This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 18/11/3303    most recent tpc.106.041103v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via ISI Web of Science (29) Citing Articles via Google Scholar Google Scholar Articles by Kang, J.-H. Articles by Baldwin, I. T. Search for Related Content PubMed PubMed Citation Articles by Kang, J.-H. Articles by Baldwin, I. T. Agricola Articles by Kang, J.-H. Articles by Baldwin, I. T.

Silencing Threonine Deaminase and JAR4 in Nicotiana attenuata Impairs Jasmonic Acid–Isoleucine–Mediated Defenses against Manduca sexta^[W]

Department of Molecular Ecology, Max-Planck-Institute of Chemical Ecology, D-07745 Jena, Germany

¹ To whom correspondence should be addressed. E-mail baldwin{at}ice.mpg.de; fax 49-3641-571102.

	ABSTRACT

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Threonine deaminase (TD) catalyzes the conversion of Thr to -keto butyrate in Ile biosynthesis; however, its dramatic upregulation in leaves after herbivore attack suggests a role in defense. In Nicotiana attenuata, strongly silenced TD transgenic plants were stunted, whereas mildly silenced TD transgenic plants had normal growth but were highly susceptible to Manduca sexta attack. The herbivore susceptibility was associated with the reduced levels of jasmonic acid–isoleucine (JA-Ile), trypsin proteinase inhibitors, and nicotine. Adding [¹³C₄]Thr to wounds treated with oral secretions revealed that TD supplies Ile for JA-Ile synthesis. Applying Ile or JA-Ile to the wounds of TD-silenced plants restored herbivore resistance. Silencing JASMONATE-RESISTANT4 (JAR4), the N. attenuata homolog of the JA-Ile–conjugating enzyme JAR1, by virus-induced gene silencing confirmed that JA-Ile plays important roles in activating plant defenses. TD may also function in the insect gut as an antinutritive defense protein, decreasing the availability of Thr, because continuous supplementation of TD-silenced plants with large amounts (2 mmol) of Thr, but not Ile, increased M. sexta growth. However, the fact that the herbivore resistance of both TD- and JAR-silenced plants was completely restored by signal quantities (0.6 µmol) of JA-Ile treatment suggests that TD's defensive role can be attributed more to signaling than to antinutritive defense.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Threonine deaminase (TD) catalyzes the formation of -keto butyrate (-KB) from Thr, the first step in the biosynthesis of Ile. Regulation of TD activity by Ile was the first recognized instance of allosteric feedback regulation by the end product of a biosynthetic pathway (Umbarger, 1956). The function of TD for Ile biosynthesis was demonstrated by analyzing the Ile auxotrophic mutant in Nicotiana plumbaginifolia, which has no detectable TD activity (Sidorov et al., 1981). When this mutant was transformed with the Saccharomyces cerevisiae ILV gene that encodes TD, the transformed lines could be grown on medium without Ile (Colau et al., 1987). These results demonstrate that TD regulates Ile production and is indispensable for plant growth. However, TD's unusual expression pattern in solanaceous plants suggests that TD plays additional roles in development and herbivore defense.

For more than a decade, TD has been recognized as a reliable marker for wounding and jasmonic acid (JA) elicitation in potato (Solanum tuberosum) and tomato (Solanum lycopersicum) (Hildmann et al., 1992; Samach et al., 1995; Dammann et al., 1997). Wound-induced TD expression is mediated by abscisic acid and JA signaling in tomato plants (Hildmann et al., 1992), and in potato, protein phosphorylation is required for TD elicitation by JA. TD is also highly expressed in flowers and has a chloroplast transit peptide in the N-terminal region (Samach et al., 1991, 1995). A strong association between JA signaling and TD expression can be inferred from the synthesis of JA–amino acid conjugates and suggests a mechanism linking TD activity and herbivore resistance.

JA synthesis begins in plastids. There, -linolenic acid is oxygenated by lipoxygenase (LOX); converted to 12-oxo-phytodienoic acid by allene oxide synthase and allene oxide cyclase before being exported to the peroxisome; and reduced by 12-oxo-phytodienoic acid reductase. JA is produced after three consecutive ß-oxidation steps in the peroxisome (Li et al., 2005). JA can be subsequently methylated to its volatile counterpart, methyl jasmonate (MeJA), or conjugated with various sugars and amino acids (Sembdner and Parthier, 1993; Sembdner et al., 1994). An Arabidopsis thaliana gene (JASMONATE-RESISTANT1 [JAR1]) involved in JA responsiveness was shown to adenylate JA before its conjugation with amino acids, of which the JA–isoleucine conjugate (JA-Ile) was the most abundant (Staswick et al., 2002; Staswick and Tiryaki, 2004). Because JA signaling is essential for resistance to a large number of herbivore taxa (Halitschke and Baldwin, 2003), TD might supply Ile for conjugation with JA at the attack site and thereby function in defense signaling. Another hypothesis for a defensive role for TD is based on a recent proteomic analysis of the midgut contents of Manduca sexta larvae that fed on tomato (Chen et al., 2005). This exciting report revealed that one of the abundant proteins in the larval midgut was TD, but TD that lacked a regulatory domain. This truncated TD might efficiently degrade Thr without being inhibited by Ile and function as an antinutritive defense by limiting the supply of Thr needed for herbivore growth (Chen et al., 2005).

Nicotiana attenuata is a particularly useful system in which to study herbivore resistance responses. Not only is it well established that JA signaling mediates herbivore resistance in the field for this species (Baldwin, 1998; Kessler and Baldwin, 2001; Kessler et al., 2004), but also the direct and indirect defense traits with which JA signaling influences herbivore resistance are known (Halitschke et al., 2004; Steppuhn et al., 2004). The responses of N. attenuata to one particular herbivore, the solanaceous specialist M. sexta, are particularly well understood. The attacked plant reorganizes its wound response when eight fatty acid–amino acid conjugates, present in the herbivore's oral secretions (OS), are introduced into plant wounds during feeding. The reorganization begins with a dramatic JA burst in the attacked leaves (Schittko et al., 2000), which alters the expression of numerous genes and the accumulation and release of secondary metabolites (Halitschke et al., 2000, 2001, 2003; Kahl et al., 2000; Roda et al., 2004). Silencing the expression of the specific lox that supplies the fatty acid hydroperoxides for JA biosynthesis in N. attenuata (LOX3) reduces the OS-elicited JA burst and all associated changes in the plant's resistance traits (Halitschke and Baldwin, 2003; Kessler et al., 2004).

