- Email: [email protected]
The systemin signaling pathway: differential activation of plant defensive genes
Biochimica et Biophysica Acta 1477 (2000) 112^121 www.elsevier.com/locate/bba
Review
The systemin signaling pathway: di¡erential activation of plant defensive genes Clarence A. Ryan Institute of Biological Chemistry, Washington State University, Pullman, WA 99164-6340, USA
Abstract Systemin, an 18-amino-acid polypeptide released from wound sites on tomato leaves caused by insects or other mechanical damage, systemically regulates the activation of over 20 defensive genes in tomato plants in response to herbivore and pathogen attacks. Systemin is processed from a larger prohormone protein, called prosystemin, by proteolytic cleavages. However, prosystemin lacks a signal sequence and glycosylation sites and is apparently not synthesized through the secretory pathway, but in the cytoplasm. The polypeptide activates a lipid-based signal transduction pathway in which the 18:3 fatty acid, linolenic acid, is released from plant membranes and converted to the oxylipin signaling molecule jasmonic acid. A wound-inducible systemin cell surface receptor with an Mr of 160 000 has recently been identified. The receptor regulates an intracellular cascade including, depolarization of the plasma membrane, the opening of ion channels, an increase in intracellular Ca2 , activation of a MAP kinase activity and a phospholipase A2 activity. These rapid changes appear to play important roles leading to the intracellular release of linolenic acid from membranes and its subsequent conversion to jasmonic acid, a potent activator of defense gene transcription. Although the mechanisms for systemin processing, release, and transport are still unclear, studies of the timing of the synthesis and of the intracellular localization of wound- and systemin-inducible mRNAs and proteins indicates that differential syntheses of signal pathway genes and defensive genes are occurring in different cell types. This signaling cascade in plants exhibits extraordinary analogies with the signaling cascade for the inflammatory response in animals. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Tomato; Prosystemin; Octadecanoid pathway; Systemic induction; Plant^herbivore interaction; Plant^pathogen interaction ; (Lycopersicon esculentum)
1. Introduction Plants have a variety of chemical defenses to protect themselves against herbivores [1,2]. These in-
Abbreviations: ABA, abscisic acid; ACC, 1-aminocyclopropane-carboxylic acid; AOS, allene oxide synthase; CaM, calmodulin; CPI, metallocarboxypeptidase inhibitor; CaMV, cauli£ower mosaic virus; CDI, cathepsin D inhibitor; GUS, L-glucuronidase; JA, jasmonic acid; LOX, lipoxygenase; MeJA, methyl jasmonate; OGA, oligogalacturonic acid; OPDA, 12-oxy-phytodienoic acid; PG, polygalacturonase; PLA2 , phospholipase A2 ; PPO, polyphenol oxidase
clude the production of chemicals, from small organics to large proteins and enzymes, that have various deterrent e¡ects on attacking herbivores that consume them. Other responses commonly found among many plant genera produce volatile signals at sites of larval attacks that attract insect predators such as wasps and mites [1]. Many plants respond to herbivore attacks by activating defense genes in leaves whose products inhibit digestive proteases of herbivores and reduce the nutritional quality of the ingested proteins, making the attackers ill [3]. In tomato plants, wounding causes a systemic reprogramming of leaf cells that results in
0167-4838 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 9 ) 0 0 2 6 9 - 1
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
113
Fig. 1. Systemic wound responsive genes. Asterisks indicate gene products identi¢ed in cell suspension cultures. Other genes or gene products have been identi¢ed in tomato leaves.
the synthesis of over 20 defense-related proteins [4^ 9]. This is analogous to the in£ammatory and acute phase responses of animals in response to trauma [4,10]. Most of the newly synthesized proteins fall into functional groups that include (i) antinutritional proteins, including both proteinase inhibitors and polyphenol oxidase (PPO), (ii) signal pathway components, and (iii) proteinases (Fig. 1). Several other genes are also activated, but they do not fall into the above groupings and their roles in defense are obscure. The wound-inducible proteinase inhibitors [6,7,9^ 11] cumulatively represent a wide range of speci¢cities that include all four known mechanistic classes of proteases. PPO is also wound-inducible in leaves [5] and, when ingested by herbivores together with phenolics, can crosslink proteins, rendering them less digestible. Together, the proteinase inhibitors and PPO provide formidable barriers to protein digestion. Wound-induced synthesis of components of the signal transduction pathway (cf. Section 3) appears to be a strategy of the plant to amplify its ability to mount a maximal defense response against the attacking predators. Wound-inducible signal pathway genes in tomato leaves include prosystemin [12], CaM [13], LOX [14,15], and AOS (G. Howe, C.A. Ryan, unpublished). The latter two enzymes are components of the octadecanoid pathway [16] and they reside in the chloroplast [14,15,17] where they participate in the conversion of linolenic acid to signaling molecules, phytodienoic acid and jasmonic acid, which are oxylipin analogs of prostaglandins [18]. Whether other members of the signal pathway-associated genes are also synthesized in response
to herbivore attacks has not been determined, but evidence using cell suspension cultures suggests that other signaling-associated genes in intact plants may also be wound-inducible, including the systemin receptor [19], ACC synthase (a key enzyme in ethylene biosynthesis) [20], and an NADPH oxidase (an enzyme involved in the production of active oxygen species) [21,22]. The ¢nal category of wound-inducible genes in Fig. 1 is comprised of several proteinases, including endopeptidases and exopeptidases. The function of these proteinases is still unclear, but some may be involved in protein processing, and others may have a role in protein turnover and remodeling. Knowledge of the localization of the synthesis of these proteinases in leaves may provide some clues to their functional signi¢cance in the defense response. 2. Systemin and prosystemin A primary wound signal for the signaling cascade is an 18-amino-acid polypeptide hormone called systemin [23] that is released at wound sites by chewing herbivores (wounding) (Fig. 2). Systemin is active at extraordinarily low levels (i.e., fmol/plant) [23] and ranks among the most potent gene activators known. Alanine scanning mutagenesis of the systemin sequence revealed that all residues except Thr17 could be replaced without completely eliminating activity. Ala17 is essentially inactive in inducing defense genes, but it is a potent antagonist of systemin, [24] suggesting that systemin-A17 can bind a receptor but cannot activate the signal transduction system. Ra-
BBAPRO 36067 23-2-00
114
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 2. Amino acid sequence of prosystemin. The underlined sequences indicate amino acids within conserved regions resulting from gene duplication^elongation events. The systemin sequence is indicated within the C-terminus.
dioactively labeled systemin, when placed on fresh wounds on tomato leaves has been shown to move from wound sites to petiole phloem, supporting its possible role as a mobile systemic wound signal [23,25]. Systemin is processed from a 200-aminoacid precursor protein called prosystemin [12] (Fig. 2). Prosystemin is found at low levels in leaves of unwounded plants, but increases several fold in response to wounding [12], apparently to amplify the wound signal in plants when under attack. The enzymes involved in the processing of prosystemin have not been identi¢ed, but possible processing enzymes may include one or more of the wound-inducible enzymes shown in Fig. 1. The processing apparently takes place as a result of wounding, which is thought to mix prosystemin with proteolytic enzymes from another cellular compartment(s), resulting in the release of systemin. A critical role for prosystemin and systemin in signaling defense genes has been established by transforming tomato plants with a fused gene comprised of a prosystemin cDNA in its antisense orientation, driven by the 35S CaMV promoter [12]. Plants ex-
pressing the gene are severely impaired in their systemic induction of both inhibitor I and II proteins in response to wounding. The plants were also compromised in their ability to defend themselves against attacking Manduca sexta larvae [26]. As expected, supplying antisense plants with systemin resulted in the expression of the defensive genes, con¢rming that the plants were capable of responding to the wound signal, but could not produce it. Additionally, placing systemin or recombinant prosystemin on fresh wounds of antisense plants reestablished the wound response to wild-type levels in distal leaves [27]. The characteristics of systemin, including its potency, its mobility in the plant, the wound-inducibility of the prosystemin gene, the e¡ects of the antisense gene in blocking systemic wound signaling, and the reestablishment of the systemic wound response in antisense plants, have led to the conclusion that systemin is a systemic wound hormone that plays a central role in regulating the expression of defense genes in response to pest attacks. Tomato plants transformed with the prosystemin cDNA in its correct or `sense' orientation regulated by a constitutive promoter expressed high levels of prosystemin mRNA in leaves [28]. Surprisingly, the plants constitutively synthesized wound-inducible defensive proteins throughout the plants in the absence of wounding, apparently due to an abnormal release of systemin from the overexpression of prosystemin. Inhibitors I and II were found to accumulate to extraordinary high levels of several hundred Wg inhibitor protein per g leaf tissue. Analyses of the proteins that accumulated in the transgenic plants (Fig. 3) have led to the identi¢cation of several systemin-inducible genes in tomato leaves that are shown in Fig. 1. When wild-type tomato plants were grafted onto the transgenic prosystemin rootstocks that overexpressed prosystemin, they synthesized high levels of proteinase inhibitors, indicating that systemin, or a defense signal produced by systemin, was transported through the graft to the wild-type plants where it also activated defense genes in the absence of wounding. Recombinant prosystemin, synthesized in Escherichia coli [29], was as active as systemin in inducing the synthesis of systemic wound response proteins when supplied to young tomato plants through their cut stems [27]. It is not known if full-length prosys-
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
115
tivity of systemin, while the prosystemin lacking the systemin sequence was totally inactive [27]. This indicated that the systemin sequence is likely the only sequence within prosystemin that can activate the signal transduction pathway. Prosystemin has been isolated from potato, black nightshade, and bell pepper [30]. All are homologous to tomato systemin, and sequence identities among them range from 73% to 88%. Systemins were synthesized from their deduced sequences and tested for their abilities to induce proteinase inhibitor synthesis in tomato. Potato and pepper were as active as tomato systemin, whereas nightshade systemin was 10fold less active. Tomato systemin is inactive in inducing proteinase inhibitor synthesis in leaves of tobacco, although a semi-puri¢ed polypeptide fraction from tobacco leaves is highly active in inducing proteinase inhibitor synthesis in tobacco leaves (G. Pearce, C.A. Ryan, unpublished). This suggests that either the systemin gene has changed signi¢cantly during evolution, or that other polypeptides may be present in tobacco that serve a similar signaling role as systemin in response to wounding. This suggests that if polypeptides are defense-signaling molecules throughout the plant kingdom, they may have minimally conserved structures or diverse structures that may require individual isolation and characterization. Fig. 3. SDS^PAGE of the soluble proteins extracted from leaves of wild-type (WT) and transgenic tomato plants constitutively expressing the prosystemin gene. Arrows indicate proteins that constitutively accumulate in the transgenic plants. These same proteins accumulate systemically in wild-type plants in response to wounding.
