathogens. SAR ispredominantly effective against pathogens that are resisted by SA-dependent basal resistancemechanisms in uninduced plants, whereas ISR is effective against pathogens that are resistedthrough JA- and ET-dependent basal defenses (Ton et al., 2002). Unlike SAR, rhizobacteria-mediated ISR in Arabidopsis is not associated with the induction of SA and PR gene expression(Pieterse et al., 1996). Instead, ISR requires responsiveness to both JA and ET, but the levels of these hormones remain unchanged (Pieterse et al., 1998; 2000). Transcriptome analysis didnot reveal changes in gene expression in the leaves after establishment of ISR (Verhagen et al.,2004). However, upon pathogen challenge, ISR-expressing Arabidopsis showed potentiatedexpression of JA-responsive defense genes, indicating that these plants were primed (Van Weeset al., 1999; Verhagen et al., 2004).Thus, the observed effectiveness of ISR against pathogensthat are resisted through JA- and ET-dependent basal defenses in uninduced plants is basedon the priming for JA- and ET-dependent defenses (Ton et al., 2002).


137

Effect of microbially induced plant defense on direct defense against insectsThe spectrum of effectiveness of SAR and ISR against pathogens has been well investigated(Sticher et al., 1997; Ton et al., 2002). However, relatively little attention has been paid to theeffect of these two types of induced resistance against insects. In Chapter 4, I investigated theeffectiveness of SAR and ISR against the tissue-chewing caterpillars P. rapae, a specialist oncrucifers, and Spodoptera exigua, a generalist herbivore. Forced feeding experiments revealedthat induction of SAR and ISR significantly reduced growth and development of S. exigua,whereas the performance P. rapae was unaffected (Fig. 1). Using NahG, S-12 and coi1-16plants with increased or decreased production of JA, or insensitivity to JA, respectively, itwas shown that the JA-dependent defense response is effective against both P. rapae and S. exigua (Fig. 1). However, the effect on P. rapae was apparent only in coi1-16 plants, com-pletely blocked in JA-signaling, indicating that this specialist insect is less sensitive to JA-dependent defense than the generalist insect.Thus, the lack of effect of SAR and ISR on P. rapaelarvae can be explained by their relative insensitivity to JA-dependent defenses.

The effectiveness of ISR against S. exigua was expected, because this type of inducedresistance is effective against attackers that are sensitive to JA-dependent defenses (Ton et al.,2002; Verhagen et al., 2006). Since SA has an antagonistic effect on JA-responsive geneexpression (Spoel et al., 2003), the effectiveness of SAR against S. exigua was unexpected.Induction of SAR by predisposal infection with avirulent P. syringae pv. tomato DC3000(avrRpt2) results not only in enhanced SA levels, but also in increased production of both JA and ET (De Vos et al., 2005, Chapter 2), suggesting that the effectiveness might be a resultof the increase in the levels of JA.Thus, SAR- and ISR-expressing Arabidopsis with increasedSA-, JA-, and ET-levels or enhanced JA- and ET-sensitivity, respectively, are effective againstattackers that are sensitive to JA-dependent defense.

138

FIGURE 1. Effectiveness of microbially induced ISR and SAR against S. exigua, but not P. rapae, is associated withpriming of defense genes (Chapter 4 and 5).

Effect of microbially induced plant defense on indirect defense against insectsPlants may release a blend of volatiles upon herbivory, leading to the attraction of parasiticor predatory enemies of the attacker (Turlings et al., 1995; Dicke, 1999). The production ofgreen leaf volatiles and JA is induced upon herbivory in Arabidopsis (Van Poecke et al., 2001)and other plants (reviewed in Paré and Tumlinson, 1999). In Arabidopsis, the decreased attractive-ness of S-12 and NahG transgenic plants to the parasitic wasp C. rubecula demonstrated thatboth JA and SA play a role in indirect defense (Van Poecke and Dicke, 2002). In various plants,herbivore-induced volatiles have been implicated in priming for enhanced defense (Kessler et al.,2006; Paschold et al., 2006; Ton et al., 2007), such as enhanced synthesis of GLVs and JAupon subsequent wounding (Engelberth et al., 2004). Because both SAR and ISR are associatedwith priming (Conrath et al., 2006), we expected that both types of microbially induced defensewould affect the indirect defense response that is triggered by insect herbivory. In a windtunnelset-up, I tested the effect of microbially induced SAR and ISR on herbivore-induced attractionof C. rubecula. Upon feeding by either S. exigua or P. rapae, Arabidopsis plants became significantly more attractive to the parasitic wasps than undamaged control plants. However,induction of either SAR or ISR did not significantly enhance the attractiveness in comparisonto uninduced plants, demonstrating that these types of microbially induced resistance did not contribute significantly to this herbivore-induced indirect defense response in Arabidopsis.Because we showed that both SAR and ISR were primed for JA-dependent defense-relatedgene expression upon challenge with S. exigua (Chapter 4), this result was unexpected.

No information is available about how microbially induced SAR and ISR affect the

139


emission of green leaf volatiles. Herbivore-induced volatile emission is dependent on thedensity and the duration of the infestation and is primed by JA in Lima bean plants (Gols et al., 2003). In Arabidopsis, prolongation of the infestation with a single L1 P. rapae larvawas shown to enhance the attractiveness to C. rubecula for up to six days (Van Poecke et al.,2001).We noticed that S. exigua larvae caused far less damage than P. rapae caterpillars. Toincrease the attractiveness of S. exigua-infested Arabidopsis to C. rubecula, we used a relativelylong duration (72 h) and high caterpillar density (10 L1 larvae per plant) for the caterpillarinfestation.These conditions correlated with enhanced expression of the defense-related markergenes PDF1.2 and HEL, indicating that the enhanced defense response had been activated by72 h. We did not study the expression of genes involved in volatile synthesis, nor did wedetermine the volatiles released. Thus, we cannot conclude whether SAR- or ISR-expressingplants emitted different volatile blends than uninduced plants after caterpillar feeding.

Recently, exposure to herbivore-induced volatiles was shown to prime maize forenhanced volatile emission after challenge with Spodoptera littoralis larvae (Ton et al., 2007).The increased volatile production rendered the plants more attractive to parasitic Cotesiamarginiventris wasps, demonstrating that maize can be primed for enhanced indirect defense.Using a set-up in which volatile emission could be coupled directly to the behavior of theparasitic wasps, it was noticed that the enhanced attraction was apparent only between 180and 300 minutes after infestation. Shorter or longer infestation did not lead to increasedeffectiveness. The period during which enhanced attraction occurred correlated with an aug-mented production of aromatic and terpenoid compounds, suggesting that the enhanced emissionof volatiles was responsible for the enhanced attraction of the parasitoid.The volatile-inducedpriming was also demonstrated to affect direct defense.These results demonstrate that primingby airborne signals can boost both direct and indirect defense in maize.

The limited period during which enhanced attraction of parasitoid wasps to herbivore-infested, volatile-primed maize plants occurred (Ton et al., 2007), suggests that the lack ofenhanced attractiveness of C. rubecula to S. exigua-infested, ISR-expressing plants (Chapter 4)was due to a feeding period that was too long to observe a priming effect.To further investigatewhether ISR-expressing Arabidopsis is primed for indirect defense upon challenge with S. exigua,the plants should be infested for a shorter period. In order to study the relationship betweenthe attractiveness to, and the volatile release from, primed plants upon herbivory, a set-upshould be used that allows the simultaneous identification of volatile emission and parasitoidbehavior.

Priming for enhanced defense gene expression

Priming of pathogen-induced genes probably allows a plant to react more effectively to theinvader encountered, because once induced the plant can respond faster and stronger topathogen attack (Conrath et al., 2006). In Chapter 4, I showed that rhizobacteria-mediatedISR is effective against the generalist herbivore S. exigua, but not against P. rapae.The effective-ness was associated with potentiated expression of the JA- and ET-responsive genes PDF1.2and HEL in SAR- and ISR-expressing plants upon feeding by S. exigua, but not upon feeding

140

by P. rapae. These results suggest that the effectiveness of microbially induced SAR and ISRagainst S. exigua is associated with priming for enhanced defense-related gene expression (Fig. 1).

In Chapter 5, a dedicated microarray containing 111 genes responsive to caterpillar feedingwas used to survey the transcriptome of uninduced and ISR-expressing Arabidopsis upon P. rapae and S. exigua feeding.An additional set of ISR-primed genes was identified, extendingour notion that ISR-expressing Arabidopsis is primed for augmented gene expression after insectattack. However, the number of ISR-primed genes (5) associated with the observed effectivenesswas rather small (Chapter 5). Although the percentage of primed genes was comparable inprevious studies (Verhagen et al., 2004; Pozo et al., in preparation), the question remains ifthe augmented expression of these five genes plays indeed a role in the reduced growth of S. exigua larvae on ISR-expressing plants.

The proposed mechanism for ISR is priming of JA/ET-responsive genes. Upon challenge,these genes are expressed faster and to higher levels in ISR-expressing than in uninduced controlplants, leading to enhanced resistance.Analysis of the promoters of the primed genes revealedthat the binding site CACATG for the transcription factor AtMYC2 was overrepresented,suggesting a regulatory function for AtMYC2 in ISR (Pozo et al., in preparation). Togetherwith the transcription factor ERF1, MYC2 was previously demonstrated to integrate signalsfrom the JA and ET pathways in activating JA- and ET-responsive defense genes (Lorenzo et al., 2003; 2004). Gene expression analysis showed elevated MYC2 mRNA levels in ISR-expressing Arabidopsis before pathogen challenge (Pozo et al., in preparation). Moreover,subsequent analysis of the Arabidopsis MYC2 mutant jin1 demonstrated that effective ISRagainst the pathogens P. syringae pv. tomato and H. parasitica could not be established,providing further evidence for a role of MYC2 in ISR (Pozo et al., in preparation). The set of five genes with augmented expression in ISR-expressing plants after feeding by S. exiguaalso contained the CACATG site in their promoters.To test whether the potentiated expressionpattern is responsible for the effect on the performance of S. exigua larvae, the MYC2 mutantjin1 should be used and show a lack of effect.

The microarray analysis revealed that not all 111 caterpillar-responsive genes were inducedin the Utrecht experiment (Chapter 5). It is possible that the target genes are not on the dedicatedmicroarray, because the genes were selected for an early response (3-5 h) after feeding byL4-L5 P. rapae larvae, whereas we tested plants after challenge for 48 h with L1 caterpillars.Using L1 larvae, the majority of the genes induced in the Utrecht experiment were inducedalso in the Lausanne experiment. Many additional genes were induced in the Lausanneexperiment, which suggests that genes induced at early time points are overrepresented onthe microarray. Therefore, the set of five S. exigua-induced, ISR-primed genes is likely to bean underestimate. Performing full genome microarrays will shed more light on the changesoccurring and is likely to strengthen our observations in Chapter 5.

Generalist vs. specialist insect herbivores

In Chapter 4, we showed that microbially induced SAR and ISR in Arabidopsis are effectiveagainst the generalist insect S. exigua, but not against the specialist P. rapae. Several previous

141


studies have demonstrated that specialist and generalist insects have different sensitivities to plant defenses (Karban and Agrawal, 2002; Stotz et al., 2002; Reymond et al., 2004).Specialists are usually rather insensitive, because they have adapted to detoxify plant metabolitesand suppress plant defense responses (Agrawal, 1999).The specialist herbivore Manduca sextawas shown to inhibit wound-inducible nicotine production in tobacco (Nicotiana attenuata)(Winz and Baldwin, 2001), demonstrating that this insect can suppress the host defense response.In Arabidopsis, the regurgitate of the specialist insect P. rapae was shown to have a suppressiveeffect on several genes, when compared to mechanical wounding (Reymond et al., 2000; DeVos, 2006), indicating that a specialist insect can suppress the induction of host plant defensethrough elicitors in the insect regurgitate. However, these studies did not include experimentswith a generalist insect, leaving the question unanswered whether the suppressive effect ofelicitors in the regurgitate is specific for specialist insects.

In tobacco, a different transcriptional response is elicited upon infestation with oligophagousM. sexta and polyphagous Heliothis virescens and S. exigua, demonstrating that the trans-criptome can be altered differentially by specialist and generalist insects (Voelckel and Baldwin,2004). In contrast, Reymond et al. (2004) observed a conserved transcriptional pattern inArabidopsis after feeding by the specialist P. rapae and the generalist S. littoralis. In Chapter 4and 5, it was demonstrated that ISR-expressing Arabidopsis plants challenged with the specialist herbivore P. rapae or the generalist insect S. exigua show different transcriptionalalterations compared to uninduced plants in response to these two herbivore species.The geneswith potentiated expression patterns upon challenge with S. exigua were not primed uponinfestation with P. rapae. Moreover, all P. rapae-induced changes in gene expression wererepressed in ISR-expressing plants. To investigate if the transcriptional changes are indeeddependent on the degree of specialization of each insect, an additional set of specialists andgeneralists should be tested. Future research should reveal whether these differences are causedby different elicitors in the regurgitate of these insect herbivores.