Many of N. attenuata's herbivore-responsive genes have been identified by cDNA differential display, subtractive hybridization, and cDNA-amplified fragment-length polymorphism display (Halitschke et al., 2001, 2003; Hermsmeier et al., 2001; Schittko et al., 2001; Hui et al., 2003; Voelckel and Baldwin, 2003). These genes have been spotted onto microarrays, and their expression behavior has been analyzed in response to various environmental stresses (Halitschke et al., 2003; Hui et al., 2003; Izaguirre et al., 2003; Voelckel and Baldwin, 2003, 2004; Lou and Baldwin, 2004). In these experiments, TD expression was consistently found to correlate with elicited herbivore resistance. TD was cloned by differential display RT-PCR, found to be encoded by a single gene, and strongly elicited when plants were attacked by M. sexta larvae, mechanically wounded, or treated with MeJA; neither Tobacco mosaic virus nor treatment with Agrobacterium tumefaciens infection, ethylene, or methyl salicylate elicited TD expression (Hermsmeier et al., 2001). Wounding and OS elicitation increase TD expression, not only in the wounded leaf but also in distal nonwounded leaves that are phyllotactically connected by common orthostichies (Schittko et al., 2001). The wound-induced expression of TD is reduced in N. attenuata plants transformed with N. attenuata LOX3 in an antisense orientation, demonstrating that TD elicitation requires JA signaling (Halitschke and Baldwin, 2003). These observations suggest that TD may be involved in defense against herbivore attack.

To examine the effect of TD on defense responses, we first expressed 1.3 kb of the N. attenuata TD in an antisense orientation. Transformed lines were readily characterized as having one of two growth phenotypes: (1) plants with severely reduced TD expression and activity, and stunted growth and development (asTDS plants), and (2) plants with mildly reduced TD expression and activity but otherwise wild-type growth and development patterns (asTDM plants). Because plant–herbivore interactions are difficult to interpret in plants that are severely stunted in their growth and development, we also silenced TD with a virus-induced gene-silencing (VIGS) system optimized for N. attenuata (Saedler and Baldwin, 2004), which allowed us to silence TD in wild-type plants. Finally, we cloned JAR4, the Arabidopsis JAR1 homolog in N. attenuata, to demonstrate that JAR4 conjugates Ile to JA to mediate defense signaling and resistance to M. sexta larvae. We also tested the hypothesis that TD functions as an antinutritive defense by adding Thr and Ile to wild-type and TD-silenced plants and examined the consequences of this supplementation for larval growth. The results of this work support both hypotheses: TD plays an important role in herbivore resistance by mediating JA-Ile signaling and also acts as an antinutritional protein by depleting Thr levels.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
We measured TD transcript accumulation in wild-type plants after elicitation by insect attack and MeJA treatment. TD mRNA in wild-type plants is strongly increased after attack from M. sexta larvae (62-fold); TD transcripts are also strongly elicited when leaves are wounded and treated with M. sexta OS (16-fold) or when MeJA is applied in a lanolin paste to intact leaves (79-fold) (see Supplemental Figure 1 online). Attack from other leaf-chewing insect herbivores (Heliothis virescens and Spodoptera exigua) as well as from a species that feeds by lacerating and flushing cells (Tupiocoris notatus) also strongly elicits TD transcript accumulation (18- to 41-fold; see Supplemental Figure 1 online), suggesting that TD is involved in plant defense.

To examine the function of TD, we first produced transgenic plants expressing TD in an antisense orientation. T2 homozygous plants from independently transformed lines, each harboring a single copy of the transgene, as verified by segregation analysis for antibiotic resistance and DNA gel blot analysis (see Supplemental Figure 2 online), were analyzed. Transformed lines were readily characterized as having one of two growth phenotypes: (1) plants with greatly reduced TD expression and activity and retarded growth (asTDS plants; Figure 1A ), and (2) plants with mildly reduced TD expression and activity but whose growth and development patterns were indistinguishable from those of wild-type plants (asTDM plants; Figure 1A). Second, we produced TD-silenced plants using the VIGS method. TD-silenced (TDVIGS) plants had less TD expression and activity but similar growth patterns compared with empty vector (EV) control VIGS plants (Figure 1A). All transgenic lines and VIGS plants were also analyzed for levels of defense-related secondary metabolites such as trypsin protease inhibitor (TPI) and nicotine.

View larger version (46K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 1. Suppressing TD in asTD and TDVIGS Plants. (A) Wild-type (55-d-old), asTDM2 (55-d-old), asTDS1 (85-d-old), EV (54-d-old), and TDVIGS (54-d-old) plants. Note that asTDM2 plants are morphologically indistinguishable from wild-type plants, but asTDS1 plants are severely stunted in their growth and morphologically different from wild-type plants. TDVIGS and EV plants grow similarly. (B) and (C) Accumulation of TD transcripts (B) and -KB concentration (C) in a pooled sample of four replicate node +1 leaves, which were treated with 20 µL of lanolin containing 150 µg of MeJA and harvested after 24 h from wild-type and three independently transformed T2 asTD plants (asTDS1, asTDM1, and asTDM2). The arrow indicates endogenous TD RNA (TD), and the arrowhead indicates antisense TD RNA (asTD). Ethidium bromide–stained 18S rRNA was used as a loading control. Asterisks represent significant differences between MeJA-treated wild-type and MeJA-treated asTD plants (unpaired t test: * P < 0.05; ** P < 0.01; **** P < 0.0001). (D) to (G) Accumulation of TD transcripts ([D] and [F]) and -KB concentration ([E] and [G]) in leaves at node +1 from three replicate wild-type, asTDS1, and asTDM2 plants, which were wounded with a fabric pattern wheel and immediately treated with 20 µL of deionized water (W) or 20 µL of OS. Asterisks represent significant differences between wild-type and asTD plants (two-way ANOVA, Fisher's PLSD: **** P < 0.0001). (H) and (I) Accumulation of TD transcripts (H) and -KB (I) in TDVIGS plants. Plants were inoculated with Agrobacterium harboring tobacco rattle virus (TRV) constructs that contain an EV or a 335-bp TD fragment (TDVIGS). Fourteen days after inoculation, leaves at node +1 from four to five replicate EV and TDVIGS plants were wounded with a fabric pattern wheel and immediately treated with 20 µL of deionized water (W) or 20 µL of OS. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; ** P < 0.01). The transcripts were analyzed by real-time PCR as means ± SE of three to five replicate leaves in arbitrary units from a calibration with 5x dilution series of cDNAs prepared from asTD plant RNA samples extracted 1 h after wounding ([D] and [F]) or from EV plant RNA samples extracted 1 h after wounding (H).