temin is active, or if a processed form interacts with the systemin receptor. Proteinases have been identi¢ed in apoplast £uids that can degrade prosystemin to small polypeptides that retain immunoreactivity against systemin antibodies [27]. Thus, it is likely that systemin is processed from prosystemin when supplied to plants via the transpiration stream. Additionally, recombinant prosystemin analogs were synthesized in E. coli in which either the prosystemin sequence was mutated at residue 195, corresponding Thr17 to Ala17 of systemin, or else the entire systemin sequence was deleted. The Ala17 analog exhibited less than 1% of the proteinase inhibitor inducing ac-
3. The systemin signaling pathway An updated original model [31] for the activation of defense genes by systemin in tomato plants is shown in Fig. 4. Using 125 I-labeled systemin, a systemin-binding protein has recently been identi¢ed in the plasma membranes of leaf cells [32] and, more recently, on the surface of Lycopersicon peruvianum suspension cultured cells [19]. The properties of the binding protein indicate that it is the receptor that is involved in systemin-mediated signal transduction. The dissociation constant of the systemin^receptor interaction with the suspension cultured cell receptor was 0.17 nM and is in the range of Kd s found for polypeptide^receptor interactions in animal systems. The receptor exhibited a speci¢c and reversible interaction with systemin, and the binding was shown to increase several fold in response to MJ, suggesting
BBAPRO 36067 23-2-00
116
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 4. A current model for the systemic signaling pathway for defensive genes in tomato plants that are activated by herbivore attacks (wounding). The interaction of systemin with its membrane receptor initiates intracellular events that activate a PLA2 . The phospholipase releases LA from membranes, the production of JA, and the activation of defensive genes.
that the receptor is a systemic wound response protein. An Mr of 160 000 was determined for the receptor by photoa¤nity labeling and electrophoretic analysis [19]. The isolation of the receptor protein and its cDNA should provide a new avenue of research to seek interactive signaling components to further elucidate the early events of signaling. The interaction of systemin with its receptor regulates a complex cascade of intracellular event s that are all orchestrated to activate a PLA2 to release linolenic acid from membranes. These events include a depolarization of the plasma membrane [33], the opening of ion channels [32^34], an increase in the concentration of intracellular Ca2 [34], the inactivation of a plasma membrane proton ATPase [35], the activation of a MAP kinase [36,37], the synthesis of calmodulin, and the activation of a PLA2 [38,39]. Release of linolenic acid from membranes leads to its conversion to the oxylipins OPDA and JA that regulate defense genes. JA, together with ethylene has been shown [40] to activate transcription of the defense-related genes by a mechanism that is still
obscure. Investigations of the individual components of this pathway are important not only for the understanding of the intracellular events leading to plant defense gene activation, but of events leading to stress and developmental responses that are regulated in plants by JA [18,41^44]. The identi¢cation of a systemin receptor on the cell surface raises important questions regarding the transport of systemin throughout the plants and its delivery to the surface of the distal cells. Although systemin was found to be mobile in the phloem, its long-distance transport following wounding is not well understood. For systemin to be regulating systemic activation of defense genes, it must be either transported throughout the plants or somehow involved in a cascade that operates over long distances. In any event, systemin appears to exert its e¡ects through an external cell surface receptor. If systemin is transported in the phloem then it is likely that it is transported to companion cells and phloem parenchyma where it could be delivered to the apoplast, presumably by a speci¢c transporter. On the other hand, when systemin is initially released by wounding it may travel through the apoplast to nearby vascular bundle cells where it can interact with receptors to activate the synthesis in the cells of both prosystemin and its processing enzymes. This in turn would produce more systemin, which could move through the phloem and apoplast to distal cells to continue a cascade through the plant. Another possibility is that the initially released systemin might activate synthesis of prosystemin and a prosystemin transporter that would deliver nascent prosystemin to the apoplast where processing and receptor binding would both occur. In this scenario, the transport of systemin through the plant would occur by apoplastic transport and phloem loading to e¡ect transport to more distal cells, cascading through the plant. This latter scenario would require systemin to be transported from the phloem to the apoplast to interact with the receptor. Both scenarios ¢t several known facts, including the timing of movement of the wound signal (systemin) through the petioles (from 15 to 90 min), the tissue speci¢city of prosystemin synthesis in the vascular bundles (see Section 4), the movement of the wound signal from prosystemin transgenic rootstalks through grafts to wildtype scions in the absence of wounding, and the ob-
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
servations that the addition of systemin or prosystemin to wound sites on leaves of prosystemin antisense tomato plants caused distal leaves to activate defense genes [27]. Other primary signals have been proposed for activating defense genes in response to herbivore attacks, including electrical pulses, changes in hydraulic pressure, and the plant hormones ABA and ethylene [45^48]. Electrical phenomena accompany wounding but no evidence has been forthcoming to de¢ne how weak electrical signals can play a role in systemic signal transduction. The propagation of both electrical and hydraulic pulses are very fast compared to the known velocity of the movement of the wound signal in plants [49]. Additionally, girdling petioles of young tomato plants with a stream of hot air that kills all cells except the xylem elements prevented systemic signaling of proteinase inhibitor I in upper, ungirdled leaves in response to wounding [49]. Under these conditions, hydraulic signals would have been able to move through the xylem to unwounded leaves. Malone and Alorcan [50] reported that wounding girdled tomato leaves induced inhibitory activity against elastase in upper, unwounded leaves. It is possible that an entirely di¡erent signaling pathway is activating elastase inhibitors, but these latter experiments did not give consistent results, and speci¢c inhibitor quanti¢cation was not recorded, so that basal levels of inhibitor at the onset of experiments could not be assessed. There is no doubt that electrical and hydraulic waves do propagate throughout the plants in response to wounding, but their e¡ects on plant defense activation still remains obscure. In this regard, in the grafted plants mentioned earlier in which `sense' prosystemin transgenic rootstalks caused the induction of proteinase inhibitors in the wild-type scions in the absence of wounding, no generation of electrical or hydraulic signals could have occurred. In wild-type plants, breaching the vascular system has been proposed to perturb membranes throughout the plants [37], so that a priming of the signal transduction system through a hydraulic event might e¡ect an early transient activation of the MAP kinases and the PLA2 that are involved in the early signaling events, similar to a `touch response'. Transient signaling in response to hydraulic signals may be part of an early alarm system to activate signaling
117
pathway components, but weak, transient, electrical, or hydraulic signals alone cannot sustain the activation of defense genes over a period of several hours. ABA appears to play an overall role in determining whether the plant can mount a defense response [51,52]. Tomato plants with a mutation in ABA biosynthesis were found to be de¢cient in the wound response, and the response could be restored by exogenously ABA applied to leaves [51]. ABA does not behave as if it is a primary component of the signaling pathway since it does not act as an inducer of proteinase inhibitor synthesis in tomato plants, as do wounding, systemin, or other elicitors [52], but it is required for the plants to respond maximally [51]. Other plant hormones are also important to the defense response, including auxin, which inhibits the response [53], and ethylene [40] which is a component of the signaling pathway. 4. Di¡erential signaling by systemin ^ a modi¢ed model In comparing the timing of the wound-inducible expression of proteinase inhibitor genes with those of the signal transduction pathway, it became clear that two classes of wound- and systemin-inducible genes were di¡erentially regulated. As shown in Fig. 5, mRNAs coding for proteinase proteins mRNAs appeared much later than mRNAs coding for signal pathway proteins. The mRNAs encoding proteinase inhibitors I and II [11], CPI ([54]; M. Diaz, C.A. Ryan, in preparation), and an aspartic proteinase inhibitor, CDI [4,7], were ¢rst detectable in northern analyses at about 2 h after the plants were wounded, and the levels continued to increase through the next 8 h. On the other hand, mRNAs encoding wound-inducible LOX [15], CaM [13], AOS (G. Howe, C.A. Ryan, unpublished), and prosystemin [12] were ¢rst detectable within about 30 min, and were maximally induced 2^3 h after wounding. This was 5^6 h before the proteinase inhibitor mRNAs reached maximal levels. AOS mRNA has been reported to be wound inducible in Arabidopsis leaves [55] with induction kinetics similar to LOX mRNAs in tomato leaves, beginning at 0.5 h after wounding and maximizing at about 2^4 h, then declining thereafter.