Conclusion

This thesis, in combination with the thesis by Martin de Vos (2006), provides a comprehensivestudy on the induced defense responses of the model plant Arabidopsis thaliana upon attackby a range of pathogens and insects.We showed that the transcriptional alterations were pre-dominantly attacker-specific, but the processes affected surprisingly similar. Despite knowledgeabout the induced signal signature and transcript changes, the cross-effectiveness of microbiallyinduced defense on insects and of herbivore-induced resistance against pathogens remaineddifficult to predict (Chapter 4, De Vos et al., 2006). Both the effectiveness of microbially inducedresistance against S. exigua and the P. rapae-induced effectiveness against TCV were associatedwith priming (De Vos et al., 2006, Chapter 4). The combined phytopathological and entomo-logical approach has moved us closer to understanding how plants react to different attackersto defend themselves. To understand the expressed phenotype, knowledge of the underlyingmechanisms is required. Therefore, future research on this exciting topic should be focusedon the unraveling of the regulatory mechanisms of cross-communicating defense signaling

142

pathways in parallel with studies about their effectiveness against different types of attackers.This combined knowledge will enable us to predict and exploit the outcome of specific plant-attacker interactions.

143


ReferencesABE, H., YAMAGUCHI-SHINOZAKI, K., URAO, T., IWASAKI, T., HOSOKAWA, D., AND SHINOZAKI, K. (1997). Roleof Arabidopsis MYC and MYB homologs in drought- and abscisic acid-regulated gene expression. Plant Cell 9: 1859-1868.AGRAWAL, A.A. (1999). Induced responses to herbivory in wild radish: Effects on several herbivores and plant fitness.Ecology 80: 1713-1723.AGRAWAL, A.A., AND KURASHIGE, N.S. (2003). A role for isothiocyanates in plant resistance against the specialistherbivore Pieris rapae. J. Chem. Ecol. 29: 1403-1415.AGRIOS, G.N. (1997). Plant Pathology. (San Diego, USA: Academic Press Inc.).AHN, I.-P., PARK, K., AND KIM, C.-H. (2002). Rhizobacteria-induced resistance perturbs viral disease progress and triggersdefense-related gene expression. Mol. Cells 13: 302-308.AJLAN, A.M., AND POTTER, D.A. (1992). Lack of effect of tobacco mosaic virus-induced systemic acquired resistanceon arthropod herbivores in tobacco. Phytopathol. 82: 647-651.AMENT, K., KANT, M.R., SABELIS, M.W., HARING, M.A., AND SCHUURINK, R.C. (2004). Jasmonic acid is a key regulatorof spider mite-induced volatile terpenoid and methyl salicylate emission in tomato. Plant Physiol. 135: 2025-2037.ARIMURA, G., OZAWA, R., SHIMODA, T., NISHIOKA, T., BOLAND, W., AND TAKABAYASHI, J. (2000). Herbivory-inducedvolatiles elicit defence genes in lima bean. Nature 406: 512-515.ARONDEL, V., VERGNOLLE, C., CANTREL, C., AND KADER, J.C. (2000). Lipid transfer proteins are encoded by a smallmultigene family in Arabidopsis thaliana. Plant Science 157: 1-12.BECKERS, G.J.M., AND SPOEL, S.H. (2006). Fine-tuning plant defence signalling: salicylate versus jasmonate. PlantBiol. 8: 1-10.BELL, E., CREELMAN, R.A., AND MULLET, J.E. (1995). A chloroplast lipoxygenase is required for wound-induced accumulation of jasmonic acid in Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 92: 8675-8679.BERROCAL-LOBO, M., AND MOLINA, A. (2004). Ethylene response factor 1 mediates Arabidopsis resistance to thesoilborne fungus Fusarium oxysporum. Mol. Plant-Microbe Interact. 17: 763-770.BODENHAUSEN, N., AND REYMOND, P. (in preparation). Signaling pathways controlling induced resistance to insectherbivores in Arabidopsis.BOSTOCK, R.M. (2005). Signal crosstalk and induced resistance: Straddling the line between cost and benefit. Annu. Rev.Phytopathol. 43: 545-580.BROEKAERT, W.F., TERRAS, F.R.G., CAMMUE, B.P.A., AND VANDERLEYDEN, J. (1990) An automated quantitative assayfor fungal growth. FEMS (Fed. Eur. Microbiol. Soc.) Microbiol. Lett. 69: 55-60.CAO, H., BOWLING, S.A., GORDON, A.S., AND DONG, X. (1994). Characterization of an Arabidopsis mutant that isnonresponsive to inducers of systemic acquired resistance. Plant Cell 6: 1583-1592.CAMERON, R.K., PAIVA, N.C., LAMB, C.J., AND DIXON, R.A. (1999). Accumulation of salicylic acid and PR gene transcriptsin relation to the systemic acquired resistance (SAR) response by Pseudomonas syringae pv. tomato in Arabidopsis.Physiol. Mol. Plant Pathol. 55: 121-130.CARDOZA, Y.J., TEAL, P.E.A., AND TUMLINSON, J.H. (2003a). Effect of peanut plant fungal infection on ovipositionpreference by Spodoptera exigua and on host-searching behavior by Cotesia marginiventris. Environm. Entomol. 32: 970-976.CARDOZA, Y.J., LAIT, C.G., SCHMELZ, E.A., HUANG, J., AND TUMLINSON, J.H. (2003b). Fungus-induced biochemicalchanges in peanut plants and their effect on development of beet armyworm, Spodoptera exigua Hubner (Lepidoptera :Noctuidae) larvae. Environm. Entomol. 32: 220-228.CHAPPLE, C. (1998). Molecular-genetic analysis of plant cytochrome P450-dependent monooxygenases. Annu. Rev. PlantPhysiol. Plant Mol. Biol. 49: 311-343.CHEN, H., WILKERSON, C.G., KUCHAR, J.A., PHINNEY, B.S., AND HOWE, G.A. (2005). Jasmonate-inducible plant enzymesdegrade essential amino acids in the herbivore midgut. Proc. Natl. Acad. Sci. U.S.A. 102: 19237-19242.CHOH, Y., SHIMODA, T., OZAWA, R., DICKE, M., AND TAKABAYASHI, J. (2004). Exposure of lima bean leaves to volatilesfrom herbivore-induced conspecific plants results in emission of carnivore attractants: active or passive process? J. Chem.Ecol. 30: 1305-1317.CIPOLLINI, D., ENRIGHT, S., TRAW, M.B., AND BERGELSON, J. (2004). Salicylic acid inhibits jasmonic acid-inducedresistance of Arabidopsis thaliana to Spodoptera exigua. Mol. Ecol. 13: 1643-1653.COLMENARES, A.J., ALEU, J., DURAN-PATRON, R., COLLADO, I.G., AND HERNANDEZ-GALAN, R. (2002). The putative role of botrydial and related metabolites in the infection mechanism of Botrytis cinerea. J. Chem. Ecol. 28:997-1005.CONA, A., REA, G., ANGELINI, R., FEDERICO, R., AND TAVLADORAKI, P. (2006). Functions of amine oxidases in plantdevelopment and defence. Trends Plant Sci. 11: 80-88.CONRATH, U., PIETERSE, C.M.J., AND MAUCH-MANI, B. (2002). Priming in plant-pathogen interactions. Trends PlantSci. 7: 210-216.CONRATH, U., BECKERS, G.J.M., FLORS, V., GARCIA-AGUSTIN, P., JAKAB, G., MAUCH, F., NEWMAN, M.A.,PIETERSE, C.M.J., POINSSOT, B., POZO, M.J., PUGIN, A., SCHAFFRATH, U., TON, J., WENDEHENNE, D., ZIMMERLI, L.,MAUCH-MANI, B., AND GRP, P.-A.-P. (2006). Priming: Getting ready for battle. Mol. Plant-Microbe Interact. 19: 1062-1071.CUI, J., JANDER, G., RACKI, L.R., KIM, P.D., PIERCE, N.E., AND AUSUBEL, F.M. (2002). Signals involved in Arabidopsisresistance to Trichoplusia ni caterpillars induced by virulent and avirulent strains of the phytopathogen Pseudomonassyringae. Plant Physiol. 129: 551-564.

144

A

B

C

CZECHOWSKI, T., BARI, R.P., STITT, M., SCHEIBLE, W.R., AND UDVARDI, M.K. (2004) Real-time RT-PCR profiling ofover 1400 Arabidopsis transcription factors: unprecedented sensitivity reveals novel root- and shoot-specific genes. PlantJ. 38: 366-379.DANGL, J.L. AND JONES, J.D.G. (2001) Plant pathogens and integrated defence responses to infection. Nature 411:826-833.DAVIES, F.T., HE, C.J., CHAU, A., HEINZ, K.M., AND CARTMILL, A.D. (2004) Fertility affects susceptibility of chrysanthemumto cotton aphids: Influence on plant growth, photosynthesis, ethylene evolution, and herbivore abundance. J. Am. Soc.Hortic. Sci. 129: 344-353.DAVIES, R.T., GOETZ, D.H., LASSWELL, J., ANDERSON, M.N., AND BARTEL, B. (1999). IAR3 encodes an auxin conjugatehydrolase from Arabidopsis. Plant Cell 11: 365-376.DE BOER, J.G., AND DICKE, M. (2004). The role of methyl salicylate in prey searching behavior of the predatory mitePhytoseiulus persimilis. J. Chem. Ecol. 30: 255-271.DE LAAT, A.M.M. AND VAN LOON, L.C. (1982) Regulation of ethylene biosynthesis in virus-infected tobacco leaves. II.Time course of levels of intermediates and in vivo conversion rates. Plant Physiol. 69: 240-245.DE MEYER, G., AUDENAERT, K., AND HÖFTE, M. (1999). Pseudomonas aeruginosa 7NSK2-induced systemic resistancein tobacco depends on in planta salicylic acid accumulation but is not associated with PR1a expression. Eur. J. Plant Pathol.105: 513-517.DE VOS, M. (2006). Signal signature, transcriptomics, and effectiveness of induced pathogen and insect resistance inArabidopsis (Ph. D. Thesis, Utrecht University, The Netherlands), pp. 145.DE VOS, M., VAN OOSTEN, V.R., VAN POECKE, R.M.P., VAN PELT, J.A., POZO, M.J., MUELLER, M.J., BUCHALA,A.J., MÉTRAUX, J.-P., VAN LOON, L.C., DICKE, M., AND PIETERSE, C.M.J (2005). Signal signature and transcriptomechanges of Arabidopsis during pathogen and insect attack. Mol. Plant-Microbe Interact. 18:, 923-937.DE VOS, M., VAN ZAANEN, W., KOORNNEEF, A., KORZELIUS, J.P., DICKE, M., VAN LOON, L.C., AND PIETERSE,C.M.J. (2006). Herbivore-induced resistance against microbial pathogens in Arabidopsis. Plant Physiol. 142: 352-363.DELANEY, T.P., FRIEDRICH, L., AND RYALS, J.A. (1995). Arabidopsis signal transduction mutant defective in chemicallyand biologically induced disease resistance. Proc. Natl. Acad. Sci. U.S.A. 92: 6602-6606.DELANEY, T.P., UKNES, S., VERNOOIJ, B., FRIEDRICH, L., WEYMANN, K., NEGROTTO, D., GAFFNEY, T., GUT-RELLA,M., KESSMANN, H., WARD, E., AND RYALS, J. (1994) A central role of salicylic acid in plant disease resistance.Science 266: 1247-1250.DICKE, M. (1999). Evolution of induced indirect defence of plants. In The Ecology and Evolution of Inducible Defenses,R. Tollrian and C.D. Harvell, eds (Princeton: Princeton University Press), pp. 62-88.DICKE, M., GOLS, R., LUDEKING, D., AND POSTHUMUS, M.A. (1999). Jasmonic acid and herbivory differentially inducecarnivore-attracting plant volatiles in lima bean plants. J. Chem. Ecol. 25: 1907-1922.DICKE, M., AND HILKER, M. (2003) Induced plant defences: from molecular biology to evolutionary ecology. Basic Appl.Ecol. 4: 3-14.DICKE, M., SABELIS, M.W., TAKABAYASHI, J., BRUIN, J., AND POSTHUMUS, M.A. (1990). Plant strategies of manipulatingpredator-prey interactions through allelochemicals: prospects for application in pest control. J. Chem. Ecol. 16: 3091-3118.DICKE, M. AND VAN POECKE, R.M.P. (2002) Signaling in plant-insect interactions: Signal transduction in direct andindirect plant defence. In Plant Signal Transduction: Frontiers in Molecular Biology, D. Scheel and C. Wasternack, eds,.(Oxford: Oxford University Press), pp. 289-316.DOARES, S.H., NARVAEZVASQUEZ, J., CONCONI, A., AND RYAN, C.A. (1995). Salicylic acid inhibits synthesis ofproteinase-inhibitors in tomato leaves induced by systemin and jasmonic acid. Plant Physiol. 108: 1741-1746.DOHERTY, H.M., SELVENDRAN, R.R., AND BOWLES, D.J. (1988). The wound response of tomato plants can be inhibitedby aspirin and related hydroxy-benzoic acids. Physiol. Mol. Plant Pathol. 33: 377-384.DONG, J.X., CHEN, C.H., AND CHEN, Z.X. (2003). Expression profiles of the Arabidopsis WRKY gene superfamily duringplant defense response. Plant Mol. Biol. 51: 21-37.DOUGLAS, A.E. (2006). Phloem-sap feeding by animals: problems and solutions. J. Exp. Bot. 57: 747-754.DURRANT, W.E., AND DONG, X.N. (2004). Systemic acquired resistance. Annu. Rev. Phytopathol. 42: 185-209.ELLIS, C., KARAFYLLIDIS, L., AND TURNER, J.G. (2002) Constitutive activation of jasmonate signaling in an Arabidopsismutant correlates with enhanced resistance to Erysiphe cichoracearum, Pseudomonas syringae, and Myzus persicae.Mol. Plant-Microbe Interact. 15: 1025-1030.ELLIS, C., AND TURNER, J.G. (2002). A conditionally fertile coi1 allele indicates cross-talk between plant hormone signalingpathways in Arabidopsis thaliana seeds and young seedlings. Planta 215: 549-556.ENGELBERTH, J., ALBORN, H.T., SCHMELZ, E.A., AND TUMLINSON, J.H. (2004). Airborne signals prime plants againstinsect herbivore attack. Proc. Natl. Acad. Sci. U.S.A. 101: 1781-1785.EULGEM, T., RUSHTON, P.J., ROBATZEK, S., AND SOMSSICH, I.E. (2000). The WRKY superfamily of plant transcriptionfactors. Trends Plant Sci. 5: 199-206.FACCHINI, P.J. (2001). Alkaloid biosynthesis in plants: Biochemistry, cell biology, molecular regulation, and metabolicengineering applications. Annu. Rev. Plant Physiol. Plant Mol. Biol. 52: 29-66.FAN, W.H., AND DONG, X.N. (2002). In vivo interaction between NPR1 and transcription factor TGA2 leads to salicylicacid-mediated gene activation in Arabidopsis. Plant Cell 14: 1377-1389.FARMER, E.E., AND RYAN, C.A. (1992). Octadecanoid precursors of jasmonic acid activate the synthesis of wound-inducibleproteinase inhibitors. Plant Cell 4: 129-134.FARMER, E.E., ALMERAS, E., AND KRISHNAMURTHY, V. (2003). Jasmonates and related oxylipins in plant responses topathogenesis and herbivory. Curr. Opin. Plant Biol. 6: 372-378.