Silencing TD Transcripts Decreases -KB Accumulation and Impairs Herbivore Resistance without Influencing Plant Growth

Plants treated with MeJA in a lanolin paste are continuously elicited, as the MeJA slowly diffuses into the plant (Zhang et al., 1997). To examine the effects of a more subtle elicitation treatment, transgenic lines and VIGS plants were wounded with a fabric pattern wheel and treated with water or M. sexta OS. TD mRNA expression was analyzed by real-time PCR. When wounded leaves were treated with water, TD mRNA attained maximum values in wild-type plants 1.5 h after wounding and waned slowly thereafter. Levels in both asTDM2 and asTDS1 plants were significantly lower than the levels in wild-type plants (Figure 1D; Fisher's protected least squares difference [PLSD], P < 0.0001). The production of -KB was slightly increased by wounding. Although -KB levels in asTDM2 plants were reduced, they did not differ significantly compared with the levels in wild-type plants (Figure 1E; Fisher's PLSD, P = 0.243), but levels of -KB in asTDS1 plants were reduced significantly compared with the levels in wild-type plants (Figure 1E; Fisher's PLSD, P < 0.0001). Leaves treated with water or OS showed similar expression patterns. TD mRNA levels in leaves treated with M. sexta OS from both asTDM2 and asTDS1 plants were significantly lower than the levels in leaves from wild-type plants (Figure 1F; Fisher's PLSD, P < 0.0001). Levels of -KB in asTDM2 plants were reduced but did not differ significantly compared with the levels in wild-type plants (Figure 1G; Fisher's PLSD, P = 0.1969); however, levels of -KB in asTDS1 plants were reduced significantly compared with the levels in wild-type plants (Figure 1G; Fisher's PLSD, P < 0.0001). TDVIGS plants also showed reduced TD mRNA levels compared with EV control plants. When wounded leaves were treated with water or M. sexta OS and then compared with EV plants, the levels of TD mRNA and -KB in TDVIGS plants were 80 and 71% lower in transcripts and 48 and 47% lower in TD activity (Figures 1H and 1I; unpaired t test, P 0.012).

To determine whether the mild suppression of TD transcripts and activity observed in the asTDM lines influenced plant growth and competitive ability, we synchronized the germination and growth of the different lines and grew them individually in 2-liter pots or in competition with each other in 2-liter pots. We measured stalk elongation, which previous experiments have revealed to accurately measure competitive ability and relative fitness (Glawe et al., 2003). No differences in stalk elongation among the lines were observed when plants were grown singly or in competition with wild-type plants (see Supplemental Figure 3 online). When TDVIGS and EV plants were grown individually in 1-liter pots, stalk lengths appeared not to differ (Figure 1A).

To determine whether TD is involved in plant defense, we measured the performance of the insect herbivore M. sexta, which is responsible for the largest losses in leaf area among N. attenuata plants growing in nature (Baldwin, 1998). asTDS1 plants were severely stunted in their growth, and their leaf developmental traits differed from those of wild-type plants (Figure 1A), which confounded comparisons of herbivore performance between wild-type and asTDS1 plants. Therefore, we first compared herbivore performance on wild-type and the morphologically indistinguishable asTDM lines. Freshly eclosed M. sexta larvae placed on the source–sink transition leaf of each of seven replicate plants of each genotype gained significantly more mass on plants of both asTDM lines than they did on wild-type plants. By day 6, larvae on asTDM2 plants had almost doubled their mass compared with larvae on wild-type plants (see Supplemental Figure 4A online; repeated-measurement analysis of variance [ANOVA], F_2,36 = 15.988; P = 0.0001; PLSD 0.0485). Similarly, M. sexta larvae placed on VIGS leaves gained significantly more mass on TDVIGS plants than on EV plants (see Supplemental Figure 4B online; repeated-measurement ANOVA, F_1,21 = 12.071; P = 0.0023; PLSD = 0.0023). These results demonstrate that reductions in TD expression and activity do not influence plant growth (even under intense intraspecific competition [see Supplemental Figure 3 online]) but impair resistance to an adapted herbivore.

TD Silencing Impairs Elicited Direct Defenses in asTD and TDVIGS Plants
To test the hypothesis that increasing Ile pools at the wound site could be used for herbivore-elicited direct defenses, we measured TPI in TD-silenced plants. Compared with wounding alone, OS treatment of puncture wounds in wild-type plants resulted in a 2.4-fold increase in TPI activity (Figure 2A ). The wound-induced accumulation of TPI activity in asTDM2 plants was 31% lower than that in wild-type plants (Figure 2A; unpaired t test, P = 0.0911), and the OS-induced accumulation of TPI in asTDM2 plants was 41% lower than that in wild-type plants (Figure 2A; unpaired t test, P = 0.0386). The wound-induced accumulation of TPI activity in asTDS1 plants was 47% lower than that in wild-type plants (Figure 2A; unpaired t test, P = 0.0365), and the OS-induced accumulation of TPI in asTDS1 plants was 76% lower than that in wild-type plants (Figure 2A; unpaired t test, P = 0.002). However, OS treatment of wounds in asTDS1 plants did not significantly increase TPI activity compared with water treatment of wounds in asTDS1 plants (Figure 2A; unpaired t test, P = 0.719), suggesting that severe nutritional deficiencies inhibited TPI production in these plants. VIGS plants also showed induced TPI activity when leaves were wounded and treated with water or M. sexta OS (Figure 2B); however, when leaves were treated with M. sexta OS, TPI activity in TDVIGS plants was reduced significantly compared with that in EV plants (Figure 2B; unpaired t test, P = 0.0006). These results suggest that diminished TPI levels are a major cause of the increased performance of M. sexta larvae feeding on asTDM2 and TDVIGS plants. When Ile was added to water or OS before being applied to the puncture wounds, TPI levels in both asTDM2 and TDVIGS plants were restored to levels found in wild-type and EV plants. In asTDS1 plants, adding Ile to water restored TPI levels to those found in wild-type plants (Figure 2A; unpaired t test, P = 0.4507). When Ile was added directly to OS, TPI levels in asTDS1 plants were still lower than those in wild-type plants (Figure 2A; unpaired t test, P = 0.0427). However, adding Ile to OS and then applying these to the puncture wounds significantly increased TPI activity in asTDS1 plants compared with OS-treated asTDS1 plants (Figure 2A; unpaired t test, P = 0.0159). The restoration of TPI activity by Ile supplementation at the wound site in asTDM and TDVIGS plants could be attributed either to the restoration of the biosynthetic needs of TPI production or to its signaling.

