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 α-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 α-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 α-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 5× 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 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).

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.

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).

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.

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 5× dilution series of cDNAs prepared from RNA samples extracted at 1 h after wounding.

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 5× 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.

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).

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 α-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).

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 pathways. LOX, lipoxygenase; AOS, allene oxide synthase; AOC, allene oxide cyclase; OPR, 12-oxo-phytodienoic acid reductase; JAR1, JA–amino synthetase.