145

References

D

E

F

FELTON, G.W. AND KORTH, K.L. (2000) Trade-offs between pathogen and herbivore resistance. Curr. Opin. Plant Biol.3: 309-314.FELTON, G.W., KORTH, K.L., BI, J.L., WESLEY, S.V., HUHMAN, D.V., MATHEWS, M.C., MURPHY, J.B., LAMB, C.,AND DIXON, R.A. (1999). Inverse relationship between systemic resistance of plants to microorganisms and to insectherbivory. Curr. Biol. 9: 317-320.FEYS, B.J. AND PARKER, J.E. (2000) Interplay of signaling pathways in plant disease resistance. Trends Genet.16: 449-455.GAFFNEY, T., FRIEDRICH, L., VERNOOIJ, B., NEGROTTO, D., NYE, G., UKNES, S., WARD, E., KESSMANN, H., AND RYALS, J. (1993). Requirement of salicylic acid for the induction of systemic acquired resistance. Science 261:754-756.GARCÍA-OLMEDO, F., MOLINA, A., SEGURA, A., AND MORENO, M. (1995). The defensive role of nonspecific lipid-transfer proteins in plants. Trends Microbiol. 3: 72-74.GEERVLIET, J.B.F., VET, L.E.M., AND DICKE, M. (1994). Volatiles from damaged plants as major cues in long-rangehost-searching by the specialist parasitoid Cotesia rubecula. Entomol. Exp. Appl. 73: 289-297.GLAZEBROOK, J. (2001) Genes controlling expression of defense responses in Arabidopsis - 2001 status. Curr. Opin.Plant Biol. 4: 301-308.GLAZEBROOK, J. (2005). Contrasting mechanisms of defense against biotrophic and necrotrophic pathogens. Annu.Rev. Phytopathol. 43: 205-227.GLAZEBROOK, J. AND AUSUBEL, F.M. (1994) Isolation of phytoalexin-deficient mutants of Arabidopsis thaliana andcharacterization of their interactions with bacterial pathogens. Proc. Natl. Acad. Sci. U.S.A. 91: 8955-8959.GLAZEBROOK, J., CHEN, W., ESTES, B., CHANG, H.-S., NAWRATH, C., MÉTRAUX, J.-P., ZHU, T., AND KATAGIRI,F. (2003) Topology of the network integrating salicylate and jasmonate signal transduction derived from global expressionphenotyping. Plant J. 34: 217-228.GOLS, R., POSTHUMUS, M.A., AND DICKE, M. (1999). Jasmonic acid induces the production of gerbera volatiles thatattract the biological control agent Phytoseiulus persimilis. Entomol. Exp. Appl. 93: 77-86.GOLS, R., ROOSJEN, M., DIJKMAN, H., AND DICKE, M. (2003). Induction of direct and indirect plant responses byjasmonic acid, low spider mite densities, or a combination of jasmonic acid treatment and spider mite infestation. J. Chem.Ecol. 29: 2651-2666.HALKIER, B.A., AND GERSHENZON, J. (2006). Biology and biochemistry of glucosinolates. Annu. Rev. Plant Biol. 57:303-333.HEIL, M., AND KOST, C. (2006). Priming of indirect defences. Ecol. Lett. 9: 813-817.HELLE, W., AND SABELIS, M.W. (1985). Spider mites: Their biology, natural enemies, and control. (Amsterdam: Elsevier).HOAGLAND, D.R. AND ARNON, D.I. (1938) The water culture method for growing plants without soil. Calif. Agric.Exp. Stn. Bull. 347: 36-39.HOFFMAN, T., SCHMIDT, J.S., ZHENG, X., AND BENT, A.F. (1999) Isolation of ethylene-insensitive soybean mutantsthat are altered in pathogen susceptibility and gene-for-gene disease resistance. Plant Physiol. 119: 935-949.HORIUCHI, J., ARIMURA, G., OZAWA, R., SHIMODA, T., TAKABAYASHI, J., AND NISHIOKA, T. (2001). ExogenousACC enhances volatiles production mediated by jasmonic acid in lima bean leaves. FEBS Letters 509: 332-336.HOWE, G.A. (2004). Jasmonates as signals in the wound response. J. Plant Growth Regul. 23: 223-237.HOWE, G.A., LIGHTNER, J., BROWSE, J., AND RYAN, C.A. (1996) An octadecanoid pathway mutant (JL5) of tomatois compromised in signaling for defense against insect attack. Plant Cell 8: 2067-2077.JAMESON, P.E. (2000). Cytokinins and auxins in plant-pathogen interactions - an overview. J. Plant Growth Regul. 32:369-380.JONGEBLOED, P.H.J., ELGERSMA, D.M., AND SABELIS, M.W. (1992). Does a vascular fungus of tomato induce adefence response or a change in host plant quality that also affects the oviposition of spider mites? Exp. Appl. Acarol.16: 227-236.KACHROO, P., YOSHIOKA, K., SHAH, J., DOONER, K.D., AND KLESSIG, D.F. (2000). Resistance to turnip crinklevirus in Arabidopsis is regulated by two host genes and is salicylic acid dependent but NPR1, ethylene, and jasmonateindependent. Plant Cell 12: 677-690.KADER, J.C. (1997). Lipid-transfer proteins: A puzzling family of plant proteins. Trends Plant Sci. 2: 66-70.KAHL, J., SIEMENS, D.H., AERTS, R.J., GABLER, R., KUHNEMANN, F., PRESTON, C.A., AND BALDWIN, I.T. (2000)Herbivore-induced ethylene suppresses a direct defense but not a putative indirect defense against an adapted herbivore.Planta 210: 336-342.KANNO, H., AND FUJITA, Y. (2003). Induced systemic resistance to rice blast fungus in rice plants infested by white-backed planthopper. Entomol. Exp. Appl. 107, 155-158.KANNO, H., SATOH, M., KIMURA, T., AND FUJITA, Y. (2005). Some aspects of induced resistance to rice blast fungus,Magnaporthe grisea, in rice plant infested by white-backed planthopper, Sogatella furcifera. Appl. Entomol. Zool. 40: 91-97.KANT, M.R., AMENT, K., SABELIS, M.W., HARING, M.A., AND SCHUURINK, R.C. (2004). Differential timing of spidermite-induced direct and indirect defenses in tomato plants. Plant Physiol. 135: 483-495.KARBAN, R., ADAMCHAK, R., AND SCHNATHORST, W.C. (1987). Induced resistance and interspecific competition betweenspider mites and a vascular wilt fungus. Science 235: 678-680.KARBAN, R., AND AGRAWAL, A.A. (2002). Herbivore offense. Annu. Rev. Ecol. Syst. 33: 641-664.KESSLER, A., AND BALDWIN, I.T. (2002) Plant responses to insect herbivory: The emerging molecular analysis. Annu. Rev.Plant Biol. 53: 299-328.KESSLER, A., AND BALDWIN, I.T. (2004) Herbivore-induced plant vaccination. Part I. The orchestration of plant defensesin nature and their fitness consequences in the wild tobacco Nicotiana attenuata. Plant J. 38: 639-649.