View larger version (15K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 2. Silencing TD in asTD and TDVIGS Plants Impairs OS-Elicited TPI Activity. (A) Mean TPI activity (±SE) in wild-type, asTDM2, and asTDS1 plants of three replicate node +1 leaves that were harvested 3 d after being wounded and treated with 20 µL of either deionized water (W) or M. sexta OS supplemented with 0.625 µmol of Ile (+Ile). Leaves from control plants (C) were left intact and untreated. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; ** P < 0.01). (B) Mean TPI activity (±SE) in EV and TDVIGS plants of four to five replicate node +1 leaves that were harvested 3 d after being wounded and treated with 20 µL of either deionized water (W) or M. sexta OS supplemented with 0.625 µmol of Ile (+Ile). Leaves from control plants (C) were left intact and untreated. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; *** P < 0.01).

TD Silencing Impairs JA Signaling in asTD and TDVIGS Plants
More JA is elicited from wounded leaves treated with M. sexta OS than from leaves that have only been wounded (Halitschke et al., 2001). To determine whether M. sexta OS elicit the same rapidly increasing and declining JA-Ile pools, leaves at node +1 from four independently treated plants from each genotype, wild type and asTDM2, were wounded, treated with OS (Figure 3A ), and analyzed by LC-MS at each harvest time. As expected, in treated wild-type leaves, a JA burst was elicited within 30 min, reached maximum levels at 1 h, and declined rapidly after 1.5 h (Figure 3B). Similar responses were observed in JA-Ile pools in treated wild-type leaves (Figure 3C). The JA burst in asTDM2 plants was similar to that in wild-type plants but waned more slowly at 1.5 h after elicitation (Figure 3B; unpaired t test, P = 0.0120). The OS-elicited JA-Ile burst in asTDM2 plants was less than that in wild-type plants, with pools being significantly lower (36 and 68%) at 0.5 and 3 h, respectively (Figure 3C; unpaired t test, P 0.0237). To determine whether JA-Ile is synthesized from JA and Ile at the wound site, node +1 leaves from wild-type plants were wounded and immediately treated with OS containing 0.625 µmol of [¹³C₄]Thr or [¹³C₆]Ile (Figure 3A). Four replicate plants were harvested for each treatment and harvest time to measure the elicited kinetics of JA and ¹³C-labeled JA-Ile by LC-MS analysis. Adding [¹³C₄]Thr and [¹³C₆]Ile to OS reduced the levels of JA compared with OS (Figures 3B and 3D). Adding Thr to an OS-elicited wound reduced the maximum JA values by 5.5 nmol/g fresh weight (cf. [¹³C₄]Thr treatments: 6 nmol/g fresh weight in Figure 3D with OS-elicited values of 11.5 nmol/g fresh weight in Figure 3B). Adding the more efficiently incorporated amino acid, Ile, reduced JA values even further (to 4 nmol/g fresh weight; Figure 3D). The reduced levels of JA were used to synthesize JA-Ile. Significant quantities of ¹³C-labeled JA-Ile were detected when either [¹³C₄]Thr or [¹³C₆]Ile was applied, demonstrating that [¹³C₄]Thr was rapidly converted to Ile at the wound site and used to synthesize ¹³C-labeled JA-Ile. As expected, Thr was incorporated less efficiently into JA-Ile than was Ile (Figure 3E), demonstrating that the conjugation capacity of an elicited leaf is limited by substrate availability. Compared with wild-type plants, asTDM2 plants were less efficient in incorporating [¹³C₄]Thr into ¹³C-labeled JA-Ile at 0.5 h after elicitation (Figure 3F; unpaired t test, P = 0.0183), suggesting that mildly silencing TD expression correlated with detectable reductions in the conversion of Thr to Ile and its subsequent incorporation into JA-Ile. When wounds were treated with 0.625 µmol of JA (Figure 3A), plants sustained increased JA-Ile pools for 2.5 h (Figure 3G), demonstrating that the level of JA regulates the level of JA-Ile. Under these experimental conditions, when JA is not limited, clear differences between the abilities of asTDM2 plants and wild-type plants to produce JA-Ile were readily discerned: the amount of JA-Ile in asTDM2 plants was significantly lower than that in wild-type plants (Figure 3G; two-way ANOVA, F_1,16 = 9.768; P = 0.0065). When both [¹³C₆]Ile and JA were applied to wounded wild-type and asTDM2 plants, the levels of JA-Ile did not differ (Figure 3H; two-way ANOVA, F_1,16 = 0.999; P = 0.3324). These results demonstrate that Ile limits JA-Ile synthesis in asTDM2 plants more than in wild-type plants and that the JA-Ile conjugation enzyme in asTDM2 plants is equally active in wild-type plants. These experiments also demonstrate that the JA and JA-Ile bursts that erupt when M. sexta OS are introduced into a wound can be simulated by adding JA to a wound.