BBAPRO 36067 23-2-00
118
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 5. The di¡erential induction of mRNAs of systemic wound response proteins. Young tomato plants were wounded on the lower leaves at time zero and assayed by Northern blotting at the times indicated. Inh I, proteinase inhibitor I; Inh II, proteinase inhibitor II; CDI, cathepsin D inhibitor (and aspartic proteinase inhibitor); CPI, metallocarboxypeptidase inhibitor; CaM, calmodulin; LOX, lipoxygenase; ProSys, prosystemin.
Ultrastructural studies of the intracellular localization of proteinase inhibitor I, synthesized in response to wounding, had revealed previously that the inhibitor is sequestered in the central vacuoles of mesophyll cells of tomato leaves [56]). Inhibitor II has also been found in leaf vacuoles [57,58], and it is likely that other proteinase inhibitors with long half lives are also sequestered in vacuoles of mesophyll and/or palisade cells. In contrast, when the promoter region of the wound-inducible prosystemin gene was fused with the GUS reporter gene, and tomato plants were transformed with the fused gene, the plants responded to wounding and MeJA by synthesizing the GUS protein [59]. The localization of GUS activity was visualized by supplying a colorimetric substrate to freehand sections of leaf tissues. The wound-inducible GUS activity was only found in the vascular bundles of the major and minor veins of the leaves. This was con¢rmed by tissue printing
using speci¢c antibodies prepared against prosystemin and visualized before and after wounding using a colorimetric substrate. The AOS gene in Arabidopsis [55] that is a component of the defense signaling pathway is also speci¢cally expressed in the vascular bundles in response to wounding (E. Weiler, personal communication). We hypothesize that the signal transduction genes as a group are `early genes' that are expressed in the vascular bundle cells, while the defensive proteinase inhibitor genes are `late genes' that are expressed in palisade and spongy mesophyll cells (Fig. 6). The di¡erences in timing and localization of these two functionally di¡erent groups of proteins suggests a scenario in which the signal transduction pathway is initially activated in the vascular bundles to amplify the production of second messengers, which could include OPDA and/or JA, derivatives of these molecules [60], or other signaling molecules that are transported to other cell types. The second messengers are then transported to the palisade and mesophyll cells where the defense proteins are synthesized and compartmented. A comparison of the relative time courses of the various components of the signal transduction pathway from the initial release of systemin to the production of proteinase inhibitors is shown in Fig. 7. This scenario di¡ers signi¢cantly from a recently proposed signaling model [48] in which systemin was suggested to be a signal released in response to jasmonic acid. In this latter model, jasmonic acid and ethylene were not considered downstream systemic signals because increases in ethylene and jasmonic acid cannot be detected levels in unwounded leaves of wounded plants [48]. The author did not consider that JA is likely transient and only present at very
Fig. 6. A modi¢ed model to indicate the di¡erential signaling of signal pathway components and the proteinase inhibitors in di¡erent cell types.
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 7. A time-course representation of the progression of the production of signal pathway components in tomato leaves in response to wounding at time 0. H, a transient hydraulic pulse due to the breaching of the vascular system by wounding, which initiates the signaling events.
low concentrations in speci¢c cell types. There is no evidence available to bear on the minimum quantities of these compounds that can activate genes, or whether OPDA or JA is the signaling component. Several examples are known of JA-inducible genes being activated without measurable increases in JA. As stated by Weiler et al. [61], `clearly, results based solely on analysis of JA can no longer be regarded as complete'. Bowles also concluded [48], based on unpublished data, that Ala17-systemin, the potent antagonist of systemin mentioned earlier, inhibits the action of OGA elicitors. In contrast, through extensive experimentation (G. Pearce, C.A. Ryan, unpublished), we have shown conclusively that Ala17-systemin has no e¡ect on the proteinase inhibitor inducing activity of either OGA or chitosan, and therefore systemin must be perceived independently from these elicitors, likely through di¡erent receptors. A recent report has demonstrated that chitosan, OGA, and systemin all activate PLA2 activity in tomato plants [39], and all three elicitors act via the octadecanoid pathway [62] although it is not yet known if the same PLA2 is involved with signaling by all three elicitors. While the downstream message that activates pro-
119
teinase inhibitor synthesis in response to wounding and systemin may be JA or a derivative in conjunction with ethylene, new evidence from our research suggests that other candidates for second messengers are produced in response to wounding and systemin that could be part of the signaling pathway for proteinase inhibitor synthesis. The candidates are (i) the OGA derived from the cell walls of vascular bundle cells by the action of a newly discovered wound-, systemin- and JA-inducible PG [6], and (ii) reactive oxygen species [22], known to be produced in tomato leaves by OGAs [63], through NADPH oxidase [63^ 65]. A speci¢c inhibitor of NADPH oxidase, diphenylene iodonium chloride, blocks wound-inducible H2 O2 generation in tomato leaves [22], and we have recently shown that this inhibitor powerfully blocks the synthesis of inhibitor I and II protein accumulation in response to wounding, systemin, and MeJA (M. Orozco-Cardenas, C.A. Ryan, in preparation), although whether the inhibitor is speci¢c for NADPH oxidase in vivo is not known. If OGA and H2 O2 are indeed second messengers for the activation of proteinase inhibitor genes in mesophyll cells, they are likely downstream in the signaling pathway since they are inducible by JA and PG in the vascular bundles. The production of OGA and H2 O2 as second messengers is kinetically compatible with the synthesis of proteinase inhibitors since both are produced later than the signal transduction mRNAs ([22]; see Fig. 5), but concurrent with the synthesis of proteinase inhibitor mRNAs (i.e., between 2 and 8 h after wounding). Reactive oxygen plays a central role in the resistance response of plants against pathogens leading to incompatibility and the development of systemic acquired resistances [66]. The increase in active oxygen during herbivore attacks has been suggested to act as a deterrent to herbivores [66], and the wound-inducible reactive oxygen might potentiate the defense response against pathogens as well. While there is no evidence to date to link wound-inducible H2 O2 in tomato leaves to pathogen resistances, the octadecanoid pathway is important to phytoalexin production in several species of plants [67,68] and the pathway has been shown through mutant analyses to be a factor in pathogen resistance in Arabidopsis [69^72]. In a broader context, the defense signaling pathways in plants contain common types of components
BBAPRO 36067 23-2-00
120
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
that have contributed to the evolution of information processing systems for the activation of defenses against both herbivores and pathogens in particular ecological niches. For example, receptors have evolved in plants that can recognize signals unique to speci¢c pests or pathogens [73^76]. These signals regulate the tissue-speci¢c and/or cell-speci¢c expression of a variety of defensive genes in di¡erent plants to protect against speci¢c herbivores and/or pathogens. Herbivore saliva can be a source of signals to activate defense genes [77], or to signal the production of chemicals to attract predators of insects chewing on the plants [78,79]. While the various signals and genes produced by herbivore and pathogen attacks are likely to di¡er among di¡erent plant species, it is also likely that some other signal pathway components, e.g., ion channels, MAP kinases, and transcription factors, will have evolved with some conservation of structures and functions. Perhaps with time these fundamental components may track back to ancestral genes common to plant and animal defense systems. Knowledge of the biochemical details of the evolution of defense signaling pathways of plants could have an enormous potential in helping to understand the biological characteristics of species that are found in various ecosystems, as well as in applying this knowledge to the agricultural community to increase plant productivity using natural defense strategies Acknowledgements Supported in part by Washington State University College of Agriculture and Home Economics, Project 1791 and Grants DCB 9104542 and DCB 9117795 from the National Science Foundation, and Grant #9801502 from the U.S. Department of Agriculture.
References [1] R. Karban, I.T. Baldwin (Eds.), Induced Responses to Herbivory, University of Chicago Press, Chicago, 1997. [2] Herbivores: Their Interaction with Secondary Plant Metabolites. (1979) G.A. Rosenthal and D.H. Janzen (Eds.), Academic Press, NY. [3] C.A. Ryan, Annu. Rev. Phytopathol. 28 (1990) 425^449.