146

G

H

J

K

KESSLER, A., HALITSCHKE, R., AND BALDWIN, I.T. (2004) Silencing the jasmonate cascade: Induced plant defensesand insect populations. Science 305: 665-668.KESSLER, A., HALITSCHKE, R., DIEZEL, C., AND BALDWIN, I.T. (2006). Priming of plant defense responses in natureby airborne signaling between Artemisia tridentata and Nicotiana attenuata. Oecologia 148: 280-292.KIM, M.S., KIM, Y.C., AND CHO, B.H. (2004). Gene expression analysis in cucumber leaves primed by root colonizationwith Pseudomonas chlororaphis O6 upon challenge-inoculation with Corynespora cassiicola. Plant Biol. 6: 105-108.KINDT, F., JOOSTEN, N.N., PETERS, D., AND TJALLINGII, W.F. (2003) Characterisation of the feeding behaviour ofWestern flower thrips in terms of electrical penetration graph (EPG) waveforms. J. Insect Physiol. 49: 183-191.KING, E.O., WARD, M.K., AND RANEY, D.E. (1954) Two simple media for the demonstration of phycocyanin and fluor-escin. J. Lab. Clin. Med. 44: 301-307.KLOEK, A.P., VERBSKY, M.L., SHARMA, S.B., SCHOELZ, J.E., VOGEL, J., KLESSIG, D.F., AND KUNKEL, B.N. (2001).Resistance to Pseudomonas syringae conferred by an Arabidopsis thaliana coronatine-insensitive (coi1) mutation occursthrough two distinct mechanisms. Plant J. 26: 509-522.KNOESTER, M., VAN LOON, L.C., VAN DEN HEUVEL, J., HENNIG, J., BOL, J.F., AND LINTHORST, H.J.M. (1998)Ethylene-insensitive tobacco lacks nonhost resistance against soil-borne fungi. Proc. Natl. Acad. Sci. U.S.A. 95: 1933-1937.KOCH, E., AND SLUSARENKO, A. (1990). Arabidopsis is susceptible to infection by a downy mildew fungus. Plant Cell2: 437-445.KOHLER, A., SCHWINDLING, S., AND CONRATH, U. (2002). Benzothiadiazole-induced priming for potentiated responsesto pathogen infection, wounding, and infiltration of water into leaves requires the NPR1/NIM1 gene in Arabidopsis.Plant Physiol. 128: 1046-1056.KONNO, K., HIRAYAMA, C., NAKAMURA, M., TATEISHI, K., TAMURA, Y., HATTORI, M., AND KOHNO, K. (2004).Papain protects papaya trees from herbivorous insects: role of cysteine proteases in latex. Plant J. 37: 370-378.KREPS, J.A., WU, Y.J., CHANG, H.S., ZHU, T., WANG, X., AND HARPER, J.F. (2002) Transcriptome changes for Arabidopsisin response to salt, osmotic, and cold stress. Plant Physiol. 130: 2129-2141.KRISHNA, P. (2003). Brassinosteroid-mediated stress responses. J. Plant Growth Regul. 22: 289-297.KUC, J. (1982). Induced immunity to plant disease. Bioscience 32: 854-860.KUNKEL, B.N. (1996) A useful weed put to work: genetic analysis of disease resistance in Arabidopsis thaliana. TrendsGenet. 12: 62-69.KUNKEL, B.N., BENT, A.F., DAHLBECK, D., INNES, R.W., AND STASKAWICZ, B.J. (1993) RPS2, an Arabidopsis diseaseresistance locus specifying recognition of Pseudomonas syringae strains expressing the avirulence gene AvrRpt2. PlantCell 5: 865-875.KUNKEL, B.N., AND BROOKS, D.M. (2002) Cross talk between signaling pathways in pathogen defense. Curr. Opin.Plant Biol. 5: 325-331.LAWTON, K.A., FRIEDRICH, L., HUNT, M., WEYMANN, K., DELANEY, T., KESSMANN, H., STAUB, T., AND RYALS,J. (1996). Benzothiadiazole induces disease resistance in Arabidopsis by activation of the systemic acquired resistancesignal transduction pathway. Plant J. 10: 71-82.LAWTON, K., WEYMANN, K., FRIEDRICH, L., VERNOOIJ, B., UKNES, S., AND RYALS, J. (1995). Systemic acquiredresistance in Arabidopsis requires salicylic acid but not ethylene. Mol. Plant-Microbe Interact. 8: 863-870.LEE, I., DATE, S.V., ADAI, A.T., AND MARCOTTE, E.M. (2004). A probabilistic functional network of yeast genes.Science 306: 1555-1558.LÉON-KLOOSTERZIEL, K.M., VERHAGEN, B.W.M., KEURENTJES, J.J.B., VAN PELT, J.A., REP, M., VAN LOON, L.C.,AND PIETERSE, C.M.J. (2005). Colonization of the Arabidopsis rhizosphere by fluorescent Pseudomonas spp. activatesa root-specific, ethylene-responsive PR-5 gene in the vascular bundle. Plant Mol. Biol. 57: 731-748.LI, C.Y., WILLIAMS, M.M., LOH, Y.T., LEE, G.I., AND HOWE, G.A. (2002) Resistance of cultivated tomato to cell content-feeding herbivores is regulated by the octadecanoid-signaling pathway. Plant Physiol. 130: 493-503.LI, J., BRADER, G., AND PALVA, E.T. (2004). The WRKY70 transcription factor: a node of convergence for jasmonate-mediated and salicylate-mediated signals in plant defense. Plant Cell 16: 319-331.LIU, Y., AHN, J.E., DATTA, S., SALZMAN, R.A., MOON, J., HUYGHUES-DESPOINTES, B., PITTENDRIGH, B., MURDOCK,L.L., KOIWA, H., AND ZHU-SALZMAN, K. (2005). Arabidopsis vegetative storage protein is an anti-insect acid phosphatase.Plant Physiol 139: 1545-1556.LORENZO, O., PIQUERAS, R., SÁNCHEZ-SERRANO, J.J., AND SOLANO, R. (2003). ETHYLENE RESPONSE FACTOR1integrates signals from ethylene and jasmonate pathways in plant defense. Plant Cell 15: 165-178.LORENZO, O., CHICO, J.M., SANCHEZ-SERRANO, J.J., AND SOLANO, R. (2004). JASMONATE-INSENSITIVE1 encodes a MYC transcription factor essential to discriminate between different jasmonate-regulated defense responses in Arabidopsis.Plant Cell 16: 1938-1950.LOU, Y.G., DU, M.H., TURLINGS, T.C.J., CHENG, J.A., AND SHAN, W.F. (2005). Exogenous application of jasmonic acidinduces volatile emissions in rice and enhances parasitism of Nilaparvata lugens eggs by the parasitoid Anagrus nilaparvatae.J. Chem. Ecol. 31: 1985-2002.LUND, S.T., STALL, R.E., AND KLEE, H.J. (1998) Ethylene regulates the susceptible response to pathogen infection intomato. Plant Cell 10: 371-382.MACKINNON, S.L., KEIFER, P., AND AYER, W.A. (1999). Components from the phytotoxic extract of Alternaria brassicicola,a black spot pathogen of canola. Phytochem. 51: 215-221.MALDONADO, A.M., DOERNER, P., DIXON, R.A., LAMB, C.J., AND CAMERON, R.K. (2002). A putative lipid transferprotein involved in systemic resistance signalling in Arabidopsis. Nature 419: 399-403.MALECK, K., LEVINE, A., EULGEM, T., MORGAN, A., SCHMID, J., LAWTON, K.A., DANGL, J.L., AND DIETRICH,

147

References

L

M

R.A. (2000). The transcriptome of Arabidopsis thaliana during systemic acquired resistance. Nature Genet. 26, 403-410.MARRS, K.A. (1996). The functions and regulation of glutathione S-transferases in plants. Annu. Rev. Plant Physiol. PlantMol. Biol. 47: 127-158.MAUCH-MANI, B., AND MÉTRAUX, J.-P. (1998). Salicylic acid and systemic acquired resistance to pathogen attack.Ann. Bot. 82: 535-540.MAUCH-MANI, B., AND MAUCH, F. (2005). The role of abscisic acid in plant-pathogen interactions. Curr. Opin. PlantBiol. 8: 409-414.MAYER, R.T., INBAR, M., MCKENZIE, C.L., SHATTERS, R., BOROWICZ, V., ALBRECHT, U., POWELL, C.A., ANDDOOSTDAR, H. (2002). Multitrophic interactions of the silverleaf whitefly, host plants, competing herbivores, and phyto-pathogens. Archiv. Insect Biochem. Physiol. 51: 151-169.MCCONN, J., CREELMAN, R.A., BELL, E., MULLET, J.E., AND BROWSE, J. (1997) Jasmonate is essential for insectdefense in Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 94: 5473-5477.MCINTYRE, J.L., DODDS, J.A., AND HARE, J.D. (1981). Effects of localized infections of Nicotiana tabacum by tobaccomosaic virus on systemic resistance against diverse pathogens and an insect. Phytopathol. 71: 297-301.MEUWLY, P., AND MÉTRAUX, J.-P. (1993) Ortho-anisic acid as internal standard for the simultaneous quantitation ofsalicylic acid and its putative biosynthetic precursors in cucumber leaves. Anal. Biochem. 214: 500-505.MEWIS, I., APPEL, H.M., HOM, A., RAINA, R., AND SCHULTZ, J.C. (2005). Major signaling pathways modulate Arabidopsisglucosinolate accumulation and response to both phloem-feeding and chewing insects. Plant Physiol. 138: 1149-1162.MEYEROWITZ, E.M., AND SOMMERVILLE, C.R. (1994). Arabidopsis. (New York, USA: Cold Spring Harbor Laboratory Press).MILES, P.W. (1999). Aphid saliva. Biol. Rev. 74: 41-85.MITCHELL-OLDS, T. (2001). Arabidopsis thaliana and its wild relatives: a model system for ecology and evolution.Trends Ecol. Evol. 16: 693-700.MOELLENBECK, D.J., QUISENBERRY, S.S., AND COLYER, P.D. (1992). Fusarium crown-rot development in alfalfa stressedby threecornered alfalfa hopper (Homoptera, Membracidae) feeding. J. Econ. Entomol. 85: 1442-1449.MORAN, P.J., CHENG, Y., CASSELL, J.L., AND THOMPSON, G.A. (2002) Gene expression profiling of Arabidopsis thalianain compatible plant-aphid interactions. Arch. Insect Biochem. Physiol. 51: 182-203.MORAN, P.J., AND THOMPSON, G.A. (2001) Molecular responses to aphid feeding in Arabidopsis in relation to plantdefense pathways. Plant Physiol. 125: 1074-1085.MUELLER, M.J., AND BRODSCHELM, W. (1994) Quantification of jasmonic acid by capillary gas chromatography-negativechemical-ionization mass spectrometry. Anal. Biochem. 218: 425-435.MUR, L.A.J., KENTON, P., ATZORN, R., MIERSCH, O., AND WASTERNACK, C. (2006). The outcomes of concentration-specific interactions between salicylate and jasmonate signaling include synergy, antagonism, and oxidative stress leadingto cell death. Plant Physiol. 140: 249-262.MUSSER, R.O., HUM-MUSSER, S.M., EICHENSEER, H., PEIFFER, M., ERVIN, G., MURPHY, J.B., AND FELTON, G.W.(2002). Herbivory: caterpillar saliva beats plant defences. Nature 416: 599-600.NAGANO, A.J., MATSUSHIMA, R., AND HARA-NISHIMURA, I. (2005). Activation of an ER-body-localized beta-glucosidasevia a cytosolic binding partner in damaged tissues of Arabidopsis thaliana. Plant Cell Physiol. 46: 1140-1148.NAWRATH, C., AND MÉTRAUX J.-P. (1999) Salicylic acid induction-deficient mutants of Arabidopsis express PR-2 andPR-5 and accumulate high levels of camalexin after pathogen inoculation. Plant Cell 11: 1393-1404.NEMHAUSER, J.L., HONG, F., AND CHORY, J. (2006). Different plant hormones regulate similar processes through largelynonoverlapping transcriptional responses. Cell 126: 467-475.NIKI, T., MITSUHARA, I., SEO, S., OHTSUBO, N., AND OHASHI, Y. (1998) Antagonistic effect of salicylic acid and jasmonicacid on the expression of pathogenesis-related (PR) protein genes in wounded mature tobacco leaves. Plant Cell Physiol.39: 500-507.NISHII, A., TAKEMURA, M., FUJITA, H., SHIKATA, M., YOKOTA, A., AND KOHCHI, T. (2000). Characterization of a novelgene encoding a putative single zinc-finger protein, ZIM, expressed during the reproductive phase in Arabidopsis thaliana.Biosci. Biotechnol. Biochem. 64: 1402-1409.NORMAN-SETTERBLAD, C., VIDAL, S., AND PALVA, T.E. (2000). Interacting signal pathways control defense geneexpression in Arabidopsis in response to cell wall-degrading enzymes from Erwinia carotovora. Mol. Plant-MicrobeInteract. 13: 430-438.O'DONNELL, P.J., CALVERT, C., ATZORN, R., WASTERNACK, C., LEYSER, H.M.O., AND BOWLES, D.J. (1996).Ethylene as a signal mediating the wound response of tomato plants. Science 274: 1914-1917.OH, I.S., PARK, A.R., BAE, M.S., KWON, S.J., KIM, Y.S., LEE, J.E., KANG, N.Y., LEE, S.M., CHEONG, H., AND PARK, O.K.(2005). Secretome analysis reveals an Arabidopsis lipase involved in defense against Alternaria brassicicola. Plant Cell17: 2832-2847.OPPENHEIMER, D.G., HERMAN, P., SIVAKUMARAN, S., ESCH, J., AND MARKS, M. (1991). A myb gene required for leaf trichome differentiation in Arabidopsis is expressed in stipules. Cell 67: 483-493.OROZCO-CARDENAS, M., MCGURL, B., AND RYAN, C.A. (1993). Expression of an antisense prosystemin gene in tomatoplants reduces resistance toward Manduca sexta larvae. Proc. Natl. Acad. Sci. U.S.A. 90: 8273-8276.OTANI, H., KOHNOBE, A., KODAMA, M., AND KOHMOTO, K. (1998). Production of a host-specific toxin by germinatingspores of Alternaria brassicicola. Physiol. Mol. Plant Pathol. 52: 285-295.PARÉ, P.W., AND TUMLINSON, J.H. (1999). Plant volatiles as a defense against insect herbivores. Plant Physiol. 121: 325-332.PARKER, B.L., SKINNER, M., AND LEWIS, T. (1995). Thrips biology and management. (New York: Plenum Press).PASCHOLD, A., HALITSCHKE, R., AND BALDWIN, I.T. (2006). Using 'mute' plants to translate volatile signals. Plant J.45: 275-291.