View larger version (29K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 3. OS-Elicited JA and JA-Ile Are Regulated by Thr, Ile, and JA; Silencing TD in asTDM2 Plants Reduces JA-Ile. (A) Numbering of leaf positions in rosette-stage N. attenuata plants. The leaf undergoing the source–sink transition (T) was designated as growing at node 0. The treated leaf growing at node +1, which is older by one leaf position than the source–sink transition leaf, was wounded with a fabric pattern wheel, and the resulting puncture wounds (W) were immediately treated with 20 µL of M. sexta OS, OS containing 0.625 µmol of 13C4-labeled Thr (13C4 Thr) or 13C6-labeled Ile (13C6 Ile), water containing 0.625 µmol of JA (JA), or water containing 0.625 µmol of JA and 13C6-labeled Ile. The treated leaves were harvested to measure JA, JA-Ile, and isotope-labeled JA-Ile. (B) and (C) Mean ± SE JA (B) and JA-Ile (C) concentrations in OS-treated leaves of four replicate wild-type and asTDM2 plants. Asterisks represent significant differences between members of a pair analyzed at the same time after OS elicitation (unpaired t test: * P < 0.05). FW, fresh weight. (D) and (E) Mean ± SE JA (D) and isotope-labeled JA-Ile (E) concentrations in [13C4]Thr- or [13C6]Ile-treated leaves of three replicate wild-type plants. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; *** P < 0.001; **** P < 0.0001). (F) Mean ± SE isotope-labeled JA-Ile concentrations in [13C4]Thr-treated leaves of three replicate wild-type and asTDM2 plants. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05). (G) Mean ± SE JA-Ile concentrations in JA-treated leaves of three replicate wild-type and asTDM2 plants. Asterisks represent significant differences between wild-type and asTDM2 plants (two-way ANOVA, Fisher's PLSD: ** P < 0.01). (H) Mean ± SE isotope-labeled JA-Ile concentrations in JA- and [13C6]Ile-treated leaves of three replicate wild-type and asTDM2 plants.

View larger version (26K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 4. Silencing TD in asTDS1 and TDVIGS Plants Reduces JA-Ile. (A) Node +1 leaves were wounded with a fabric pattern wheel and the resulting puncture wounds (W) immediately treated with 20 µL of M. sexta OS, OS containing 0.625 µmol of 13C4-labeled Thr (13C4 Thr), or water containing 0.625 µmol of JA (JA). The treated leaves were harvested to measure JA, JA-Ile, and isotope-labeled JA-Ile. (B) and (C) Mean ± SE JA (B) and JA-Ile (C) concentrations in OS-treated leaves of four replicate wild-type and asTDS1 plants. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; ** P < 0.01; *** P < 0.001). FW, fresh weight. (D) Mean ± SE isotope-labeled JA-Ile concentrations in [13C4]Thr-treated leaves of three replicate wild-type and asTDS1 plants. Asterisks represent significant differences between members of a pair (unpaired t test: ** P < 0.01; *** P < 0.001). (E) and (F) Mean ± SE JA-Ile concentrations in leaves of five replicate EV and TDVIGS plants. Leaves were harvested 2 h after JA (E) or 1 h after OS (F) treatment. Asterisks represent significant differences between EV and TDVIGS plants (unpaired t test: * P < 0.05; ** P < 0.01). (G) Mean ± SE isotope-labeled JA-Ile concentrations of five replicate EV and TDVIGS plants 1 h after leaves were wounded and treated with OS and [13C4]Thr. Asterisks represent significant differences between EV and TDVIGS plants (unpaired t test: ** P < 0.01).

Adding JA or JA-Ile to wounds of both wild-type and asTDM2 plants significantly increased TPI above the levels reached when nothing was added to wounds (Figure 5A ; unpaired t test, P 0.0035). JA addition, which was demonstrated to produce sustained differences in endogenous JA-Ile levels between wild-type and asTDM2 plants (Figure 3G), elicited TPI in asTDM2 plants at levels that were 40% lower than those in wild-type plants (Figure 5A; unpaired t test, P = 0.0361). When plants were treated with JA-Ile, the induced TPI responses did not differ between wild-type and asTDM2 lines (Figure 5A; unpaired t test, P = 0.8197), although they were significantly lower than the TPI responses elicited in wild-type plants by JA treatment. Higher levels of nicotine resulted when plants were treated with JA or JA-Ile and not only wounded (Figure 5A; unpaired t test, P 0.0022). Nineteen percent less nicotine accumulated in response to JA treatment in asTDM2 plants compared with wild-type plants (Figure 5A; unpaired t test, P = 0.0396), whereas the responses to JA-Ile treatment did not differ between asTDM2 and wild-type plants (Figure 5A; unpaired t test, P = 0.4214). However, unlike levels of TPI, the levels of nicotine elicited in plants treated with JA-Ile were much higher than in plants treated with JA (Figure 5A). Levels of chlorogenic acid in JA or JA-Ile treatment were the same in the untreated control. Although similar levels of diterpene glycosides were elicited by either JA or JA-Ile treatment, the levels did not differ between asTDM2 and wild-type plants (see Supplemental Figure 5 online), demonstrating that these secondary metabolites are not differentially elicited by JA and JA-Ile. Differences in the ability of JA and JA-Ile to elicit nicotine and TPI may reflect different rates of absorption in treated leaves or their transport within the plant. Most important, the elicited nicotine and TPI responses, which were significantly lower in JA-treated asTDM2 plants than in wild-type plants, did not differ between wild-type and asTDM2 plants when plants were treated with JA-Ile. These results demonstrated that JA-Ile could restore the direct defense responses of asTDM2 plants to the levels of these responses in wild-type plants; the next step was to determine whether resistance to M. sexta larvae could be similarly restored.

View larger version (10K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 5. JA-Elicited Herbivore Resistance, Nicotine, and TPI Production Are Impaired in asTDM2 Plants Compared with Wild-Type Plants but Restored by Adding JA-Ile. (A) Mean (±SE) TPI and nicotine levels in leaves from three replicate wild-type and asTDM2 plants growing at node +1, 3 d after being wounded and treated with 20 µL of deionized water (W), Ile (W+Ile), JA (W+JA), or JA-Ile conjugate (W+JA-Ile), all at 0.625 µmol. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05). FW, fresh weight. (B) Mean (±SE) mass of M. sexta larvae after 3, 6, and 9 d of feeding on 16 replicate wild-type plants and two lines of T2 transgenic plants (asTDM2 and asLOX3). Leaves were treated with 0.625 µmol of JA-Ile or left untreated (C). Top graph, asterisks represent significant differences between untreated wild-type plants and two lines of untreated T2 transgenic plants on day 9 (unpaired t test: * P < 0.05; *** P < 0.001). Bottom graph, asterisks represent significant differences between untreated and JA-Ile–treated plants on day 9 (unpaired t test: ** P < 0.01). asLOX3 plants, which are largely defenseless because of their impaired JA signaling (Halitschke and Baldwin, 2003), were included as a positive control for herbivore resistance.