[4] D. Bergey, G. Howe, C.A. Ryan, Proc. Natl. Acad. Sci. USA 93 (1996) 12053^12058. [5] C.P. Constabel, D.R. Bergey, C.A. Ryan, in: J. Romeo et al. (Eds.), Phytochemical Diversity and Redundancy in Ecological Interactions, Plenum Press, New York, 1996, pp. 231^ 251. [6] D.R. Bergey, M. Orozco-Cardenas, D. Moura, C.A. Ryan, Proc. Natl. Acad. Sci. USA 96 (1999) 1756^1760. [7] T. Hildemann, M. Ebneth, H. Pen¬a-Cortes, J.J. SanchezSerrano, L. Wilmitzer, S. Prat, Plant Cell 4 (1992) 1157^ 1170. [8] V. Pautoªt, F.M. Holzer, B. Reisch, L.L. Walling, Proc. Natl. Acad Sci. USA 90 (1993) 9906^9910. [9] C.J. Bolter, Plant Physiol. 103 (1993) 1347^1353. [10] C.A. Ryan, G. Pearce, Annu. Rev. Cell Dev. Biol. 14 (1998) 1^17. [11] J.S. Graham, G. Hall, G. Pearce, C.A. Ryan, Planta 169 (1986) 399^405. [12] B. McGurl, G. Pearce, M. Orozco-Cardenas, C.A. Ryan, Science 255 (1992) 1570^1573. [13] D.R. Bergey, C.A. Ryan, Plant Mol. Biol. 40 (1999) 815^ 823. [14] J. Royo, G. Vancanney, A.G. Perez, C. Sanz, K. Stormann, S. Rosahl, J.J. Sanchez-Serrano, J. Biol. Chem. 271 (1996) 21012^21019. [15] T. Heitz, D. Bergey, C.A. Ryan, Plant Physiol. 114 (1997) 1805^1903. [16] B. Vick, D.C. Zimmerman, Plant Physiol. 75 (1984) 458^461. [17] E. Blee, J. Joyard, Plant Physiol. 110 (1996) 445^454. [18] M. Hamberg, H.W. Gardner, Biochim. Biophys. Acta 1165 (1992) 1^18. [19] J.M. Scheer, C.A. Ryan, Plant Cell 11 (1999) 1525^1535. [20] G. Felix, T. Boller, Plant J. 7 (1995) 381^389. [21] M.J. Stennis, C. Sreeganga, C.A. Ryan, P.S. Low, Plant Physiol. 117 (1998) 1031^1036. [22] M. Orozco-Cardenas, C.A. Ryan, Proc. Natl. Acad. Sci. USA 96 (1999) 6553^7667. [23] G. Pearce, D. Strydom, S. Johnson, C.A. Ryan, Science 253 (1991) 895^897. [24] G. Pearce, S. Johnson, C.A. Ryan, J. Biol. Chem. 268 (1993) 212^216. [25] J. Narva©ez-Vasque©z, G. Pearce, M.L. Orozco-Cardenas, V.R. Franceschi, C.A. Ryan, Planta 195 (1995) 593^600. [26] M. Orozco-Cardenas, B. McGurl, C.A. Ryan, Proc. Natl. Acad. Sci. USA 90 (1993) 8273^8276. [27] J.E. Dombrowski, G. Pearce, C.A. Ryan, Proc. Natl. Acad. Sci. USA 96 (1999) 12947^12952. [28] B. McGurl, M. Orozco-Cardenas, G. Pearce, C.A. Ryan, Proc. Natl. Acad. Sci. USA 91 (1994) 9799^9802. [29] J. Delano-Frier, J.E. Dombrowski, C.A. Ryan, Protein Exp. Purif. 17 (1999) 74^82. [30] C.P. Constabel, L. Yip, C.A. Ryan, Plant Mol. Biol. 36 (1998) 55^62. [31] E.E. Farmer, C.A. Ryan, Plant Cell 4 (1992) 129^134. [32] T. Meindl, T. Boller, G. Felix, Plant Cell 10 (1998) 1561^ 1570.
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121 [33] C. Moyen, E. Johannes, Plant Cell Environ. 19 (1996) 464^ 470. [34] C. Moyen, K.E. Hammond-Kosack, J. Jones, M.R. Knight, E. Johannes, Plant Cell Environ. 21 (1998) 1101^1111. [35] A. Schaller, C. Oecking, Plant Cell 11 (1999) 263^272. [36] S. Usami, H. Banno, Y. Ito, R. Nishihama, Y. Machida, Proc. Natl. Acad. Sci. USA 92 (1995) 8660^8664. [37] J.W. Stratmann, C.A. Ryan, Proc. Natl. Acad. Sci. USA 94 (1997) 11085^11089. [38] S. Lee, S. Suh, S. Kim, R.C. Crain, J.M. Kwak, H.-G. Nam, Y. Lee, Plant J. 12 (1997) 547^556. [39] J. Narva¨ez-Vasquez, J. Florin-Christensen, C.A. Ryan, The Plant Cell 11 (1999) 1^13. [40] P.J. O'Donnell, C.M. Calvert, R. Atzorn, C. Wasternack, H.M.O. Leyser, D.J. Bowles, Science 274 (1998) 1914^1917. [41] E.W. Weiler, Naturwissenschaften 84 (1997) 340^349. [42] T. Farmer, Plant Mol. Biol. 26 (1994) 1423^1437. [43] R.A. Creelman, J.E. Mullet, Annu. Rev. Plant Physiol. Plant Biol. 48 (1997) 355^381. [44] S. Seo, H. Sano, O. Ohashi, Physiol. Plant. 101 (1997) 740^ 745. [45] D.C. Wildon, J.F. Thain, P.E.H. Michin, I. Gubb, A. Reilly, Y. Skipper, H. Doherty, P. O'Donnell, D.J. Bowles, Nature 360 (1993) 62^65. [46] O. Herde, H. Pen¬a Cortes, C. Wasternack, L. Willmitzer, J. Fisahn, Plant Physiol. 119 (1999) 213^218. [47] M. Malone, L. Palumbo, F. Boari, M. Monteleone, H.G. Jones, Plant Cell Environ. 17 (1994) 81^87. [48] D. Bowles, Phil. Trans. R. Soc. Lond. B 353 (1998) 1495^ 1510. [49] C. Nelson, M. Walker-Simmons, D. Makus, G. Zuroske, J. Graham, C.A. Ryan, in: P. Hedin (Ed.), Mechanisms of Plant Resistance to Insects, ACS Monograph, 1983, pp. 103^122. [50] M. Malone, J.-J. Alarcon, Planta 196 (1995) 740^746. [51] O. Herde, R. Atzorn, J. Fisahn, C. Wasternack, L. Willmitzer, H. Pena-Cortes, Plant Physiol. 112 (1996) 853^860. [52] G.F. Birkenmeier, C.A. Ryan, Plant Physiol. 117 (1998) 687^693. [53] A. Kernan, R.W. Thornburg, Plant Physiol. 91 (1989) 73^78. [54] B. Martineau, K.E. McBride, C.M. Mouck, Mol. Gen. Genet. 228 (1991) 281^286. [55] D.L. Laudert, U. Pfannschmidt, F. Lottspeich, H. Hollander-Czytko, E.W. Weiler, Plant Mol. Biol. 31 (1996) 323^335.
121
[56] L.K. Shumway, J.M. Rancour, C.A. Ryan, Planta 93 (1970) 1^14. [57] M.K. Walker-Simmons, C.A. Ryan, Plant Physiol. 60 (1977) 61^63. [58] G.-Y. Jauh, A.M. Fischer, H.D. Grimes, C.A. Ryan, J.C. Rogers, Proc. Natl. Acad. Sci. USA 95 (1998) 12995^12999. [59] T. Jacinto, B. McGurl, V. Franceschi, J. Delano-Freier, C.A. Ryan, Planta 203 (1997) 406^411. [60] S. Parchmann, H. Gundlach, M.J. Mueller, Plant Physiol. 115 (1997) 1057^1064. [61] E.W. Weiler, Recent Adv. Phytochem. 32 (1998) 179^1205. [62] S. Doares, T. Syrovets, E.W. Weiler, C.A. Ryan, Proc. Natl. Acad. Sci. USA 92 (1995) 4095^4098. [63] M.J. Stennis, C. Sreeganga, C.A. Ryan, P.S. Low, Plant Physiol. 117 (1998) 1031^1036. [64] T. Keller, H.G. Damude, D. Werner, P. Doener, R.A. Dixon, C. Lamb, Plant Cell 10 (1998) 255^266. [65] C. Lamb, R.A. Dixon, Annu. Rev. Plant Physiol. Plant Mol. Biol. 48 (1997) 251^275. [66] J.L. Bi, G.W. Felton, J. Chem. Ecol. 21 (1995) 1511^1530. [67] H. Gundlach, M.J. Muller, T.M. Kutchan, M.H. Zenk, Proc. Natl. Acad. Sci. USA 89 (1992) 2389^2393. [68] M.J. Mueller, W. Brodschelm, E. Spannagl, M.H. Zenk, Proc. Natl. Acad. Sci. USA 90 (1993) 7490^7494. [69] V. Perumal, J. Shockey, C.A. Levesque, R.J. Cook, J. Browse, Proc. Natl. Acad. Sci. USA 95 (1998) 7209^7214. [70] P.E. Staswick, G.Y. Yuen, C.C. Lehman, Plant J. 15 (1998) 747^754. [71] B.P.H.J. Thomma, K. Eggermont, I.A.M.A. Penninckz, B. Mauchmani, R. Vogelsang, P.P.A. Cammue, W.F. Broekaert, Proc. Natl. Acad. Sci. USA 95 (1998) 15107^15111. [72] J.S. Thaler, Nature 399 (1999) 686^688. [73] D. Nennstiel, D. Scheel, T. Nurnberger, FEBS Lett. 431 (1998) 405^410. [74] N. Umemoto, M. Kakitani, A. Iwamatsu, M. Yoshikawa, N. Yamaoka, I. Ishida, Proc. Natl. Acad. Sci. USA 94 (1997) 1029^1034. [75] J. Ebel, BioEssays 20 (1998) 569^576. [76] A. Fath, T. Boller, Plant Physiol. 112 (1996) 1659^1668. [77] K.L. Korth, R.A. Dixon, Plant Physiol. 115 (1997) 1299^ 1305. [78] C.M. De Moraes, W.J. Lewis, P.W. Pare, H.T. Abora, J.H. Tumlinson, Nature 393 (1998) 570^573. [79] L. Mattiacci, M. Dicke, M.A. Posthumus, Proc. Natl. Acad. Sci. USA 92 (1995) 2036^2040.