148

N

O

P

PEÑA-CORTÉS, H., ALBRECHT, T., PRAT, S., WEILER, E.W., AND WILLMITZER, L. (1993) Aspirin prevents wound-inducedgene expression in tomato leaves by blocking jasmonic acid biosynthesis. Planta 191: 123-128.PENNINCKX, I.A.M.A., THOMMA, B.P.H.J., BUCHALA, A., MÉTRAUX, J.-P., AND BROEKAERT, W.F. (1998).Concomitant activation of jasmonate and ethylene response pathways is required for induction of a plant defensin genein Arabidopsis. Plant Cell 10: 2103-2113.PIETERSE, C.M.J., TON, J., AND VAN LOON, L.C. (2001) Cross-talk between plant defence signaling pathways: Boostor burden? AgBiotechNet 3: ABN 068.PIETERSE, C.M.J., AND VAN LOON, L.C. (1999) Salicylic acid-independent plant defence pathways. Trends Plant Sci.4: 52-58.PIETERSE, C.M.J., AND VAN LOON, L.C. (2004) NPR1: the spider in the web of induced resistance signaling pathways.Curr. Opin. Plant Biol. 7: 456-464.PIETERSE, C.M.J., VAN PELT, J.A., TON, J., PARCHMANN, S., MUELLER, M.J., BUCHALA, A.J., MÉTRAUX, J.-P.,AND VAN LOON, L.C. (2000). Rhizobacteria-mediated induced systemic resistance (ISR) in Arabidopsis requires sensitivityto jasmonate and ethylene but is not accompanied by an increase in their production. Physiol. Mol. Plant Pathol. 57:123-134.PIETERSE, C.M.J., VAN WEES, S.C.M., HOFFLAND, E., VAN PELT, J.A., AND VAN LOON, L.C. (1996). Systemic resistancein Arabidopsis induced by biocontrol bacteria is independent of salicylic acid accumulation and pathogenesis-related geneexpression. Plant Cell 8: 1225-1237.PIETERSE, C.M.J., VAN WEES, S.C.M., TON, J., VAN PELT, J.A., AND VAN LOON, L.C. (2002). Signalling in rhizobacteria-induced systemic resistance in Arabidopsis thaliana. Plant Biol. 4: 535-544.PIETERSE, C.M.J., VAN WEES, S.C.M., VAN PELT, J.A., KNOESTER, M., LAAN, R., GERRITS, H., WEISBEEK, P.J.,AND VAN LOON L.C. (1998) A novel signaling pathway controlling induced systemic resistance in Arabidopsis. Plant Cell10: 1571-1580.POZO, M.J., VAN DER ENT, S., VAN LOON, L.C., AND PIETERSE, C.M.J. (in preparation). Biocontrol bacteria induceplant systemic resistance through MYC2-regulated priming of JA-dependent defenses.POZO, M.J., VAN LOON, L.C., AND PIETERSE, C.M.J. (2005) Jasmonates - Signals in plant-microbe interactions. J. Plant Growth Regul. 23: 211-222.PRESTON, C.A., LEWANDOWSKI, C., ENYEDI, A.J., AND BALDWIN, I.T. (1999). Tobacco mosaic virus inoculationinhibits wound-induced jasmonic acid-mediated responses within but not between plants. Planta 209: 87-95.PRUITT, R.E., AND MEYEROWITZ, E.M. (1986) Characterization of the genome of Arabidopsis thaliana. J. Mol. Biol.187: 169-183.RAVEN, J.A. (1983). Phytophages of xylem and phloem. Adv. Ecol. Res. 13: 135-234.REDMAN, J.C., HAAS, B.J., TANIMOTO, G., AND TOWN, C.D. (2004) Development and evaluation of an Arabidopsiswhole genome Affymetrix probe array. Plant J. 38: 545-561.REYMOND, P., BODENHAUSEN, N., VAN POECKE, R.M.P., KRISHNAMURTHY, V., DICKE, M., AND FARMER E.E.(2004) A conserved transcript pattern in response to a specialist and a generalist herbivore. Plant Cell 16: 3132-3147.REYMOND, P., AND FARMER, E.E. (1998) Jasmonate and salicylate as global signals for defense gene expression. Curr.Opin. Plant Biol. 1: 404-411.REYMOND, P., WEBER, H., DAMOND, M., AND FARMER, E.E. (2000) Differential gene expression in response tomechanical wounding and insect feeding in Arabidopsis. Plant Cell 12: 707-719.RHEE, S.Y., BEAVIS, W., BERARDINI, T.Z., CHEN, G.H., DIXON, D., DOYLE, A., GARCIA-HERNANDEZ, M., HUALA,E., LANDER, G., MONTOYA, M., MILLER, N., MUELLER, L.A., MUNDODI, S., REISER, L., TACKLIND, J., WEEMS,D.C., WU, Y.H., XU, I., YOO, D., YOON, J., AND ZHANG, P.F. (2003) The Arabidopsis Information Resource (TAIR): a model organism database providing a centralized, curated gateway to Arabidopsis biology, research materials andcommunity. Nucleic Acids Res. 31: 224-228.ROJO, E., SOLANO, R., AND SANCHEZ-SERRANO, J.J. (2003) Interactions between signaling compounds involved in plant defense. J. Plant Growth Regul. 22: 82-98.ROSS, A.F. (1961). Systemic acquired resistance induced by localized virus infections in plants. Virology 14: 340-358.ROSTAS, M., TON, J., MAUCH-MANI, B., AND TURLINGS, T.C. (2006). Fungal infection reduces herbivore-inducedplant volatiles of maize, but does not affect naive parasitoids. J. Chem. Ecol. 32: 1897-1909.RUSHTON, P.J., AND SOMSSICH, I.E. (1998). Transcriptional control of plant genes responsive to pathogens. Curr.Opin. Plant Biol. 1: 311-315.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: 1808-1819.RYAN, C.A. (1992). The search for the proteinase inhibitor-inducing factor, Piif. Plant Mol. Biol. 19: 123-133.RYU, C.M., MURPHY, J.F., MYSORE, K.S., AND KLOEPPER, J.W. (2004). Plant growth-promoting rhizobacteria systemicallyprotect Arabidopsis thaliana against Cucumber mosaic virus by a salicylic acid and NPR1-independent and jasmonicacid-dependent signaling pathway. Plant J. 39: 381-392.SAMBROOK, J., FRITSCH, E.F., AND MANIATIS, T. (1989) Molecular Cloning: A Laboratory Manual, 2nd edition. ColdSpring Harbor Laboratory Press, Cold Spring Harbor, NY, U.S.A.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. Proc. Natl. Acad. Sci.U.S.A. 97: 11655-11660.SCHIPPERS, B., BAKKER, A.W., AND BAKKER, P.A.H.M. (1987). Interactions of deleterious and beneficial rhizospheremicoorganisms and the effect of cropping practices. Annu. Rev. Phytopathol. 115: 339-358.

149

References

R

S

SCHOONHOVEN, L.M., JERMY, T., AND VAN LOON, J.J.A. (1998) Insect-Plant Biology. From Physiology to Evolution.Chapman and Hall, London.SEGURA, A., MORENO, M., AND GARCIAOLMEDO, F. (1993). Purification and antipathogenic activity of lipid transferproteins (Ltps) from the leaves of Arabidopsis and spinach. FEBS Lett. 332: 243-246.SHAH, J., TSUI, F., AND KLESSIG, D.F. (1997). Characterization of a salicylic acid-insensitive mutant (sai1) of Arabidopsisthaliana, identified in a selective screen utilizing the SA-inducible expression of the tms2 gene. Mol. Plant-Microbe Interact.10: 69-78.SIMON, M., AND HILKER, M. (2003). Herbivores and pathogens on willow: do they affect each other? Agr. Forest Entomol.5: 275-284.SMALLE, J., AND VIERSTRA, R.D. (2004). The ubiquitin 26S proteasome proteolytic pathway. Annu. Rev. of Plant Biol.55: 555-590.SMITS, P.H., VAN DE VRIE, M., AND VLAK, J.M. (1986). Oviposition of beet armyworm (Lepidoptera: Noctuidae) ongreenhouse crops. Environm. Entomol. 15: 1189-1191.SPOEL, S.H., KOORNNEEF, A., CLAESSENS, S.M.C., KORZELIUS, J.P., VAN PELT, J.A., MUELLER, M.J., BUCHALA,A.J., MÉTRAUX, J.-P., BROWN, R., KAZAN, K., VAN LOON, L.C., DONG, X.N., AND PIETERSE, C.M.J. (2003) NPR1modulates cross-talk between salicylate-and jasmonate-dependent defense pathways through a novel function in thecytosol. Plant Cell 15: 760-770.STASWICK, P.E., YUEN, G.Y., AND LEHMAN, C.C. (1998) Jasmonate signaling mutants of Arabidopsis are susceptibleto the soil fungus Pythium irregulare. Plant J. 15: 747-754.STICHER, L., MAUCH-MANI, B., AND MÉTRAUX, J.-P. (1997). Systemic acquired resistance. Annu. Rev. Phytopathol.35: 235-270.STINTZI, A., WEBER, H., REYMOND, P., BROWSE, J., AND FARMER, E.E. (2001). Plant defense in the absence of jasmonic acid: The role of cyclopentenones. Proc. Natl. Acad. Sci. U.S.A. 98: 12837-12842.STOTZ, H.U., KOCH, T., BIEDERMANN, A., WENIGER, K., BOLAND, W., AND MITCHELL-OLDS, T. (2002) Evidencefor regulation of resistance in Arabidopsis to Egyptian cotton worm by salicylic and jasmonic acid signaling pathways.Planta 214: 648-652.STOTZ, H.U., PITTENDRIGH, B.R., KROYMANN, J., WENIGER, K., FRITSCHE, J., BAUKE, A., AND MITCHELL-OLDS, T.(2000). Induced plant defense responses against chewing insects. Ethylene signaling reduces resistance of Arabidopsisagainst Egyptian cotton worm but not diamondback moth. Plant Physiol. 124: 1007-1017.STOUT, M.J., FIDANTSEF, A.L., DUFFEY, S.S., AND BOSTOCK, R.M. (1999) Signal interactions in pathogen and insectattack: Systemic plant-mediated interactions between pathogens and herbivores of the tomato, Lycopersicon esculentum.Physiol. Mol. Plant Pathol. 54: 115-130.STOUT, M.J., THALER, J.S., AND THOMMA, B.P.H.J. (2006). Plant-mediated interactions between pathogenic micro-organisms and herbivorous arthropods. Annu. Rev. Entomol. 51: 663-689.TAKABAYASHI, J., AND DICKE, M. (1996). Plant-carnivore mutualism through herbivore-induced carnivore attractants.Trends Plant Sci. 1: 109-113.TAO, Y., XIE, Z.Y., CHEN, W.Q., GLAZEBROOK, J., CHANG, H.-S., HAN, B., ZHU, T., ZOU, G.Z., AND KATAGIRI, F.(2003) Quantitative nature of Arabidopsis responses during compatible and incompatible interactions with the bacterialpathogen Pseudomonas syringae. Plant Cell 15: 317-330.TELANG, A., SANDSTROM, J., DYRESON, E., AND MORAN, N.A. (1999). Feeding damage by Diuraphis noxia resultsin a nutritionally enhanced phloem diet. Entomol. Exp. Appl. 91: 403-412.THALER, J.S., AND BOSTOCK, R.M. (2004). Interactions between abscisic-acid-mediated responses and plant resistanceto pathogens and insects. Ecology 85: 48-58.THALER, J.S., FARAG, M.A., PARÉ, P.W., AND DICKE, M. (2002a) Jasmonate-deficient plants have reduced direct andindirect defences against herbivores. Ecol. Lett. 5: 764-774.THALER, J.S., FIDANTSEF, A.L., AND BOSTOCK, R.M. (2002b) Antagonism between jasmonate- and salicylate-mediatedinduced plant resistance: Effects of concentration and timing of elicitation on defense-related proteins, herbivore, andpathogen performance in tomato. J. Chem. Ecol. 28: 1131-1159.THALER, J.S., FIDANTSEF, A.L., DUFFEY, S.S., AND BOSTOCK, R.M. (1999) Trade-offs in plant defense against pathogensand herbivores: A field demonstration of chemical elicitors of induced resistance. J. Chem. Ecol. 25: 1597-1609.THALER, J.S., OWEN, B., AND HIGGINS, V.J. (2004). The role of the jasmonate response in plant susceptibility to diversepathogens with a range of lifestyles. Plant Physiol. 135: 530-538.THIMM, O., BLASING, O., GIBON, Y., NAGEL, A., MEYER, S., KRUGER, P., SELBIG, J., MULLER, L.A., RHEE, S.Y.,AND STITT, M. (2004). MAPMAN: a user-driven tool to display genomics data sets onto diagrams of metabolic pathwaysand other biological processes. Plant J. 37: 914-939.THOMMA, B.P.H.J., EGGERMONT, K., PENNINCKX, I.A.M.A., MAUCH-MANI, B., VOGELSANG, R., CAMMUE, B.P.A.,AND BROEKAERT, W.F. (1998) Separate jasmonate-dependent and salicylate-dependent defense-response pathways in Arabidopsis are essential for resistance to distinct microbial pathogens. Proc. Natl. Acad. Sci. U.S.A. 95: 15107-15111.THOMMA, B.P.H.J., NELISSEN, I., EGGERMONT, K., AND BROEKAERT, W.F. (1999) Deficiency in phytoalexin productioncauses enhanced susceptibility of Arabidopsis thaliana to the fungus Alternaria brassicicola. Plant J. 19: 163-171.THOMMA, B.P.H.J., PENNINCKX, I.A.M.A., CAMMUE, B.P.A., AND BROEKAERT, W.F. (2001) The complexity of diseasesignaling in Arabidopsis. Curr. Opin. Immunol. 13: 63-68.THOMPSON, G.A., AND GOGGIN, F.L. (2006). Transcriptomics and functional genomics of plant defence induction byphloem-feeding insects. J. Exp. Bot. 57: 755-766.THULKE, O., AND CONRATH, U. (1998). Salicylic acid has a dual role in the activation of defence-related genes in parsley.