Because we now understood that the traits responsible for the defects in herbivore resistance were associated with TD silencing, we were ready to examine herbivore resistance in the developmentally challenged asTDS plants. M. sexta larvae that fed on JA-treated asTDS plants gained less mass compared with those that fed on untreated asTDS plants, but the difference was not significant by day 9 (Figure 5B; unpaired t test, P = 0.134). Feeding on JA-Ile–treated asTDS plants, however, the larvae gained significantly less mass (45%) compared with those that fed on untreated asTDS plants by day 9 (Figure 5B; unpaired t test, P = 0.0027). These results demonstrate that even in plants with severely silenced TD that suffer from severe nutritional deficiencies, Ile is conjugated to JA at the wound site to mediate defense signaling. Supplementing wounds with JA-Ile restores a modicum of induced resistance in these severely growth-impaired plants.

Suppressing TD and JAR4 by VIGS Impairs JA Signaling and Herbivore Resistance
To further examine whether JA-Ile is the signal molecule that elicits herbivore resistance, we cloned the Arabidopsis JAR1 homolog JAR4 (GenBank accession number DQ359729) from N. attenuata using RT-PCR. To investigate whether JAR4 encodes the enzyme that conjugates amino acids to JA in N. attenuata, we collected amino acid sequences of JAR-like proteins using N. attenuata JAR4 as a query. Phylogenetic analysis revealed that these proteins clustered into three groups; JAR4 and JAR1 cluster together with three functionally unknown proteins (see Supplemental Figure 6 online) that share >60% amino acid identity (see Supplemental Figure 7 online), suggesting that they share similar functions as JAR1, namely, conjugating amino acid to JA (Staswick et al., 2002; Staswick and Tiryaki, 2004). The other Arabidopsis JAR family members, GH3.1, GH3.2, GH3.5, and GH3.17, which conjugate amino acids to indole-3-acetic acid (Staswick et al., 2005), clustered together in a separate group. DNA gel blotting revealed that JAR4 is a single-copy gene in the N. attenuata genome (see Supplemental Figure 8 online). These results suggested that JAR4 is a good candidate for the JA-conjugating enzyme in N. attenuata.

To determine whether JAR4 mRNA is elicited by wounding or OS treatment of wounds, plants were wounded with a fabric pattern wheel and treated with water or OS, and JAR4 mRNA accumulation was analyzed by quantitative real-time PCR. In response to wounding alone, JAR4 mRNA levels increased within 30 min, reached a maximum at 1.5 h, and declined after 3 h (Figure 6 ). Similar patterns of transcript accumulation were observed in OS-treated wild-type leaves, but these levels waned more slowly and did not return to control levels after 12 h (Figure 6).

View larger version (9K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 6. Accumulation of JAR4 Transcripts after Elicitation by Wounding and OS Treatments. Leaves at node +1 were wounded with a fabric pattern wheel, and the resulting wounds were immediately treated with 20 µL of deionized water (W) or with M. sexta OS in five replicate wild-type plants. The transcripts were analyzed by real-time PCR as means ± SE of five replicate leaves in arbitrary units from calibration with a 5x dilution series of cDNAs prepared from RNA samples extracted at 1 h after wounding.

View larger version (10K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 7. Silencing TD and JAR4 by VIGS Reduces Transcript and JA-Ile Accumulation. (A) VIGS of TD and JAR4 transcripts. Plants were inoculated with Agrobacterium harboring TRV constructs, which contain an EV, a 335-bp TD fragment (TDVIGS), or a 292-bp JAR4 fragment (JAR4VIGS). Fourteen days after inoculation, leaves were wounded, treated with 0.625 µmol of JA-Ile, and harvested 1 h later; or they were made available for M. sexta larvae to feed on for another 12 d, after which they were harvested (H); or they were harvested immediately from untreated plants (C). The transcripts were analyzed by real-time PCR as means ± SE of five replicate leaves in arbitrary units from a calibration with a 5x dilution series of cDNAs prepared from EV control RNA samples. Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; ** P < 0.01; **** P < 0.0001). (B) Mean ± SE JA-Ile concentrations in leaves of four to five replicate EV, TDVIGS, and JAR4VIGS plants. Fourteen days after inoculation, leaves were wounded, treated with 20 µL of either M. sexta OS or water containing 0.625 µmol of JA, and harvested 1 h later. Asterisks represent significant differences between EV and VIGS plants (unpaired t test: * P < 0.05; ** P < 0.01). FW, fresh weight.

View larger version (10K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 8. Silencing TD and JAR4 by VIGS Reduces Herbivore Resistance and TPI Activity; Adding JA-Ile Restores Them. (A) Mean ± SE mass of M. sexta larvae after 6 and 9 d of feeding on 10 to 13 replicate plants, each inoculated with Agrobacterium harboring TRV constructs, which contain an EV, a 335-bp TD fragment (TDVIGS), or a 292-bp JAR4 fragment (JAR4VIGS). Fourteen days after inoculation, leaves were either wounded and treated with 0.625 µmol of JA-Ile (JA-Ile) or left untreated (control). Asterisks represent significant differences between EV and VIGS plants (repeated-measurement ANOVA, Fisher's PLSD: * P < 0.05; ** P < 0.01). (B) Mean ± SE TPI activity of five replicate EV, TDVIGS, and JAR4VIGS plants. Fourteen days after inoculation, leaves were wounded, treated with 0.625 µmol of JA-Ile, and harvested 3 d later; or they were made available for M. sexta larvae to feed on for another 12 d, after which they were harvested (H); or they were harvested immediately from untreated plants (C). Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; ** P < 0.01).