BBAPRO 36067 23-2-00
Review
The systemin signaling pathway: di¡erential activation of plant defensive genes Clarence A. Ryan Institute of Biological Chemistry, Washington State University, Pullman, WA 99164-6340, USA
Abstract Systemin, an 18-amino-acid polypeptide released from wound sites on tomato leaves caused by insects or other mechanical damage, systemically regulates the activation of over 20 defensive genes in tomato plants in response to herbivore and pathogen attacks. Systemin is processed from a larger prohormone protein, called prosystemin, by proteolytic cleavages. However, prosystemin lacks a signal sequence and glycosylation sites and is apparently not synthesized through the secretory pathway, but in the cytoplasm. The polypeptide activates a lipid-based signal transduction pathway in which the 18:3 fatty acid, linolenic acid, is released from plant membranes and converted to the oxylipin signaling molecule jasmonic acid. A wound-inducible systemin cell surface receptor with an Mr of 160 000 has recently been identified. The receptor regulates an intracellular cascade including, depolarization of the plasma membrane, the opening of ion channels, an increase in intracellular Ca2 , activation of a MAP kinase activity and a phospholipase A2 activity. These rapid changes appear to play important roles leading to the intracellular release of linolenic acid from membranes and its subsequent conversion to jasmonic acid, a potent activator of defense gene transcription. Although the mechanisms for systemin processing, release, and transport are still unclear, studies of the timing of the synthesis and of the intracellular localization of wound- and systemin-inducible mRNAs and proteins indicates that differential syntheses of signal pathway genes and defensive genes are occurring in different cell types. This signaling cascade in plants exhibits extraordinary analogies with the signaling cascade for the inflammatory response in animals. ß 2000 Elsevier Science B.V. All rights reserved. Keywords: Tomato; Prosystemin; Octadecanoid pathway; Systemic induction; Plant^herbivore interaction; Plant^pathogen interaction ; (Lycopersicon esculentum)
1. Introduction Plants have a variety of chemical defenses to protect themselves against herbivores [1,2]. These in-
Abbreviations: ABA, abscisic acid; ACC, 1-aminocyclopropane-carboxylic acid; AOS, allene oxide synthase; CaM, calmodulin; CPI, metallocarboxypeptidase inhibitor; CaMV, cauli£ower mosaic virus; CDI, cathepsin D inhibitor; GUS, L-glucuronidase; JA, jasmonic acid; LOX, lipoxygenase; MeJA, methyl jasmonate; OGA, oligogalacturonic acid; OPDA, 12-oxy-phytodienoic acid; PG, polygalacturonase; PLA2 , phospholipase A2 ; PPO, polyphenol oxidase
clude the production of chemicals, from small organics to large proteins and enzymes, that have various deterrent e¡ects on attacking herbivores that consume them. Other responses commonly found among many plant genera produce volatile signals at sites of larval attacks that attract insect predators such as wasps and mites [1]. Many plants respond to herbivore attacks by activating defense genes in leaves whose products inhibit digestive proteases of herbivores and reduce the nutritional quality of the ingested proteins, making the attackers ill [3]. In tomato plants, wounding causes a systemic reprogramming of leaf cells that results in
0167-4838 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 3 8 ( 9 9 ) 0 0 2 6 9 - 1
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
113
Fig. 1. Systemic wound responsive genes. Asterisks indicate gene products identi¢ed in cell suspension cultures. Other genes or gene products have been identi¢ed in tomato leaves.
the synthesis of over 20 defense-related proteins [4^ 9]. This is analogous to the in£ammatory and acute phase responses of animals in response to trauma [4,10]. Most of the newly synthesized proteins fall into functional groups that include (i) antinutritional proteins, including both proteinase inhibitors and polyphenol oxidase (PPO), (ii) signal pathway components, and (iii) proteinases (Fig. 1). Several other genes are also activated, but they do not fall into the above groupings and their roles in defense are obscure. The wound-inducible proteinase inhibitors [6,7,9^ 11] cumulatively represent a wide range of speci¢cities that include all four known mechanistic classes of proteases. PPO is also wound-inducible in leaves [5] and, when ingested by herbivores together with phenolics, can crosslink proteins, rendering them less digestible. Together, the proteinase inhibitors and PPO provide formidable barriers to protein digestion. Wound-induced synthesis of components of the signal transduction pathway (cf. Section 3) appears to be a strategy of the plant to amplify its ability to mount a maximal defense response against the attacking predators. Wound-inducible signal pathway genes in tomato leaves include prosystemin [12], CaM [13], LOX [14,15], and AOS (G. Howe, C.A. Ryan, unpublished). The latter two enzymes are components of the octadecanoid pathway [16] and they reside in the chloroplast [14,15,17] where they participate in the conversion of linolenic acid to signaling molecules, phytodienoic acid and jasmonic acid, which are oxylipin analogs of prostaglandins [18]. Whether other members of the signal pathway-associated genes are also synthesized in response
to herbivore attacks has not been determined, but evidence using cell suspension cultures suggests that other signaling-associated genes in intact plants may also be wound-inducible, including the systemin receptor [19], ACC synthase (a key enzyme in ethylene biosynthesis) [20], and an NADPH oxidase (an enzyme involved in the production of active oxygen species) [21,22]. The ¢nal category of wound-inducible genes in Fig. 1 is comprised of several proteinases, including endopeptidases and exopeptidases. The function of these proteinases is still unclear, but some may be involved in protein processing, and others may have a role in protein turnover and remodeling. Knowledge of the localization of the synthesis of these proteinases in leaves may provide some clues to their functional signi¢cance in the defense response. 2. Systemin and prosystemin A primary wound signal for the signaling cascade is an 18-amino-acid polypeptide hormone called systemin [23] that is released at wound sites by chewing herbivores (wounding) (Fig. 2). Systemin is active at extraordinarily low levels (i.e., fmol/plant) [23] and ranks among the most potent gene activators known. Alanine scanning mutagenesis of the systemin sequence revealed that all residues except Thr17 could be replaced without completely eliminating activity. Ala17 is essentially inactive in inducing defense genes, but it is a potent antagonist of systemin, [24] suggesting that systemin-A17 can bind a receptor but cannot activate the signal transduction system. Ra-
BBAPRO 36067 23-2-00
114
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 2. Amino acid sequence of prosystemin. The underlined sequences indicate amino acids within conserved regions resulting from gene duplication^elongation events. The systemin sequence is indicated within the C-terminus.
dioactively labeled systemin, when placed on fresh wounds on tomato leaves has been shown to move from wound sites to petiole phloem, supporting its possible role as a mobile systemic wound signal [23,25]. Systemin is processed from a 200-aminoacid precursor protein called prosystemin [12] (Fig. 2). Prosystemin is found at low levels in leaves of unwounded plants, but increases several fold in response to wounding [12], apparently to amplify the wound signal in plants when under attack. The enzymes involved in the processing of prosystemin have not been identi¢ed, but possible processing enzymes may include one or more of the wound-inducible enzymes shown in Fig. 1. The processing apparently takes place as a result of wounding, which is thought to mix prosystemin with proteolytic enzymes from another cellular compartment(s), resulting in the release of systemin. A critical role for prosystemin and systemin in signaling defense genes has been established by transforming tomato plants with a fused gene comprised of a prosystemin cDNA in its antisense orientation, driven by the 35S CaMV promoter [12]. Plants ex-
pressing the gene are severely impaired in their systemic induction of both inhibitor I and II proteins in response to wounding. The plants were also compromised in their ability to defend themselves against attacking Manduca sexta larvae [26]. As expected, supplying antisense plants with systemin resulted in the expression of the defensive genes, con¢rming that the plants were capable of responding to the wound signal, but could not produce it. Additionally, placing systemin or recombinant prosystemin on fresh wounds of antisense plants reestablished the wound response to wild-type levels in distal leaves [27]. The characteristics of systemin, including its potency, its mobility in the plant, the wound-inducibility of the prosystemin gene, the e¡ects of the antisense gene in blocking systemic wound signaling, and the reestablishment of the systemic wound response in antisense plants, have led to the conclusion that systemin is a systemic wound hormone that plays a central role in regulating the expression of defense genes in response to pest attacks. Tomato plants transformed with the prosystemin cDNA in its correct or `sense' orientation regulated by a constitutive promoter expressed high levels of prosystemin mRNA in leaves [28]. Surprisingly, the plants constitutively synthesized wound-inducible defensive proteins throughout the plants in the absence of wounding, apparently due to an abnormal release of systemin from the overexpression of prosystemin. Inhibitors I and II were found to accumulate to extraordinary high levels of several hundred Wg inhibitor protein per g leaf tissue. Analyses of the proteins that accumulated in the transgenic plants (Fig. 3) have led to the identi¢cation of several systemin-inducible genes in tomato leaves that are shown in Fig. 1. When wild-type tomato plants were grafted onto the transgenic prosystemin rootstocks that overexpressed prosystemin, they synthesized high levels of proteinase inhibitors, indicating that systemin, or a defense signal produced by systemin, was transported through the graft to the wild-type plants where it also activated defense genes in the absence of wounding. Recombinant prosystemin, synthesized in Escherichia coli [29], was as active as systemin in inducing the synthesis of systemic wound response proteins when supplied to young tomato plants through their cut stems [27]. It is not known if full-length prosys-
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
115
tivity of systemin, while the prosystemin lacking the systemin sequence was totally inactive [27]. This indicated that the systemin sequence is likely the only sequence within prosystemin that can activate the signal transduction pathway. Prosystemin has been isolated from potato, black nightshade, and bell pepper [30]. All are homologous to tomato systemin, and sequence identities among them range from 73% to 88%. Systemins were synthesized from their deduced sequences and tested for their abilities to induce proteinase inhibitor synthesis in tomato. Potato and pepper were as active as tomato systemin, whereas nightshade systemin was 10fold less active. Tomato systemin is inactive in inducing proteinase inhibitor synthesis in leaves of tobacco, although a semi-puri¢ed polypeptide fraction from tobacco leaves is highly active in inducing proteinase inhibitor synthesis in tobacco leaves (G. Pearce, C.A. Ryan, unpublished). This suggests that either the systemin gene has changed signi¢cantly during evolution, or that other polypeptides may be present in tobacco that serve a similar signaling role as systemin in response to wounding. This suggests that if polypeptides are defense-signaling molecules throughout the plant kingdom, they may have minimally conserved structures or diverse structures that may require individual isolation and characterization. Fig. 3. SDS^PAGE of the soluble proteins extracted from leaves of wild-type (WT) and transgenic tomato plants constitutively expressing the prosystemin gene. Arrows indicate proteins that constitutively accumulate in the transgenic plants. These same proteins accumulate systemically in wild-type plants in response to wounding.