150

T

Plant J. 14: 35-42.TJALLINGII, W.F. AND HOGEN ESCH, T. (1993) Fine structure of aphid stylet routes in plant tissues in correlation withEPG-signals. Physiol. Entomol. 18: 317-328.TJAMOS, S.E., FLEMETAKIS, E., PAPLOMATAS, E.J., AND KATINAKIS, P. (2005). Induction of resistance to Verticilliumdahliae in Arabidopsis thaliana by the biocontrol agent K-165 and pathogenesis-related proteins gene expression. Mol.Plant-Microbe Interact. 18: 555-561.TON, J., D'ALESSANDRO, M., JOURDIE, V., JAKAB, G., KARLEN, D., HELD, M., MAUCH-MANI, B., AND TURLINGS,T.C.J. (2007). Priming by air-borne signals boosts direct and indirect resistance in maize. Plant J. 49: 16-26.TON, J., JAKAB, G., TOQUIN, V., FLORS, V., IAVICOLI, A., MAEDER, M.N., MÉTRAUX, J.-P., AND MAUCH-MANI, B.(2005). Dissecting the ß-aminobutyric acid-induced priming phenomenon in Arabidopsis. Plant Cell 17: 987-999.TON, J., VAN PELT, J.A., VAN LOON, L.C., AND PIETERSE, C.M.J. (2002) Differential effectiveness of salicylate-dependentand jasmonate/ethylene-dependent induced resistance in Arabidopsis. Mol. Plant-Microbe Interact. 15: 27-34.TURLINGS, T.C.J., LOUGHRIN, J.H., MCCALL, P.J., ROSE, U.S.R., LEWIS, W.J., AND TUMLINSON, J.H. (1995). Howcaterpillar-damaged plants protect themselves by attracting parasitic wasps. Proc. Natl. Acad. Sci. U.S.A. 92: 4169-4174.USADEL, B., NAGEL, A., THIMM, O., REDESTIG, H., BLAESING, O.E., PALACIOS-ROJAS, N., SELBIG, J., HANNEMANN,J., PIQUES, M.C., STEINHAUSER, D., SCHEIBLE, W.R., GIBON, Y., MORCUENDE, R., WEICHT, D., MEYER, S., ANDSTITT, M. (2005). Extension of the visualization tool MapMan to allow statistical analysis of arrays, display of corespondinggenes, and comparison with known responses. Plant Physiol. 138: 1195-1204.VAN LOON, J.J.A., DE BOER, J.G., AND DICKE, M. (2000) Parasitoid-plant mutualism: Parasitoid attack of herbivoreincreases plant reproduction. Entomol. Exp. Appl. 97: 219-227.VAN LOON, L.C. (2000). Systemic induced resistance. In Mechanisms of Resistance to Plant Diseases, A.J. Slusarenko,R.S.S. Fraser, and L.C. Van Loon, eds (Dordrecht, The Netherlands: Kluwer Academic Publishers), pp. 521-574.VAN LOON, L.C., AND BAKKER, P.A.H.M. (2006). Root-associated bacteria inducing systemic resistance. In Plant-associatedbacteria., S.S. Gnanamanickam, ed (Dordrecht: Springer), pp. 269-316.VAN LOON, L.C., BAKKER, P.A.H.M., AND PIETERSE, C.M.J. (1998). Systemic resistance induced by rhizosphere bacteria.Annu. Rev. Phytopathol. 36: 453-483.VAN LOON, L.C., REP, M., AND PIETERSE, C.M.J. (2006). Significance of inducible defense-related proteins in infectedplants. Annu. Rev. Phytopathol. 44: 135-162.VAN LOON, L.C., GERAATS, B.P., AND LINTHORST, H.J. (2006). Ethylene as a modulator of disease resistance in plants.Trends Plant Sci. 11: 184-191.VAN LOON, L.C., AND VAN STRIEN, E.A. (1999). The families of pathogenesis-related proteins, their activities, andcomparative analysis of PR-1 type proteins. Physiol. Mol. Plant Pathol. 55: 85-97.VAN POECKE, R.M.P. (2002). Indirect defence of Arabidopsis against herbivorous insects. Combining parasitoid behaviourand chemical analyses with a molecular genetic approach. (Ph. D. Thesis, Wageningen University, The Netherlands), pp. 180.VAN POECKE, R.M.P., AND DICKE, M. (2002) Induced parasitoid attraction by Arabidopsis thaliana: Involvement of theoctadecanoid and salicylic acid pathway. J. Exp. Bot. 53: 1793-1799.VAN POECKE, R.M.P., AND DICKE, M. (2003). Signal transduction downstream of salicylic and jasmonic acid in herbivory-induced parasitoid attraction by Arabidopsis is independent of JAR1 and NPR1. Plant Cell Environm. 26: 1541-1548.VAN POECKE, R.M.P., AND DICKE, M. (2004) Indirect defence of plants against herbivores: Using Arabidopsis thalianaas a model plant. Plant Biol. 6: 387-401.VAN POECKE, R.M.P., POSTHUMUS, M.A., AND DICKE, M. (2001) Herbivore-induced volatile production by Arabidopsisthaliana leads to attraction of the parasitoid Cotesia rubecula: Chemical, behavioral, and gene-expression analysis. J. Chem.Ecol. 27: 1911-1928.VAN POECKE, R.M.P., ROOSJEN, M., PUMARINO, L., AND DICKE, M. (2003) Attraction or the specialist parasitoidCotesia rubecula to Arabidopsis thaliana infested by host or non-host herbivore species. Entomol. Exp. Appl. 107: 229-236.VAN WEES, S.C.M., CHANG, H.-S., ZHU, T., AND GLAZEBROOK, J. (2003) Characterization of the early response ofArabidopsis to Alternaria brassicicola infection using expression profiling. Plant Physiol. 132: 606-617.VAN WEES, S.C.M., DE SWART, E.A.M., VAN PELT, J.A., VAN LOON, L.C., AND PIETERSE, C.M.J. (2000).Enhancement of induced disease resistance by simultaneous activation of salicylate- and jasmonate-dependent defensepathways in Arabidopsis thaliana. Proc. Natl. Acad. Sci. U.S.A. 97: 8711-8716.VAN WEES, S.C.M., LUIJENDIJK, M., SMOORENBURG, I., VAN LOON, L.C., AND PIETERSE, C.M.J. (1999).Rhizobacteria-mediated induced systemic resistance (ISR) in Arabidopsis is not associated with a direct effect on expressionof known defense-related genes but stimulates the expression of the jasmonate-inducible gene Atvsp upon challenge.Plant Mol. Biol. 41: 537-549.VERHAGEN, B.W.M., GLAZEBROOK, J., ZHU, T., CHANG, H.-S., VAN LOON, L.C., AND PIETERSE, C.M.J. (2004)The transcriptome of rhizobacteria-induced systemic resistance in Arabidopsis. Mol. Plant-Microbe Interact. 17: 895-908.VERHAGEN, B.W.M., VAN LOON, L.C., AND PIETERSE, C.M.J. (2006). Induced disease resistance signaling in plants.In Floriculture, Ornamental and Plant Biotechnology, Vol III, J.A. Teixeira da Silva, ed (London: Global Science Books),pp. 334-343.VIJAYAN, P., SHOCKEY, J., LEVESQUE, C.A., COOK, R.J., AND BROWSE, J. (1998). A role for jasmonate in pathogendefense of Arabidopsis. Proc. Natl. Acad. Sci. U.S.A. 95: 7209-7214.VOELCKEL, C., AND BALDWIN, I.T. (2004). Generalist and specialist lepidopteran larvae elicit different transcriptionalresponses in Nicotiana attenuata, which correlate with larval FAC profiles. Ecol. Lett. 7: 770-775.WAGNER, U., EDWARDS, R., DIXON, D.P., AND MAUCH, F. (2002). Probing the diversity of the Arabidopsis glutathioneS-transferase gene family. Plant Mol. Biol. 49: 515-532.

151

References

U

V

W

WALLING, L.L. (2000). The myriad plant responses to herbivores. J. Plant Growth Regul. 19: 195-216.WANG, Y.Q., OHARA, Y., NAKAYASHIKI, H., TOSA, Y., AND MAYAMA, S. (2005). Microarray analysis of the geneexpression profile induced by the endophytic plant growth-promoting rhizobacteria, Pseudomonas fluorescens FPT9601-T5in Arabidopsis. Mol. Plant-Microbe Interact. 18: 385-396.WHALEN, M.C., INNES, R.W., BENT, A.F., AND STASKAWICZ, B.J. (1991) Identification of Pseudomonas syringaepathogens of Arabidopsis and a bacterial locus determining avirulence on both Arabidopsis and soybean. Plant Cell 3: 49-59.WILDERMUTH, M.C., DEWDNEY, J., WU, G., AND AUSUBEL, F.M. (2001) Isochorismate synthase is required to synthesizesalicylic acid for plant defence. Nature 414: 562-565.WINZ, R.A., AND BALDWIN, I.T. (2001). Molecular interactions between the specialist herbivore Manduca sexta (Lepidoptera,Sphingidae) and its natural host Nicotiana attenuata. IV. Insect-induced ethylene reduces jasmonate-induced nicotineaccumulation by regulating putrescine N-methyltransferase transcripts. Plant Physiol. 125: 2189-2202.WITTSTOCK, U., AGERBIRK, N., STAUBER, E.J., OLSEN, C.E., HIPPLER, M., MITCHELL-OLDS, T., GERSHENZON, J., AND VOGEL, H. (2004). Successful herbivore attack due to metabolic diversion of a plant chemical defense. Proc. Natl.Acad. Sci. U.S.A. 101: 4859-4864.XU, X., CHEN, C., FAN, B., AND CHEN, Z. (2006). Physical and functional interactions between pathogen-inducedArabidopsis WRKY18, WRKY40, and WRKY60 transcription factors. Plant Cell 18: 1310-1326.YAMADA, K., MATSUSHIMA, R., NISHIMURA, M., AND HARA-NISHIMURA, I. (2001). A slow maturation of a cysteineprotease with a granulin domain in the vacuoles of senescing Arabidopsis leaves. Plant Physiol. 127: 1626-1634.YANO, S., AND OHSAKI, N. (1993). The phenology and intrinsic quality of wild crucifers that determine the communitystructure of their herbivorous insects. Res. Pop. Ecol. 35: 151-170.YUDIN, L.S., CHO, J.J., AND MITCHELL, W.L. (1986) Host range of western flower thrips, Frankliniella occidentalis(Thysantoptera: Thripidae) with special reference to Leucaena glausa. Environ. Entomol. 15: 1292-1295.ZEHNDER, G.W., MURPHY, J.F., SIKORA, E.J., AND KLOEPPER, J.W. (2001). Application of rhizobacteria for inducedresistance. Eur. J. Plant Pathol. 107: 39-50.ZHOU, N., TOOTLE, T.L., TSUI, F., KLESSIG, D.F., AND GLAZEBROOK, J. (1998). PAD4 functions upstream from salicylicacid to control defense responses in Arabidopsis. Plant Cell 10: 1021-1030.ZHU, T., BUDWORTH, P., HAN, B., BROWN, D., CHANG, H.-S., ZOU, G.Z., AND WANG, X. (2001) Toward elucidatingthe global gene expression patterns of developing Arabidopsis: Parallel analysis of 8300 genes by a high-densityoligonucleotide probe array. Plant Physiol. Biochem. 39: 221-242.ZHU-SALZMAN, K., SALZMAN, R.A., AHN, J.E., AND KOIWA, H. (2004). Transcriptional regulation of sorghum defensedeterminants against a phloem-feeding aphid. Plant Physiol. 134: 420-431.ZIMMERLI, L., JAKAB, G., MÉTRAUX, J.-P., AND MAUCH-MANI, B. (2000). Potentiation of pathogen-specific defensemechanisms in Arabidopsis by ß-aminobutyric acid. Proc. Natl. Acad. Sci. U.S.A. 97: 12920-12925.

ZOU, J.J., RODRIGUEZ-ZAS, S., ALDEA, M., LI, M., ZHU, J., GONZALEZ, D.O., VODKIN, L.O., DELUCIA, E., AND

CLOUGH, S.J. (2005). Expression profiling soybean response to Pseudomonas syringae reveals new defense-related

genes and rapid HR-specific downregulation of photosynthesis. Mol. Plant-Microbe Interact. 18: 1161-1174.

