Planta (1999) 209: 87±95
Tobacco mosaic virus inoculation inhibits wound-induced jasmonic acid-mediated responses within but not between plants Catherine A. Preston1,2, Carrie Lewandowski2, Alexander J. Enyedi3, Ian T. Baldwin1,2 Max-Planck-Institut fuÈr Chemische OÈkologie, Tatzendpromenade 1A, D-07745 Jena, Germany Department of Biological Sciences, SUNY at Bualo, Bualo, NY 14260-1300, USA 3 Department of Biological Science, Western Michigan University, Kalamazoo, MI 49008-3899, USA 1 2
Received: 6 January 1999 / Accepted: 11 January 1999
Abstract. Jasmonic acid (JA) and salicylic acid (SA) have both been implicated as important signal molecules mediating induced defenses of Nicotiana tabacum L. against herbivores and pathogens. Since the application of SA to a wound site can inhibit both wound-induced JA and a defense response that it elicits, namely nicotine production, we determined if tobacco mosaic virus (TMV) inoculation, with its associated endogenous systemic increase in SA, reduces a plant's ability to increase JA and nicotine levels in response to mechanical damage, and evaluated the consequences of these interactions for the amount of tissue removed by a nicotine-tolerant herbivore, Manduca sexta. Additionally, we determined whether the release of volatile methyl salicylic acid (MeSA) from inoculated plants can reduce wound-induced JA and nicotine responses in uninoculated plants sharing the same chamber. The TMVinoculated plants, though capable of inducing nicotine normally in response to methyl jasmonate applications, had attenuated wound-induced JA and nicotine responses. Moreover, larvae consumed 1.7- to 2.7-times more leaf tissue from TMV-inoculated plants than from mock-inoculated plants. Uninoculated plants growing in chambers downwind of either TMV-inoculated plants or vials releasing MeSA at 83- to 643-times the amount TMV-inoculated plants release, exhibited normal wound-induced responses. We conclude that tobacco plants, when inoculated with TMV, are unable to elicit normal wound responses, due likely to the inhibition of JA production by the systemic increase in SA induced by virus-inoculation. The release of volatile MeSA from inoculated plants is not sucient to in¯uence the wound-induced responses of neighboring plants.
Abbreviations: JA = jasmonic acid; MeJA = methyl jasmonate; MeSA = methyl salicylic acid; SA = salicylic acid; TMV = tobacco mosaic virus Correspondence to: I.T. Baldwin; E-mail:
[email protected]; Fax: (3641) 643653
Key words: Herbivory ± Jasmonic acid ± Nicotiana (defense responses) ± Methyl salicylic acid ± Nicotine ± Tobacco mosaic virus
Introduction Jasmonic acid (JA) and salicylic acid (SA) are both important components of signal transduction cascades activating plant defense responses against herbivore and pathogen attack. A notable dierence between these two cascades is their dierent kinetics of elicitation. After leaf wounding, JA concentrations in wounded leaves increase rapidly and transiently, with pools waxing and waning within minutes (Baldwin et al. 1994; Blechert et al. 1995; Baldwin et al. 1997; Creelman and Mullet 1997). This JA burst precedes the activation of several defense-related genes (Creelman et al. 1992; Doares et al. 1995b) and the accumulation of plant defense compounds, including proteins (Doares et al. 1995b; Howe et al. 1996), alkaloids (Dittrich and Kutchan 1991; Baldwin et al. 1994, 1997; Baldwin 1996), ¯avonoids (Gundlach et al. 1992; Mueller et al. 1993), and phenolics (Mizukami et al. 1993). In comparison, SA displays a slower, more attenuated response to tobacco mosaic virus (TMV) inoculation, with systemic increases occurring several days after inoculation and lasting for several weeks (Raskin 1992). Salicylic acid (as well as other mammalian lipoxygenase and cyclooxygenase inhibitors) is known to inhibit the production of JA and JA-related defenses when applied to the wound site immediately after damage (White 1979; PenÄa-CorteÂs et al. 1993; Doares et al. 1995a,b; Baldwin et al. 1996, 1997), suggesting that the SA signal cascade may inhibit the JA signal cascade. Recently, Creelman and Mullet (1997) proposed that the potential antagonism between these two pathways may be minimized by the dierent kinetics of elicitation or through the speci®c recognition of pathogens and herbivores. Mur et al. (1997) report unpublished results that both