temin is active, or if a processed form interacts with the systemin receptor. Proteinases have been identi¢ed in apoplast £uids that can degrade prosystemin to small polypeptides that retain immunoreactivity against systemin antibodies [27]. Thus, it is likely that systemin is processed from prosystemin when supplied to plants via the transpiration stream. Additionally, recombinant prosystemin analogs were synthesized in E. coli in which either the prosystemin sequence was mutated at residue 195, corresponding Thr17 to Ala17 of systemin, or else the entire systemin sequence was deleted. The Ala17 analog exhibited less than 1% of the proteinase inhibitor inducing ac-
3. The systemin signaling pathway An updated original model [31] for the activation of defense genes by systemin in tomato plants is shown in Fig. 4. Using 125 I-labeled systemin, a systemin-binding protein has recently been identi¢ed in the plasma membranes of leaf cells [32] and, more recently, on the surface of Lycopersicon peruvianum suspension cultured cells [19]. The properties of the binding protein indicate that it is the receptor that is involved in systemin-mediated signal transduction. The dissociation constant of the systemin^receptor interaction with the suspension cultured cell receptor was 0.17 nM and is in the range of Kd s found for polypeptide^receptor interactions in animal systems. The receptor exhibited a speci¢c and reversible interaction with systemin, and the binding was shown to increase several fold in response to MJ, suggesting
BBAPRO 36067 23-2-00
116
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 4. A current model for the systemic signaling pathway for defensive genes in tomato plants that are activated by herbivore attacks (wounding). The interaction of systemin with its membrane receptor initiates intracellular events that activate a PLA2 . The phospholipase releases LA from membranes, the production of JA, and the activation of defensive genes.
that the receptor is a systemic wound response protein. An Mr of 160 000 was determined for the receptor by photoa¤nity labeling and electrophoretic analysis [19]. The isolation of the receptor protein and its cDNA should provide a new avenue of research to seek interactive signaling components to further elucidate the early events of signaling. The interaction of systemin with its receptor regulates a complex cascade of intracellular event s that are all orchestrated to activate a PLA2 to release linolenic acid from membranes. These events include a depolarization of the plasma membrane [33], the opening of ion channels [32^34], an increase in the concentration of intracellular Ca2 [34], the inactivation of a plasma membrane proton ATPase [35], the activation of a MAP kinase [36,37], the synthesis of calmodulin, and the activation of a PLA2 [38,39]. Release of linolenic acid from membranes leads to its conversion to the oxylipins OPDA and JA that regulate defense genes. JA, together with ethylene has been shown [40] to activate transcription of the defense-related genes by a mechanism that is still
obscure. Investigations of the individual components of this pathway are important not only for the understanding of the intracellular events leading to plant defense gene activation, but of events leading to stress and developmental responses that are regulated in plants by JA [18,41^44]. The identi¢cation of a systemin receptor on the cell surface raises important questions regarding the transport of systemin throughout the plants and its delivery to the surface of the distal cells. Although systemin was found to be mobile in the phloem, its long-distance transport following wounding is not well understood. For systemin to be regulating systemic activation of defense genes, it must be either transported throughout the plants or somehow involved in a cascade that operates over long distances. In any event, systemin appears to exert its e¡ects through an external cell surface receptor. If systemin is transported in the phloem then it is likely that it is transported to companion cells and phloem parenchyma where it could be delivered to the apoplast, presumably by a speci¢c transporter. On the other hand, when systemin is initially released by wounding it may travel through the apoplast to nearby vascular bundle cells where it can interact with receptors to activate the synthesis in the cells of both prosystemin and its processing enzymes. This in turn would produce more systemin, which could move through the phloem and apoplast to distal cells to continue a cascade through the plant. Another possibility is that the initially released systemin might activate synthesis of prosystemin and a prosystemin transporter that would deliver nascent prosystemin to the apoplast where processing and receptor binding would both occur. In this scenario, the transport of systemin through the plant would occur by apoplastic transport and phloem loading to e¡ect transport to more distal cells, cascading through the plant. This latter scenario would require systemin to be transported from the phloem to the apoplast to interact with the receptor. Both scenarios ¢t several known facts, including the timing of movement of the wound signal (systemin) through the petioles (from 15 to 90 min), the tissue speci¢city of prosystemin synthesis in the vascular bundles (see Section 4), the movement of the wound signal from prosystemin transgenic rootstalks through grafts to wildtype scions in the absence of wounding, and the ob-
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
servations that the addition of systemin or prosystemin to wound sites on leaves of prosystemin antisense tomato plants caused distal leaves to activate defense genes [27]. Other primary signals have been proposed for activating defense genes in response to herbivore attacks, including electrical pulses, changes in hydraulic pressure, and the plant hormones ABA and ethylene [45^48]. Electrical phenomena accompany wounding but no evidence has been forthcoming to de¢ne how weak electrical signals can play a role in systemic signal transduction. The propagation of both electrical and hydraulic pulses are very fast compared to the known velocity of the movement of the wound signal in plants [49]. Additionally, girdling petioles of young tomato plants with a stream of hot air that kills all cells except the xylem elements prevented systemic signaling of proteinase inhibitor I in upper, ungirdled leaves in response to wounding [49]. Under these conditions, hydraulic signals would have been able to move through the xylem to unwounded leaves. Malone and Alorcan [50] reported that wounding girdled tomato leaves induced inhibitory activity against elastase in upper, unwounded leaves. It is possible that an entirely di¡erent signaling pathway is activating elastase inhibitors, but these latter experiments did not give consistent results, and speci¢c inhibitor quanti¢cation was not recorded, so that basal levels of inhibitor at the onset of experiments could not be assessed. There is no doubt that electrical and hydraulic waves do propagate throughout the plants in response to wounding, but their e¡ects on plant defense activation still remains obscure. In this regard, in the grafted plants mentioned earlier in which `sense' prosystemin transgenic rootstalks caused the induction of proteinase inhibitors in the wild-type scions in the absence of wounding, no generation of electrical or hydraulic signals could have occurred. In wild-type plants, breaching the vascular system has been proposed to perturb membranes throughout the plants [37], so that a priming of the signal transduction system through a hydraulic event might e¡ect an early transient activation of the MAP kinases and the PLA2 that are involved in the early signaling events, similar to a `touch response'. Transient signaling in response to hydraulic signals may be part of an early alarm system to activate signaling
117
pathway components, but weak, transient, electrical, or hydraulic signals alone cannot sustain the activation of defense genes over a period of several hours. ABA appears to play an overall role in determining whether the plant can mount a defense response [51,52]. Tomato plants with a mutation in ABA biosynthesis were found to be de¢cient in the wound response, and the response could be restored by exogenously ABA applied to leaves [51]. ABA does not behave as if it is a primary component of the signaling pathway since it does not act as an inducer of proteinase inhibitor synthesis in tomato plants, as do wounding, systemin, or other elicitors [52], but it is required for the plants to respond maximally [51]. Other plant hormones are also important to the defense response, including auxin, which inhibits the response [53], and ethylene [40] which is a component of the signaling pathway. 4. Di¡erential signaling by systemin ^ a modi¢ed model In comparing the timing of the wound-inducible expression of proteinase inhibitor genes with those of the signal transduction pathway, it became clear that two classes of wound- and systemin-inducible genes were di¡erentially regulated. As shown in Fig. 5, mRNAs coding for proteinase proteins mRNAs appeared much later than mRNAs coding for signal pathway proteins. The mRNAs encoding proteinase inhibitors I and II [11], CPI ([54]; M. Diaz, C.A. Ryan, in preparation), and an aspartic proteinase inhibitor, CDI [4,7], were ¢rst detectable in northern analyses at about 2 h after the plants were wounded, and the levels continued to increase through the next 8 h. On the other hand, mRNAs encoding wound-inducible LOX [15], CaM [13], AOS (G. Howe, C.A. Ryan, unpublished), and prosystemin [12] were ¢rst detectable within about 30 min, and were maximally induced 2^3 h after wounding. This was 5^6 h before the proteinase inhibitor mRNAs reached maximal levels. AOS mRNA has been reported to be wound inducible in Arabidopsis leaves [55] with induction kinetics similar to LOX mRNAs in tomato leaves, beginning at 0.5 h after wounding and maximizing at about 2^4 h, then declining thereafter.