View larger version (13K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 9. Effects on Herbivore Resistance of Thr or Ile Supplementation to Leaves of TD-Silenced and Wild-Type Plants. (A) and (B) Mean ± SE JA-Ile concentrations (A) and TPI levels (B) in leaves of four to five replicate EV and TDVIGS plants, each inoculated with Agrobacterium harboring TRV constructs, which contain either an EV or a 335-bp TD fragment (TDVIGS). Fourteen days after inoculation, leaves were wounded, treated with 20 µL of either M. sexta OS or OS containing 0.625 µmol of Thr or Ile, and then harvested 1 h after treatment for JA-Ile measurement. These plants were supplemented daily by spraying leaves with either water or 0.25 M Thr or Ile. Three days after OS treatment, newly hatched M. sexta larvae were placed on these plants. TPI activity was measured after 12 d of herbivore feeding. Asterisks represent significant differences between members of a pair (unpaired t test: ** P < 0.01; **** P < 0.0001). FW, fresh weight. (C) Mean ± SE mass of M. sexta larvae after 12 d of feeding on 16 to 19 replicate EV and TDVIGS plants treated as described for (A). Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05; ** P < 0.01). (D) Mean ± SE -KB levels in M. sexta frass from larvae that fed on EV and TDVIGS plants. Frass was collected from third- and fourth-instar larvae feeding on plants treated as described for (A). Asterisks represent significant differences between members of a pair (unpaired t test: * P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION METHODS REFERENCES

 
Because of the discovery, more than two decades ago, of the genes responsible for the biosynthesis of amino acids, plant biologists were able to determine which were essential for growth and development. In an attempt to improve the nutritional value of cereal crops, which have low levels of Lys and Thr, biologists have focused attention on the essential amino acids, Thr, Lys, Met, and Ile, which are synthesized via a common pathway (Azevedo et al., 1997). TD catalyzes the conversion of Thr to -KB, the first committed step in Ile biosynthesis (Umbarger, 1978).

The research presented here highlights the challenges of disentangling the multiple roles that the enzymes involved in amino acid biosynthesis can play in plants as well as after the plant has been ingested by an herbivore. This research also highlights the value of analyzing subtle phenotypes in plants for which the determinants of ecological performance are well understood.

TD's role in herbivore resistance was discovered with the transformants (asTDM) in which TD expression was mildly silenced (Figure 1). These plants had completely normal growth phenotypes, even under stringent competition regimes, but their resistance to herbivores was impaired (see Supplemental Figure 4 online), allowing researchers to understand TD's unusual transcriptional behavior in response to wounding, herbivore attack, and JA elicitation (Hildmann et al., 1992; Halitschke et al., 2001; Hermsmeier et al., 2001; Schittko et al., 2001). The susceptibility of asTDM plants to attack from M. sexta larvae was associated with the reduced levels of two inducible direct defenses: TPIs and nicotine (Figures 2 and 5). Previous research has demonstrated that silencing either of these defenses in N. attenuata plants increases the susceptibility of plants to attack from M. sexta larvae and enhances larval performance (Steppuhn et al., 2004; Zavala et al., 2004a, 2004b). Moreover, both of these direct defenses are elicited by JA signaling (Halitschke et al., 2004). The kinetics of the JA and JA-Ile bursts induced by larval elicitors were found to be subtly altered in asTDM plants (Figure 3). This observation led to the discoveries that JA is conjugated with Ile at the wound site and that herbivory-elicited TD supplies the Ile required for the formation of JA-Ile. Supplementing wounds in asTDM plants with Ile restored the wild-type kinetics of the JA-Ile burst and also elicited direct defenses. Treating asTDM plants with JA-Ile restored the plant's ability to elicit direct defenses and thereby the resistance of asTDM plants to attack from M. sexta larvae (Figure 5). These results highlight the dynamic role that JA-Ile plays in defense signaling and suggest that subtle changes in the kinetics of JA and JA-Ile accumulation after herbivore attack can profoundly affect defense elicitation.

JA-Ile's role as a defense signal was confirmed in the analysis of the asTDS plants, in which all of the subtle changes in defense signaling observed in asTDM plants were exaggerated. In asTDS plants, the OS-elicited JA and JA-Ile bursts observed in wild-type plants were much slower (Figure 4). The OS-elicited JA-Ile production was lower (Figure 4), and JA-Ile treatment effectively restored herbivore resistance (Figure 5). Hence, although the developmental defects of asTDS plants prevented direct comparisons of herbivore resistance with that in wild-type plants, some of the defensive deficiencies of asTDS plants could be compensated for by Ile or JA-Ile treatment. The ability to complement these defensive deficiencies in plants suffering from severe nutritional deficiencies underscores the importance of JA-Ile in defense signaling. The VIGS experiments in N. attenuata, in which TD activity could be strongly silenced in a developmentally normal plant, reconfirmed TD's involvement in Ile synthesis and indicated that JA-Ile conjugation is limited by the supply of Ile in wounded tissues (Figure 4) and that JA-Ile regulated direct defenses and herbivore resistance (Figure 5).

It has long been known that JA is metabolized to its volatile counterpart, MeJA, and numerous conjugates with O-glucosides, hydroxylation, and amino acids (Sembdner and Parthier, 1993; Sembdner et al., 1994). The glycosylated forms and amino acid derivatives have been viewed as mere conjugates of JA, which may be important for hormone homeostasis. Because all of the applied conjugates could be deesterified to JA, JA and JA conjugates were thought to have the same effect (Schaller et al., 2004). However, recent reports have demonstrated that JA conjugates have their own activities. Transgenic Arabidopsis plants that constitutively express an S-adenosyl-L-Met:JA carboxyl methyltransferase expressed JA-responsive genes, including VSP and PDF1.2. Furthermore, the transgenic plants showed enhanced resistance to the virulent fungus Botrytis cinerea (Seo et al., 2001). Studies that have applied synthetic JA–amino acid conjugates to plants suggest that the spheres of activity within JA–amino acid conjugates differ widely. For example, treatment of barley (Hordeum vulgare) leaves with JA-Ile elicits JA-induced protein without Ile cleavage from JA (Kramell et al., 1997). JA-Ile, JA-Phe, and JA-Leu conjugates elicit accumulation of the flavonoid phytoalexin, sakuranetin, in rice (Oryza sativa) leaves, but JA-Trp does not (Tamogami et al., 1997). However, it had not been previously appreciated that JA conjugates had elicitor-induced dynamics that were comparable to those of JA and that subtle changes in these dynamics were associated with changes in defense function.