152

XY

Z

Nederlandse samenvattingPlanten kunnen worden aangevallen door zeer diverse ziekteverwekkers en insecten. Om zichte verdedigen, hebben zij uiteenlopende strategieën ontwikkeld. Resistentie tegen pathogenenen insecten kan worden bewerkstelligd door verdedingsmechanismen die altijd aanwezig zijn(constitutief), of door mechanismen die pas actief worden na een aanval (geïnduceerd).Tijdens mijn promotie-onderzoek heb ik mij gericht op geïnduceerde resistentie in de model-plant zandraket (Arabidopsis thaliana), na aanval door zowel pathogenen als insecten. Deaanpak van dit project was interdisciplinair: het onderzoek is uitgevoerd in nauwe samenwerkingtussen de leerstoelgroepen Entomologie (Wageningen Universiteit) en Fytopathologie (UniversiteitUtrecht). Bovendien is mijn promotieonderzoek sterk verwant aan dat van Martin de Vos, diebinnen hetzelfde samenwerkingsverband werkte en in 2006 promoveerde aan de UniversiteitUtrecht. Remco van Poecke legde tijdens zijn promotieonderzoek, afgerond in 2002, de funda-menten voor het werk aan indirecte verdediging tegen insecten in Arabidopsis. Hij werktedaarna als Post-doc mee aan dit samenwerkingsproject.

“Cross-talk” tussen signalen bij de verdediging van plantentegen pathogenen en insecten

Een belangrijke vraag in het onderzoek over signaaltransductie tijdens de verdediging vanplanten is: hoe zijn planten in staat om de signalen die geïnduceerd worden na aanval doorpathogene micro-organismen en herbivore insecten dusdanig te integreren dat dit leidt tot eenverdedigingsrespons die specifiek werkzaam is tegen de betrokken aanvaller? Het is bekenddat de plantenhormonen salicylzuur (SA), jasmonzuur (JA), en ethyleen (ET) een belangrijkerol spelen bij de regulatie van de signaalnetwerken die ten grondslag liggen aan geïnduceerdeafweer. Zodra de plant een belager herkent, verhoogt zij de productie van een of meer vandeze signaalmoleculen.Vervolgens wordt een specifieke afweerreactie actief in het aangetasteweefsel.

Een verhoogde productie van SA wordt vaak in verband gebracht met geïnduceerdeafweer tegen biotrofe pathogenen, die alleen voedsel kunnen onttrekken aan levende cellen.Verhoogde biosynthese van JA bevordert afweer tegen insecten en tegen necrotrofe pathogenen,die het weefsel eerst doden alvorens zij daarin kunnen groeien. Bij beide verdedigingsstrate-gieën speelt het plantenhormoon ET een modulerende rol. Onderzoek naar de rol van SA, JAen ET heeft aangetoond dat deze drie signaalmoleculen interacteren in een complex netwerkvan signaal-transductieroutes, die gezamenlijk bepalend zijn voor de inductie van de afweer-reactie tegen het pathogeen of insect dat de plant belaagt. De interactie tussen de verschillendesignaal-transductieroutes heet “cross-talk”.Wetenschappers denken dat “cross-talk” een manieris waarop de plant een optimale afweer kan activeren als zij wordt aangevallen door een ofmeer belagers. Sommige gespecialiseerde aanvallers maken juist gebruik van zulke “cross-talk”om de afweerreactie van de plant zodanig te manipuleren dat zij daarvan geen last ondervinden.Zo voorkomen zij een afweerreactie die wel effectief geweest zou zijn tegen een niet-gespecia-liseerde aanvaller.

153

Nederlandse samenvatting

“Hormoon handtekening”Om beter te begrijpen hoe “cross-talk” tussen de signaalmoleculen SA, JA en ET werkt tijdensgeïnduceerde resistentie, is de reactie bestudeerd van de modelplant Arabidopsis thaliana op deaanval door elk van een vijftal pathogenen en insecten (hoofdstuk 2 en 3). Arabidopsis werdgeïnfecteerd met de bacterie Pseudomonas syringae pv. tomato, of de schimmel Alternariabrassicicola, of de volgende insecten werden aangebracht: rupsen van het kleine koolwitje(Pieris rapae), larven van de Californische trips (Frankliniella occidentalis) of groene perzik-luizen (Myzus persicae). Vervolgens werden de productieniveaus bekeken van de planten-hormonen SA, JA en ET (“signal signature” of “hormoon handtekening”) . Elke plant-belagerinteractie bleek een specifieke “hormoon handtekening” te induceren; de productie van SA,JA en ET in het aangetaste weefsel verschilde in samenstelling, relatieve hoeveelheden en inde snelheid waarmee de toenames optreden. Dit geeft aan dat de “hormoon handtekening”een regulerend potentieel bezit voor de afstemming van de afweerreactie.

GenexpressieVervolgens werd onderzocht hoe de “hormoon handtekening” wordt vertaald in het activerenvan afweerreacties. Hiertoe werd de expressie bestudeerd van alle ~23.000 genen vanArabidopsis met behulp van zogenaamde genenchips (Arabidopsis whole-genome AffymetrixATH1 GeneChips).Analyse van de geactiveerde en gerepresseerde genen maakte duidelijk datiedere “hormoon handtekening” in de plant wordt vertaald in een specifiek en uiterst complexgenexpressie-profiel. Aanval door de bacterie (P. syringae), de schimmel (A. brassicicola), derupsen (P. rapae) en de tripsen (F. occidentalis) leidde vooral tot verhoogde productie vanJA in de plant, en ook tot veranderde expressie van genen die door JA worden gereguleerd.Echter, de overlap tussen deze genen was gering. Dit duidt erop dat behalve de primaire sig-naalmoleculen SA, JA en ET andere regulatiemechanismen actief moeten zijn die de afstemmingvan de afweerreactie van de plant op een bepaalde belager beïnvloeden.Wel bleek het zo tezijn dat veel geïnduceerde genen, hoewel niet identiek, een rol speelden in vergelijkbare stof-wisselingsprocessen. Dit betekent dat de plant in staat is dezelfde processen via verschillendetranscriptionele veranderingen te reguleren.

In tegenstelling tot de reactie op de andere belagers, leidde beschadiging door bladluizenniet tot meetbaar verhoogde productie van SA, JA of ET.Wel veranderde de transcriptioneleactiviteit van een groot aantal genen (ruim 2.000). Het merendeel van deze genen kwam mindersterk tot expressie dan in onbeschadigde planten, terwijl de andere Arabidopsis-belager inter-acties, net als het toedienen van JA, voornamelijk resulteerden in verhoogde genexpressie.Dit geeft aan dat de interactie tussen bladluizen en Arabidopsis atypisch is vergeleken met de andere belagers.

154

“Cross-effectiveness” tussen door pathogenen en insectengeïnduceerde resistentie

De kennis over de geïnduceerde afweer van Arabidopsis tegen deze vijf belagers leidde totde vraag wat het spectrum van effectiviteit hiervan zou zijn. Is de afweer tegen pathogenenalleen effectief tegen de aanvaller zelf, of ook tegen insecten? Leidt de reactie op beschadigingdoor insecten slechts tot afweer tegen de specifieke belager, of is de plant nu ook beter beschermdtegen pathogenen? Er is al veel bekend over de rol van SA, JA en ET in de afweer van planten.Het toedienen van JA aan een plant kan bijvoorbeeld leiden tot verhoogde afweer tegen som-mige insecten of necrotrofe pathogenen vergeleken met onbehandelde planten. Omgekeerdzijn planten die geen JA kunnen maken minder goed beschermd tegen deze belagers. Het toe-dienen van SA kan juist verhoogde resistentie tegen biotrofe pathogenen tot gevolg hebben,terwijl planten die dit signaalmolecuul niet kunnen maken verminderd resistent zijn ten opzichtevan onbehandelde of wildtype planten.Verder is aangetoond dat SA de door JA geïnduceerdeafweerrespons kan onderdrukken, een bekend voorbeeld van “cross-talk”.Als planten met SAbehandeld zijn, kunnen zij zich minder goed verweren tegen insecten die gevoelig zijn vooreen JA-afhankelijk afweermechanisme. De verwachting was dan ook dat resistentie, geïnduceerddoor belagers die hoofdzakelijk JA-afhankelijke afweermechanismen stimuleren, tevens effectiefzou zijn tegen andere belagers die gevoelig zijn voor JA-afhankelijke afweer. Omgekeerd wasde verwachting dat deze vorm van resistentie juist niet effectief zou zijn tegen belagers diegevoelig zijn voor SA-afhankelijke afweermechanismen.

Effectiviteit van door insecten geïnduceerde resistentie tegen pathogenenMartin de Vos wijdde zijn onderzoek aan de vraag in hoeverre door insecten geïnduceerderesistentie effectief is tegen pathogenen, terwijl ik me verdiepte in de effectiviteit van micro-biëel geïnduceerde resistentie tegen insecten. Martin verwachtte dat door P. rapae geïnduceerde,JA-afhankelijke resistentie effectief zou zijn tegen de schimmel A. brassicicola, waarvan bekendis dat deze gevoelig is voor JA-afhankelijke afweer. Echter, hij nam geen verhoogde beschermingtegen deze schimmel waar op planten die eerst waren aangevreten door rupsen van P. rapae.Wel vond hij dat de aangevreten planten beter beschermd waren tegen het Turnip Crinkle Virus(TCV). Dit was onverwacht, omdat bekend is dat dit virus alleen door SA-, en niet door JA-afhankelijke afweermechanismen wordt geremd. Martin analyseerde vervolgens de expressievan het door SA geïnduceerde PR1 gen, en ontdekte dat de aangevreten planten dit gen ver-sneld activeerden na infectie met TCV. Dit betekende dat de planten door de rupsenvraat ver-hoogd gevoelig werden voor SA-afhankelijke afweer. Dit fenomeen wordt “priming” genoemd,en wordt in verband gebracht met verschillende typen geïnduceerde resistentie.Ge”prime”de planten kunnen geïnduceerde resistentie versneld en sterker activeren na aan-tasting door pathogenen, insecten of andere stress-factoren. Het feit dat vraat door P. rapaeleidt tot versterkte SA-afhankelijke afweerreacties laat zien dat door insecten geïnduceerdeafweer verrassend complex is.

155


Effectiviteit van microbiëel geïnduceerde resistentie tegen insectenIk richtte me op microbiëel geïnduceerde resistentie en bestudeerde de effectiviteit tegeninsecten (hoofdstuk 4 en 5). In het bijzonder keek ik naar twee goed gekarakteriseerde vormenvan microbiëel geïnduceerde resistentie: “systemic aquired resistance” (SAR), geïnduceerddoor infectie met een necrotiserend pathogeen, en “induced systemic resistance” (ISR), diegeïnduceerd wordt na kolonisatie van de wortels door niet-pathogene bacteriën. Beide typengeïnduceerde resistentie zijn systemisch van aard, en zijn werkzaam tegen een breed, maarslechts ten dele overlappend spectrum van pathogenen. Pathogenen die gevoelig zijn voor SAR,zijn gevoelig voor SA-afhankelijke afweer. Daarentegen zijn pathogenen waartegen ISR werk-zaam is, meestal gevoelig voor JA- en ET-afhankelijke afweer. Ik bestudeerde de werkzaamheidvan door de avirulente bacterie Pseudomonas syringae pv. tomato geïnduceerde SAR en doorde Pseudomonas fluorescens stam WCS417r geïnduceerde ISR tegen de rupsen van het kleinkoolwitje (P. rapae) en de floridamot (Spodoptera exigua). De rupsen van het kleine koolwitjezijn gespecialiseerd in het eten van kruisbloemigen, een plantenfamilie waartoe ook de zand-raket behoort. De floridamot eet juist veel verschillende plantensoorten, en wordt daarom eengeneralist genoemd. Beide soorten zijn gevoelig voor JA-afhankelijke afweermechanismen,al is de floridamot gevoeliger dan het kleine koolwitje. Ik verwachtte daarom dat ISR wel enSAR niet werkzaam zou zijn tegen deze rupsen. Bovendien veronderstelde ik dat mogelijkeverschillen groter zouden zijn voor de floridamot dan voor het kleine koolwitje.

Resistentie tegen insecten treedt op op twee niveaus: directe verdediging, gericht tegenhet aanvallende insect, zoals de productie van toxines, en indirecte verdediging, gericht op hetaantrekken van vijanden van het aanvallende insect (“de vijand van mijn vijand is mijn vriend”).Dit laatste komt tot stand doordat de plant na vraat geurstoffen maakt die sluipwespen lokken.De sluipwespen leggen een eitje in de rups (parasitoïde), of ze doden de rups (rover). In mijnonderzoek heb ik de invloed van SAR en ISR op zowel directe als indirecte verdediging tegende rupsen van de twee plantenetende insectensoorten bestudeerd.