BBAPRO 36067 23-2-00
118
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 5. The di¡erential induction of mRNAs of systemic wound response proteins. Young tomato plants were wounded on the lower leaves at time zero and assayed by Northern blotting at the times indicated. Inh I, proteinase inhibitor I; Inh II, proteinase inhibitor II; CDI, cathepsin D inhibitor (and aspartic proteinase inhibitor); CPI, metallocarboxypeptidase inhibitor; CaM, calmodulin; LOX, lipoxygenase; ProSys, prosystemin.
Ultrastructural studies of the intracellular localization of proteinase inhibitor I, synthesized in response to wounding, had revealed previously that the inhibitor is sequestered in the central vacuoles of mesophyll cells of tomato leaves [56]). Inhibitor II has also been found in leaf vacuoles [57,58], and it is likely that other proteinase inhibitors with long half lives are also sequestered in vacuoles of mesophyll and/or palisade cells. In contrast, when the promoter region of the wound-inducible prosystemin gene was fused with the GUS reporter gene, and tomato plants were transformed with the fused gene, the plants responded to wounding and MeJA by synthesizing the GUS protein [59]. The localization of GUS activity was visualized by supplying a colorimetric substrate to freehand sections of leaf tissues. The wound-inducible GUS activity was only found in the vascular bundles of the major and minor veins of the leaves. This was con¢rmed by tissue printing
using speci¢c antibodies prepared against prosystemin and visualized before and after wounding using a colorimetric substrate. The AOS gene in Arabidopsis [55] that is a component of the defense signaling pathway is also speci¢cally expressed in the vascular bundles in response to wounding (E. Weiler, personal communication). We hypothesize that the signal transduction genes as a group are `early genes' that are expressed in the vascular bundle cells, while the defensive proteinase inhibitor genes are `late genes' that are expressed in palisade and spongy mesophyll cells (Fig. 6). The di¡erences in timing and localization of these two functionally di¡erent groups of proteins suggests a scenario in which the signal transduction pathway is initially activated in the vascular bundles to amplify the production of second messengers, which could include OPDA and/or JA, derivatives of these molecules [60], or other signaling molecules that are transported to other cell types. The second messengers are then transported to the palisade and mesophyll cells where the defense proteins are synthesized and compartmented. A comparison of the relative time courses of the various components of the signal transduction pathway from the initial release of systemin to the production of proteinase inhibitors is shown in Fig. 7. This scenario di¡ers signi¢cantly from a recently proposed signaling model [48] in which systemin was suggested to be a signal released in response to jasmonic acid. In this latter model, jasmonic acid and ethylene were not considered downstream systemic signals because increases in ethylene and jasmonic acid cannot be detected levels in unwounded leaves of wounded plants [48]. The author did not consider that JA is likely transient and only present at very
Fig. 6. A modi¢ed model to indicate the di¡erential signaling of signal pathway components and the proteinase inhibitors in di¡erent cell types.
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
Fig. 7. A time-course representation of the progression of the production of signal pathway components in tomato leaves in response to wounding at time 0. H, a transient hydraulic pulse due to the breaching of the vascular system by wounding, which initiates the signaling events.
low concentrations in speci¢c cell types. There is no evidence available to bear on the minimum quantities of these compounds that can activate genes, or whether OPDA or JA is the signaling component. Several examples are known of JA-inducible genes being activated without measurable increases in JA. As stated by Weiler et al. [61], `clearly, results based solely on analysis of JA can no longer be regarded as complete'. Bowles also concluded [48], based on unpublished data, that Ala17-systemin, the potent antagonist of systemin mentioned earlier, inhibits the action of OGA elicitors. In contrast, through extensive experimentation (G. Pearce, C.A. Ryan, unpublished), we have shown conclusively that Ala17-systemin has no e¡ect on the proteinase inhibitor inducing activity of either OGA or chitosan, and therefore systemin must be perceived independently from these elicitors, likely through di¡erent receptors. A recent report has demonstrated that chitosan, OGA, and systemin all activate PLA2 activity in tomato plants [39], and all three elicitors act via the octadecanoid pathway [62] although it is not yet known if the same PLA2 is involved with signaling by all three elicitors. While the downstream message that activates pro-
119
teinase inhibitor synthesis in response to wounding and systemin may be JA or a derivative in conjunction with ethylene, new evidence from our research suggests that other candidates for second messengers are produced in response to wounding and systemin that could be part of the signaling pathway for proteinase inhibitor synthesis. The candidates are (i) the OGA derived from the cell walls of vascular bundle cells by the action of a newly discovered wound-, systemin- and JA-inducible PG [6], and (ii) reactive oxygen species [22], known to be produced in tomato leaves by OGAs [63], through NADPH oxidase [63^ 65]. A speci¢c inhibitor of NADPH oxidase, diphenylene iodonium chloride, blocks wound-inducible H2 O2 generation in tomato leaves [22], and we have recently shown that this inhibitor powerfully blocks the synthesis of inhibitor I and II protein accumulation in response to wounding, systemin, and MeJA (M. Orozco-Cardenas, C.A. Ryan, in preparation), although whether the inhibitor is speci¢c for NADPH oxidase in vivo is not known. If OGA and H2 O2 are indeed second messengers for the activation of proteinase inhibitor genes in mesophyll cells, they are likely downstream in the signaling pathway since they are inducible by JA and PG in the vascular bundles. The production of OGA and H2 O2 as second messengers is kinetically compatible with the synthesis of proteinase inhibitors since both are produced later than the signal transduction mRNAs ([22]; see Fig. 5), but concurrent with the synthesis of proteinase inhibitor mRNAs (i.e., between 2 and 8 h after wounding). Reactive oxygen plays a central role in the resistance response of plants against pathogens leading to incompatibility and the development of systemic acquired resistances [66]. The increase in active oxygen during herbivore attacks has been suggested to act as a deterrent to herbivores [66], and the wound-inducible reactive oxygen might potentiate the defense response against pathogens as well. While there is no evidence to date to link wound-inducible H2 O2 in tomato leaves to pathogen resistances, the octadecanoid pathway is important to phytoalexin production in several species of plants [67,68] and the pathway has been shown through mutant analyses to be a factor in pathogen resistance in Arabidopsis [69^72]. In a broader context, the defense signaling pathways in plants contain common types of components
BBAPRO 36067 23-2-00
120
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121
that have contributed to the evolution of information processing systems for the activation of defenses against both herbivores and pathogens in particular ecological niches. For example, receptors have evolved in plants that can recognize signals unique to speci¢c pests or pathogens [73^76]. These signals regulate the tissue-speci¢c and/or cell-speci¢c expression of a variety of defensive genes in di¡erent plants to protect against speci¢c herbivores and/or pathogens. Herbivore saliva can be a source of signals to activate defense genes [77], or to signal the production of chemicals to attract predators of insects chewing on the plants [78,79]. While the various signals and genes produced by herbivore and pathogen attacks are likely to di¡er among di¡erent plant species, it is also likely that some other signal pathway components, e.g., ion channels, MAP kinases, and transcription factors, will have evolved with some conservation of structures and functions. Perhaps with time these fundamental components may track back to ancestral genes common to plant and animal defense systems. Knowledge of the biochemical details of the evolution of defense signaling pathways of plants could have an enormous potential in helping to understand the biological characteristics of species that are found in various ecosystems, as well as in applying this knowledge to the agricultural community to increase plant productivity using natural defense strategies Acknowledgements Supported in part by Washington State University College of Agriculture and Home Economics, Project 1791 and Grants DCB 9104542 and DCB 9117795 from the National Science Foundation, and Grant #9801502 from the U.S. Department of Agriculture.
References [1] R. Karban, I.T. Baldwin (Eds.), Induced Responses to Herbivory, University of Chicago Press, Chicago, 1997. [2] Herbivores: Their Interaction with Secondary Plant Metabolites. (1979) G.A. Rosenthal and D.H. Janzen (Eds.), Academic Press, NY. [3] C.A. Ryan, Annu. Rev. Phytopathol. 28 (1990) 425^449.