The pioneering work of Staswick and colleagues (Staswick et al., 2002; Staswick and Tiryaki, 2004) has demonstrated that the JA-responsive gene in Arabidopsis (JAR1) adenylates JA's carboxyl group and that adenylated JA is actively conjugated with various amino acids, of which Ile is quantitatively the most important. The mutant defective in JAR1 (jar1-1) exhibits decreased resistance to the soil fungus Pythium irregulare (Staswick et al., 1998), implying that JA-Ile is involved in pathogen resistance. The analysis of JAR4VIGS plants demonstrated that JAR4 is involved in JA-Ile conjugation (Figure 7) and that JAR4VIGS plants are susceptible to attack by M. sexta (Figure 8), indicating that JA-Ile is involved in herbivore resistance in N. attenuata. Further analyses of other JA–amino acid conjugates at the attack site will be required to determine whether other JA conjugates are equally as dynamically elicited and whether these conjugates also elicit specific developmental and defense responses in the plant.

Reduced levels of JA-Ile in asTD and TDVIGS plants resulted in reduced levels of TPI and nicotine. Treatment of plants with JA and JA-Ile elicited different TPI and nicotine responses, which may be attributable to different absorption and transport rates or to the different elicitation activities of these chemicals. The JA-Ile burst can account for 13% of the elicited JA burst (Figures 3B and 3C). That the quantities of JA-Ile are smaller than those of JA may be attributable more to a rapid metabolism of JA-Ile to unknown structures than to the conversion of JA to JA-Ile. Alternatively, JA may be converted to MeJA or other conjugates. The rapid declines in JA and JA-Ile may be attributable to their binding to putative receptor(s), which have not yet been identified. Identification of the JA receptor(s), when it occurs, will be a breakthrough that will clarify the structural basis for the differences in levels as well as the contents of these dynamic metabolites.

The role of JA in systemic signaling was recently demonstrated in an elegant set of reciprocal grafting experiments. Li and coworkers (2002, 2003) grafted the JA biosynthetic mutant spr-2, known to be defective in fatty acid desaturase required for JA biosynthesis, onto the JA response mutant jai-1, known to be defective in a homolog of the Arabidopsis CORONATINE-INSENSITIVE1 gene (Xie et al., 1998), in different combinations and analyzed the resulting wound-induced expression of the proteinase inhibitor II gene. Their results demonstrated that the JA biosynthetic pathway was required to produce the long-distance signal, suggesting that JA or related compounds derived from the octadecanoid pathway function as systemically transmitted signals in tomato. In various Nicotiana species, nicotine synthesis in the roots is activated by leaf wounding. In N. attenuata, this systemic response is known to require JA signaling (Halitschke and Baldwin, 2003). JA-Ile elicits a larger accumulation of nicotine compared with JA (Figure 5A). This finding suggests that JA-Ile is the long-distance signal that elicits nicotine in roots. However, JA-Ile was not detected in roots when leaves were treated with OS, JA, or JA-Ile (data not shown), indicating that subsequent metabolites of JA-Ile or unknown molecules elicited by JA-Ile may be involved.

When attacked by herbivores, N. attenuata plants produce JA and activate TD in the attacked tissues. Ile synthesized from Thr by TD is conjugated with JA by JAR4. The resulting JA-Ile elicits the accumulation of the direct defenses, TPI and nicotine, which in turn decrease the performance of Nicotiana-adapted herbivores. The fact that the herbivore resistance of both TD- and JAR4-silenced plants could be restored by treatment with signal quantities (0.625 µmol/plant) of JA-Ile suggests that TD's defensive role can be largely attributed to its role in defense signaling. When leaves of TD-silenced plants were supplemented with large nutritional quantities (2.1 mmol/plant) of Thr, larval performance increased. This result, in conjunction with the demonstration that TD levels in frass of larvae that fed on water-treated TD-silenced plants were 54% of those in frass of larvae that fed on water-treated EV plants, suggests that TD could play a defensive role by reducing the availability of Thr to feeding larvae. However, it remains unclear how often the dietary Thr levels are sufficiently high for plants to realize a defensive benefit of delivering TD to the midgut of larvae, because resistance of the JAR4- and TD-silenced plants was similarly impaired and both could be restored by JA-Ile treatment (Figure 8A).

Supplementation of Thr or Ile in plants with different levels of TD activity helped us to evaluate TD's antinutritive role. Providing additional Thr to leaves had positive effects on herbivore performance in TD-silenced plants. However, the insufficient supply of Ile in these plants impaired JA-Ile–mediated signaling, and providing these plants with supplemental Ile restored their direct defenses and TPI levels and increased herbivore resistance. Our data also demonstrate that in normal TD-expressing plants, Ile supplementation benefited larvae. One possible explanation is that extra Ile provides feedback that inhibits TD activity in plants or in the insect midgut. We plan to explore the molecular mechanisms involved in the regulation of TD activity in the plant as well as in the insect midgut, following the lead of Chen et al. (2005), who demonstrated that the regulatory domain of tomato TD is missing in the TD protein isolated from insect midgut and frass; this domain is responsible for feedback inhibition by Ile in tomato.

To summarize, we propose dual roles for TD in herbivore defense: JA-Ile–mediated signaling and antinutritive defense (Figure 10 ). When attacked by herbivores, plants produce JA and activate TD in the attacked tissues. Ile synthesized from Thr by TD is conjugated with JA by JAR4. The resulting JA-Ile elicits the accumulation of the direct defenses, TPI and nicotine. TD also plays additional defensive roles by limiting the supply of amino acids for herbivore growth when leaves are ingested by herbivores.

View larger version (13K): [in this window] [in a new window] [as a PowerPoint slide]   Figure 10. Proposed Roles of TD in Herbivore Resistance. Two roles are proposed: (1) as an antinutritive defense that decreases Thr availability in the digestive tracts of herbivores after ingestion; (2) as a mediator of JA signaling, by supplying Ile for conjugation with JA at the wound site and subsequently eliciting various direct defenses. JA biosynthetic enzymes and a JA–amino synthetase are boxed. Dashed arrows represent signal transduction