Directe verdedigingWanneer ik net uitgekomen rupsen van het kleine koolwitje liet eten van SAR-, ISR- en niet-geïnduceerde planten, zag ik dat de rupsen zich identiek ontwikkelden op planten van alledriede behandelingen. Op verschillende tijdspunten waren de rupsen even zwaar, en ook deontwikkeling tot pop verliep even snel. Echter, de rupsen van de floridamot bleven achter in groei op zowel SAR- als ISR-geïnduceerde planten in vergelijking tot onbehandelde planten.Ook verpopten de rupsen zich later op de planten met een ISR-behandeling. Analyse van deexpressie van de JA- en ET-afhankelijke genen PDF1.2 en HEL toonde aan dat vraat doorbeide soorten rupsen deze genen activeerde in onbehandelde planten. Echter, in zowel SAR-als ISR-geïnduceerde planten kwamen deze genen sterker verhoogd tot expressie na vraatdoor rupsen van de floridamot. Dit was niet het geval na vraat door rupsen van het kleinekoolwitje. De genen PDF1.2 en HEL waren dus ge”prime”d voor verhoogde expressie in zowelSAR- als ISR-geïnduceerde planten, specifiek na vraat door rupsen van de floridamot. Hetverschil in “priming” van door rupsenvraat geïnduceerde genen heb ik bevestigd door eenaantal extra genen te analyseren met behulp van genenchips. Deze experimenten heb ik uit-gevoerd in samenwerking met Natacha Bodenhausen en Philippe Reymond in Lausanne,

156

Zwitserland. De gevonden resultaten suggereren dat de effectiviteit van SAR en ISR tegen defloridamot wordt veroorzaakt door de verhoogde activiteit van afweergenen. Het effect vanSAR was onverwacht, maar kan worden verklaard doordat SAR biologisch werd geïnduceerddoor bacteriën (P. syringae pv. tomato) die bij infectie van Arabidopsis de productie van zowelSA als JA en ET verhogen. Als een volgende belager de productie van JA en ET verhoogt, ishet onderdrukkende effect van SA op de JA-signalering blijkbaar te verwaarlozen.

Indirecte verdedigingIn een windtunnel bekeek ik het effect van SAR en ISR op het gedrag van de sluipwesp Cotesiarubecula. Het induceren van SAR of ISR had geen effect op het gedrag van de sluipwespen.Na vraat door rupsen van het kleine koolwitje of de floridamot waren zowel SAR- of ISR-geïnduceerde als niet-geïnduceerde planten aantrekkelijker voor de sluipwespen dan plantenzonder vraat. Echter, de sluipwespen maakten geen onderscheid tussen SAR-, ISR- en niet-geïnduceerde planten. Het lijkt er dus op dat SAR en ISR geen effect hebben op de indirecteafweerrespons van Arabidopsis tegen deze twee soorten insecten.

Conclusie

Mijn proefschrift, samen met dat van Martin de Vos (2006), bevat een geïntegreerd onder-zoek naar geïnduceerde verdediging van Arabidopsis tegen een scala van pathogenen en insecten.We hebben aangetoond dat de signaalmoleculen SA, JA en ET in specifieke combinaties enhoeveelheden worden geproduceerd tijdens de verschillende interacties. De transcriptioneleveranderingen waren hoofdzakelijk belager-specifiek, maar de beïnvloede processen verrassendvergelijkbaar. Ondanks de verkregen kennis over de “hormoon handtekening” en de vertalingdaarvan in genexpressie profielen, bleek het lastig om de effectiviteit van de respons te voor-spellen. Zowel de effectiviteit van microbiëel geïnduceerde resistentie tegen de floridamot (S. exigua), als de door het kleine koolwitje (P. rapae) geïnduceerde resistentie tegen het virusTCV, zijn geassociëerd met “priming”. De geïntegreerde fytopathologische en entomologischeaanpak heeft geleid tot een beter begrip van de verdedigingsmechanismen van planten tegenpathogenen en insecten. Om het fenotype van de geïnduceerde respons te kunnen begrijpen,is kennis van de onderliggende mechanismen noodzakelijk. Daarom moet toekomstig onder-zoek zich richten op het ophelderen van de regulerende mechanismen die ten grondslag liggenaan de signaaltransductieroutes tijdens de geïnduceerde verdediging van planten.Tegelijkertijd moet de effectiviteit tegen verschillende soorten belagers bepaald worden. Diegecombineerde kennis zal het mogelijk maken het resultaat van specifieke plant-belager inter-acties te voorspellen.

157


Curriculum vitae Vivian van OostenBorn in Heinkenszand, on the 1st of November, 1976,Vivian Roosje van Oosten grew up in theDutch village ‘s-Heer Abtskerke in the province of Zeeland. Fascinated by plants, insects, birds,and basically all water creatures, it was no surprise that, in 1994, she moved to Amsterdamto study Biology at the Free University. She developed into a Biologist with broad interests,ranging from processes taking place in single cells to the role of biology in society. During twointernships,Vivian specialized in the molecular biology of plants interacting with their environ-ment.The first internship focussed on the cloning of a putative resistance gene of tomato againstthe fungus Alternaria alternata f.sp. lycopersici.The research was performed at the MolecularGenetics department of the Free University of Amsterdam, the Netherlands, and supervisedby Bas Brandwagt and Prof. Dr. J. Hille. The second project aimed to unravel the biochemicalproperties of light-signalling phytochromes in Arabidopsis thaliana. This research was super-vised by Dr. C. Fankhauser and Prof. Dr. J. Chory at the Department of Molecular and CellularBiology of Plants at the Salk Institute for Biological Studies, California, USA.

Vivian went on to learn an interdisciplinary approach between the natural and socialsciences at the Institute of Environmental Sciences of Leiden University, the Netherlands.This approach proved valuable for tackling an agricultural problem during a field study inCameroon,Africa. In 2000, she graduated from the Free University of Amsterdam as a Masterof Science in Biology, with two specializations: Molecular Plant Biology and EnvironmentalSocial Science. She continued her studies at the University of Geneva, Switzerland. Undersupervision of Prof. Dr. C. Fankhauser, she continued working on phytochrome signalling inArabidopsis. This specialization resulted in a DEA degree in Molecular Biology in 2001.

From 2002 till 2006, she worked on her PhD research “Induced pathogen and insectresistance in Arabidopsis: transcriptomics and specificity of defense”.This project was a colla-boration between the Laboratory of Entomology, Wageningen University, and the SectionPhytopathology, Utrecht University. She enjoyed the interdisciplinary approach of the PhDproject, i.e. the collaboration with Entomologists and Phytopathologists, and the variety oftechniques used for molecular plant biology and bioassays with insects and pathogens. Shealso performed a series of microarray experiments in collaboration with Natacha Bodenhausenand Dr. P. Reymond.To this end, she spent one month at the University of Lausanne, Switzerland.

During her PhD project,Vivian became interested in the translation of scientific researchon plant-insect interactions into applications for agriculture. Currently, she is based at thevegetable breeding and seed development company Nickerson-Zwaan in Made, the Netherlands.Nickerson-Zwaan is part of the French “Groupe Limagrain”, the fourth largest organisationin the world dedicated to the supply of seeds and seed products.Vivian works on a Post-docproject in Plant-Insect Interactions for the Scientific Committee of Limagrain.

Curriculum vitae

159

160

List of publications Van Oosten, V.R., Bodenhausen, N., Reymond, P., Van Pelt, J.A., Van Loon, L.C., Dicke, M. andPieterse, C.M.J. Differential effectiveness of microbially induced resistance against herbivorousinsects in Arabidopsis. Submitted.

Ton, J., Van der Ent, S., Van Hulten, M., Pozo, M., Van Oosten, V.R., Van Loon, L.C., Mauch-Mani, B., Turlings, T.C.J. and Pieterse, C.M.J. 2007. Priming as a mechanism behind induced resistanceagainst pathogens, insects and abiotic stress. In: Breeding for inducible resistance against pests and diseases (B. Mauch-Mani, M. Dicke, A. Schmidt, eds.), IOBC/wprs Bulletin, in press.

Pieterse, C.M.J., Van Pelt, J.A., Verhagen, B.W.M., De Vos, M., Van Oosten, V.R., Van der Ent, S., Koorneef, A., Van Hulten, M.H.A., Pozo, M.J., Ton, J., Dicke, M. and Van Loon, L.C. 2006.Molecular mechanisms involved in induced resistance signaling in Arabidopsis. In: Biology of Molecular Plant-Microbe Interactions, Vol. 5 (F. Sánchez, C. Quinto, I.M. López-Lara and O. Geiger,eds.), The International Society for Molecular Plant-Microbe Interactions, St. Paul, MN, pp. 188-194.

De Vos, M.*, Van Oosten, V.R.*, Van Poecke, R.M.P., Van Pelt, J.A., Pozo, M.J., Mueller, M.J., Buchala,A.J., Métraux, J.-P., Van Loon, L.C., Dicke, M. and Pieterse, C.M.J. 2005. Signal signature and transcriptome changes of Arabidopsis during pathogen and insect attack. Mol. Plant-Microbe Interact.18: 923-937. (* De Vos and Van Oosten contributed equally to this paper)

Van Oosten, V.R., Van Pelt, J.A., Van Loon, L.C., Pieterse, C.M.J. & Dicke, M. (2005). Micro-organismenbeschermen planten tegen rupsenvraat. Gewasbescherming 36 (6): 267.

Pieterse, C.M.J., Van Pelt, J.A., Van Wees, S.C.M., Ton, J., Verhagen, B.W.M., Léon-Kloosterziel, K., Hase, S., De Vos, M., Van Oosten, V.R., Pozo, M., Spoel, S., Van der Ent, S., Koornneef, A., Chalfun-Junior, A., Resende, M.L.V. and Van Loon, L.C. (2005). Indução de resistência sistêmica por rizobactérias e comunicação na rota de sinalização para uma defesa refinada. Revisão de Patologia de Plantas 13:277-319.

Pieterse, C.M.J., Van Pelt, J.A., Van Wees, S.C.M., Ton, J., Verhagen, B.W.M., Léon-Kloosterziel, K., Hase, S., De Vos, M., Van Oosten, V.R., Pozo, M., Spoel, S., Van der Ent, S., Koornneef, A. andVan Loon, L.C. (2004). Rhizobacteria-induced systemic resistance and pathway cross talk to fine-tunedefense. In: Proceedings of 2nd Brasilian Meeting on Induced Resistance in Plants, 9-11 November2004, University of Lavras, Lavras, Brazil, p. 44-58.

Van Oosten, V.R., De Vos, M., Van Pelt, J.A., Van Loon, L.C., Van Poecke, R.M.P., Dicke, M. andPieterse, C.M.J. (2004). Signal signature of Arabidopsis induced upon pathogen and insect attack. In: Biology of Plant-Microbe Interactions, Vol. 4 (I. Tikhonovich, B. Lugtenberg and N. Provorov, eds.),The International Society for Molecular Plant-Microbe Interactions, St. Paul, MN, pp. 199-202.

De Vos, M., Van Oosten, V.R., Van Pelt, J.A., Van Loon, L.C., Dicke, M. and Pieterse, C.M.J. (2004).Herbivore-induced resistance: differential effectiveness against a set of microbial pathogens inArabidopsis thaliana. In: Biology of Plant-Microbe Interactions, Vol. 4 (I. Tikhonovich, B. Lugtenbergand N. Provorov, eds.), The International Society for Molecular Plant-Microbe Interactions, St. Paul,MN, pp. 40-43.

Duek, P.D., Elmer, M., Van Oosten, V.R., and Fankhauser, C. (2004). The degradation of HFR1, aputative bHLH class transcription factor involved in light signalling, is regulated by phosphorylationand requires COP1. Current Biology 14 (24): 2296-2301.

List of publications

Education statement of the Graduate School Experimental Plant Sciences

161

Colophon

162

The research described in this thesis has been performed at the Laboratory of Entomology, Wageningen

University and the Section Phytopathology, Utrecht University. The research was supported, in part,

by grant 811.36.004 of the Earth and Life Sciences Foundation (ALW), which is subsidized by the

Netherlands Organisation of Scientific Research (NWO).

The printing of this thesis was financially supported by the Limagrain Group. www.limagrain.com

Cover and artwork were designed by Nevel Karaali graphic design, The Netherlands. www.nevel.nl

Layout of figures and tables by Marjolein Kortbeek-Smithuis, Beeldverwerking en Vormgeving,

Utrecht University.

Printed at Wöhrmann Print Service in Zutphen, The Netherlands.

vivian van oosten induced pathogen and insect resistance ...vivian van oosten induced pathogen and...

Documents