[4] D. Bergey, G. Howe, C.A. Ryan, Proc. Natl. Acad. Sci. USA 93 (1996) 12053^12058. [5] C.P. Constabel, D.R. Bergey, C.A. Ryan, in: J. Romeo et al. (Eds.), Phytochemical Diversity and Redundancy in Ecological Interactions, Plenum Press, New York, 1996, pp. 231^ 251. [6] D.R. Bergey, M. Orozco-Cardenas, D. Moura, C.A. Ryan, Proc. Natl. Acad. Sci. USA 96 (1999) 1756^1760. [7] T. Hildemann, M. Ebneth, H. Pen¬a-Cortes, J.J. SanchezSerrano, L. Wilmitzer, S. Prat, Plant Cell 4 (1992) 1157^ 1170. [8] V. Pautoªt, F.M. Holzer, B. Reisch, L.L. Walling, Proc. Natl. Acad Sci. USA 90 (1993) 9906^9910. [9] C.J. Bolter, Plant Physiol. 103 (1993) 1347^1353. [10] C.A. Ryan, G. Pearce, Annu. Rev. Cell Dev. Biol. 14 (1998) 1^17. [11] J.S. Graham, G. Hall, G. Pearce, C.A. Ryan, Planta 169 (1986) 399^405. [12] B. McGurl, G. Pearce, M. Orozco-Cardenas, C.A. Ryan, Science 255 (1992) 1570^1573. [13] D.R. Bergey, C.A. Ryan, Plant Mol. Biol. 40 (1999) 815^ 823. [14] J. Royo, G. Vancanney, A.G. Perez, C. Sanz, K. Stormann, S. Rosahl, J.J. Sanchez-Serrano, J. Biol. Chem. 271 (1996) 21012^21019. [15] T. Heitz, D. Bergey, C.A. Ryan, Plant Physiol. 114 (1997) 1805^1903. [16] B. Vick, D.C. Zimmerman, Plant Physiol. 75 (1984) 458^461. [17] E. Blee, J. Joyard, Plant Physiol. 110 (1996) 445^454. [18] M. Hamberg, H.W. Gardner, Biochim. Biophys. Acta 1165 (1992) 1^18. [19] J.M. Scheer, C.A. Ryan, Plant Cell 11 (1999) 1525^1535. [20] G. Felix, T. Boller, Plant J. 7 (1995) 381^389. [21] M.J. Stennis, C. Sreeganga, C.A. Ryan, P.S. Low, Plant Physiol. 117 (1998) 1031^1036. [22] M. Orozco-Cardenas, C.A. Ryan, Proc. Natl. Acad. Sci. USA 96 (1999) 6553^7667. [23] G. Pearce, D. Strydom, S. Johnson, C.A. Ryan, Science 253 (1991) 895^897. [24] G. Pearce, S. Johnson, C.A. Ryan, J. Biol. Chem. 268 (1993) 212^216. [25] J. Narva©ez-Vasque©z, G. Pearce, M.L. Orozco-Cardenas, V.R. Franceschi, C.A. Ryan, Planta 195 (1995) 593^600. [26] M. Orozco-Cardenas, B. McGurl, C.A. Ryan, Proc. Natl. Acad. Sci. USA 90 (1993) 8273^8276. [27] J.E. Dombrowski, G. Pearce, C.A. Ryan, Proc. Natl. Acad. Sci. USA 96 (1999) 12947^12952. [28] B. McGurl, M. Orozco-Cardenas, G. Pearce, C.A. Ryan, Proc. Natl. Acad. Sci. USA 91 (1994) 9799^9802. [29] J. Delano-Frier, J.E. Dombrowski, C.A. Ryan, Protein Exp. Purif. 17 (1999) 74^82. [30] C.P. Constabel, L. Yip, C.A. Ryan, Plant Mol. Biol. 36 (1998) 55^62. [31] E.E. Farmer, C.A. Ryan, Plant Cell 4 (1992) 129^134. [32] T. Meindl, T. Boller, G. Felix, Plant Cell 10 (1998) 1561^ 1570.
BBAPRO 36067 23-2-00
C.A. Ryan / Biochimica et Biophysica Acta 1477 (2000) 112^121 [33] C. Moyen, E. Johannes, Plant Cell Environ. 19 (1996) 464^ 470. [34] C. Moyen, K.E. Hammond-Kosack, J. Jones, M.R. Knight, E. Johannes, Plant Cell Environ. 21 (1998) 1101^1111. [35] A. Schaller, C. Oecking, Plant Cell 11 (1999) 263^272. [36] S. Usami, H. Banno, Y. Ito, R. Nishihama, Y. Machida, Proc. Natl. Acad. Sci. USA 92 (1995) 8660^8664. [37] J.W. Stratmann, C.A. Ryan, Proc. Natl. Acad. Sci. USA 94 (1997) 11085^11089. [38] S. Lee, S. Suh, S. Kim, R.C. Crain, J.M. Kwak, H.-G. Nam, Y. Lee, Plant J. 12 (1997) 547^556. [39] J. Narva¨ez-Vasquez, J. Florin-Christensen, C.A. Ryan, The Plant Cell 11 (1999) 1^13. [40] P.J. O'Donnell, C.M. Calvert, R. Atzorn, C. Wasternack, H.M.O. Leyser, D.J. Bowles, Science 274 (1998) 1914^1917. [41] E.W. Weiler, Naturwissenschaften 84 (1997) 340^349. [42] T. Farmer, Plant Mol. Biol. 26 (1994) 1423^1437. [43] R.A. Creelman, J.E. Mullet, Annu. Rev. Plant Physiol. Plant Biol. 48 (1997) 355^381. [44] S. Seo, H. Sano, O. Ohashi, Physiol. Plant. 101 (1997) 740^ 745. [45] D.C. Wildon, J.F. Thain, P.E.H. Michin, I. Gubb, A. Reilly, Y. Skipper, H. Doherty, P. O'Donnell, D.J. Bowles, Nature 360 (1993) 62^65. [46] O. Herde, H. Pen¬a Cortes, C. Wasternack, L. Willmitzer, J. Fisahn, Plant Physiol. 119 (1999) 213^218. [47] M. Malone, L. Palumbo, F. Boari, M. Monteleone, H.G. Jones, Plant Cell Environ. 17 (1994) 81^87. [48] D. Bowles, Phil. Trans. R. Soc. Lond. B 353 (1998) 1495^ 1510. [49] C. Nelson, M. Walker-Simmons, D. Makus, G. Zuroske, J. Graham, C.A. Ryan, in: P. Hedin (Ed.), Mechanisms of Plant Resistance to Insects, ACS Monograph, 1983, pp. 103^122. [50] M. Malone, J.-J. Alarcon, Planta 196 (1995) 740^746. [51] O. Herde, R. Atzorn, J. Fisahn, C. Wasternack, L. Willmitzer, H. Pena-Cortes, Plant Physiol. 112 (1996) 853^860. [52] G.F. Birkenmeier, C.A. Ryan, Plant Physiol. 117 (1998) 687^693. [53] A. Kernan, R.W. Thornburg, Plant Physiol. 91 (1989) 73^78. [54] B. Martineau, K.E. McBride, C.M. Mouck, Mol. Gen. Genet. 228 (1991) 281^286. [55] D.L. Laudert, U. Pfannschmidt, F. Lottspeich, H. Hollander-Czytko, E.W. Weiler, Plant Mol. Biol. 31 (1996) 323^335.
121
[56] L.K. Shumway, J.M. Rancour, C.A. Ryan, Planta 93 (1970) 1^14. [57] M.K. Walker-Simmons, C.A. Ryan, Plant Physiol. 60 (1977) 61^63. [58] G.-Y. Jauh, A.M. Fischer, H.D. Grimes, C.A. Ryan, J.C. Rogers, Proc. Natl. Acad. Sci. USA 95 (1998) 12995^12999. [59] T. Jacinto, B. McGurl, V. Franceschi, J. Delano-Freier, C.A. Ryan, Planta 203 (1997) 406^411. [60] S. Parchmann, H. Gundlach, M.J. Mueller, Plant Physiol. 115 (1997) 1057^1064. [61] E.W. Weiler, Recent Adv. Phytochem. 32 (1998) 179^1205. [62] S. Doares, T. Syrovets, E.W. Weiler, C.A. Ryan, Proc. Natl. Acad. Sci. USA 92 (1995) 4095^4098. [63] M.J. Stennis, C. Sreeganga, C.A. Ryan, P.S. Low, Plant Physiol. 117 (1998) 1031^1036. [64] T. Keller, H.G. Damude, D. Werner, P. Doener, R.A. Dixon, C. Lamb, Plant Cell 10 (1998) 255^266. [65] C. Lamb, R.A. Dixon, Annu. Rev. Plant Physiol. Plant Mol. Biol. 48 (1997) 251^275. [66] J.L. Bi, G.W. Felton, J. Chem. Ecol. 21 (1995) 1511^1530. [67] H. Gundlach, M.J. Muller, T.M. Kutchan, M.H. Zenk, Proc. Natl. Acad. Sci. USA 89 (1992) 2389^2393. [68] M.J. Mueller, W. Brodschelm, E. Spannagl, M.H. Zenk, Proc. Natl. Acad. Sci. USA 90 (1993) 7490^7494. [69] V. Perumal, J. Shockey, C.A. Levesque, R.J. Cook, J. Browse, Proc. Natl. Acad. Sci. USA 95 (1998) 7209^7214. [70] P.E. Staswick, G.Y. Yuen, C.C. Lehman, Plant J. 15 (1998) 747^754. [71] B.P.H.J. Thomma, K. Eggermont, I.A.M.A. Penninckz, B. Mauchmani, R. Vogelsang, P.P.A. Cammue, W.F. Broekaert, Proc. Natl. Acad. Sci. USA 95 (1998) 15107^15111. [72] J.S. Thaler, Nature 399 (1999) 686^688. [73] D. Nennstiel, D. Scheel, T. Nurnberger, FEBS Lett. 431 (1998) 405^410. [74] N. Umemoto, M. Kakitani, A. Iwamatsu, M. Yoshikawa, N. Yamaoka, I. Ishida, Proc. Natl. Acad. Sci. USA 94 (1997) 1029^1034. [75] J. Ebel, BioEssays 20 (1998) 569^576. [76] A. Fath, T. Boller, Plant Physiol. 112 (1996) 1659^1668. [77] K.L. Korth, R.A. Dixon, Plant Physiol. 115 (1997) 1299^ 1305. [78] C.M. De Moraes, W.J. Lewis, P.W. Pare, H.T. Abora, J.H. Tumlinson, Nature 393 (1998) 570^573. [79] L. Mattiacci, M. Dicke, M.A. Posthumus, Proc. Natl. Acad. Sci. USA 92 (1995) 2036^2040.
BBAPRO 36067 23-2-00