88
C.A. Preston et al.: TMV inoculation prevents wound-induced defenses
SA and JA accumulation occur simultaneously in tobacco following bacterial inoculation, suggesting that concentrations of endogenous SA elicited by inoculation or its intracellular location prevent the inhibition of JAinduced responses and that the ability of SA to inhibit the JA pathway may only occur when SA is applied at high concentrations not normally attained after inoculation. Nicotiana tabacum L. cv. Xanthi-nc (NN genotype) is an ideal plant for examining the interaction of these two signal transduction cascades. The endogenous accumulation of SA in response to inoculation by TMV is well characterized in this genotype (Enyedi et al. 1992; Raskin 1992; Klessig and Malamy 1994; Yalpani et al. 1993; Shulaev et al. 1995). Salicylic levels increase by nearly 50-fold in TMV-inoculated leaves and 10-fold systemically in uninoculated leaves (Malamy et al. 1990). Additionally, Shulaev et al. (1997) report that inoculated plants release the methyl ester of SA, methyl salicylate (MeSA), into the air in quantities sucient to increase resistance in adjacent, uninoculated plants growing in connected bell jars. When the leaves of N. sylvestris are wounded, JA concentrations increase transiently in leaves minutes after damage in proportion to the amount of wounding and elicits proportional increases in whole-plant nicotine concentrations days later (Baldwin et al. 1994, 1997; Ohnmeiss et al. 1997). The increase in JA in wounded leaves and its apparent transport to the roots, where nicotine is synthesized, is necessary and sucient to activate de-novo nicotine production in this species (Baldwin et al. 1997; Zhang and Baldwin 1997). Jasmonic acid and its methyl ester, MJ, increase nicotine production in other Nicotiana species (Baldwin and Ohnmeiss 1993; Baldwin 1996). By increasing the endogenous SA levels and releasing MeSA into the air, a virus-inoculated tobacco plant may not only inhibit its own abilities to synthesize JA and accumulate nicotine in response to wounding or herbivore attack, but may also compromise the defensive ability of nearby, uninoculated plants. To examine these interactions, we inoculated plants with TMV on Day 0 and wounded them on Day 4, to allow for an increase in systemic levels of SA prior to wounding and measuring JA and nicotine inductions. Virus-inoculated plants may not be able to increase nicotine production due either to impaired abilities to (i) produce JA or (ii) respond to JA with increased nicotine production, and we tested the latter possibility by treating the roots of infected plants with MJ. To examine the consequences of TMV-inoculation on herbivore resistance, we compared the amount of plant material consumed by the larvae of tobacco hornworms (Manduca sexta) feeding on wounded and unwounded, TMV-inoculated and mock-inoculated plants. To examine potential between-plant interactions, we measured the induced responses of control and TMV-inoculated plants grown together or alone in small growth chambers. Additionally, we grew uninoculated plants in small growth chambers with devices that released MeSA at rates 83- to 643-fold greater than that reported from a TMV-inoculated tobacco plant (283 ng h)1 plant)1)
(Shulaev et al. 1997) and examined their wound-induced responses. Materials and methods Plant growth. Nicotiana tabacum L. cv. Xanthi-nc seeds carrying the N gene were germinated in soil for 11±13 d and then transferred to 40-L hydroboxes for 5±6 d of growth in a complete nutrient solution as described in Baldwin et al. (1994). Plants were then transferred to individual opaque 1-L containers with a no-nitrogen hydroponic nutrient solution (Ohnmeiss and Baldwin 1994), which was supplemented with 28 mg N as KNO3. Plants were grown in a glasshouse and received supplemental lighting from 400-W sodium vapor lamps for 16 h d)1, which provided a minimum photon ¯ux density of 220 lmol m)2 s)1 photosynthetically active radiation. For the duration of all experiments, the glasshouse temperature was kept below 32 °C, necessary to maintain the plants' ability to produce PR-1 proteins and increase resistance to TMV (Malamy et al. 1992). After 9 d of growth, all plants received an additional 28 mg N as KNO3 before starting the experiments on Day 0. Experiment 1. Can TMV-inoculated plants respond to MJ treatments? Plants were randomly assigned to four treatment groups: (1) mock-inoculated, unwounded plants (C); (2) plants wounded with a fabric pattern-tracing wheel as described below (W); (3) TMV-inoculated, unwounded plants (V); and (4) TMV-inoculated, wounded plants (VW). Treatments 1, 3, and 4 received either 0, 5, 25, or 100 lg of MJ in the hydroponic solution to quantitatively induce nicotine production in the roots. There were 5 replicates per treatment and MJ addition level, for a total of 60 plants. On Day 0, carborundum powder (320 grit; Fisher Scienti®c, Pittsburg, Penn., USA; C192-500), was dispersed over the plants which were then inoculated by treating each of the largest four leaves with 1.25 lg of TMV suspended in 50 lL distilled water (virus) or sprinkled with distilled water (mock). Treated leaves were then lightly rubbed with a gloved hand. Lesions (hypersensitive response) were evident within 48 h after inoculation. The U1 strain of TMV was originally isolated from 1000 g of TMV-infected `nn' genotype Xanthi tobacco leaves exhibiting typical symptoms of foliar mosaic and stored as a suspension (20 mg ml)1) in distilled water. On Day 4, MJ was administered to the roots, plants were wounded, and on Day 8, all plants were harvested for the determination of whole-plant nicotine concentration as described below. Since necrotic tissues are not likely to accumulate nicotine and their inclusion may underestimate the whole-plant nicotine response, the experiment was repeated (with the exception of treatment group 2) and all non-necrotic tissues were harvested for nicotine analysis. Experiment 2. Eect of TMV inoculation on wound-induced JA and nicotine responses within and between plants. To determine if TMVinoculated plants release sucient quantities of MeSA into the air to inhibit the wound-induced defenses of nearby healthy plants, mock- and virus-inoculated plants were placed upwind of uninoculated plants in small growth chambers with uni-directional air¯ow. Six plants were placed into transparent plastic 18.5-L growth chambers (Rubbermaid; Wooster, Ohio, USA) and covered with a piece of UV transparent Plexiglas (UV-T) sealed with vacuum grease (Fig. 1A). Air¯ow was controlled with a fan (7 cm diameter; Howard Industries; Milford, Ill., USA) attached to the outside of each chamber, which pulled air through ®ve small holes drilled on the opposite side of the growth chamber and across the plants. Fan speed was regulated by a rheostat and used to control the temperature within the growth chambers. Air¯ow through the chambers varied from 1 L min)1 to 24 L min)1, increasing on warmer days to keep the chamber temperature below 32 °C. All chambers were placed on benches in a glasshouse with metal-halide lights providing supplemental illumination. Large fans at the edge of the glasshouse bench circulated air exiting the chambers. On
C.A. Preston et al.: TMV inoculation prevents wound-induced defenses Day 0, 64 plants were haphazardly chosen and assigned one of two inoculation treatments, mock (40 plants) and virus (24 plants). The largest four leaves were each treated as in Experiment 1 but with 2.5 lg of TMV suspended in 100 lL distilled water on each leaf for virus-inoculation. Environment 1 contained four mock-inoculated plants upwind of two uninoculated plants in each replicate chamber, and Environment 2 had three virus-inoculated plants and one mock-inoculated plant upwind of two uninoculated plants in each replicate chamber (Fig. 1C). There were 8 replicate chambers for each of these environments and hence 96 plants used in this experiment. Experiment 3. Eect of enhanced MeSA emissions on wound-induced JA responses. To investigate the levels of airborne MeSA required to aect the wound-induced defenses of plants downwind, MeSAreleasing devices were placed in the small growth chambers with uninoculated tobacco plants (Environment 3, Fig. 1C). For these devices, we used crimped-top 2-mL GC-MS vials with Te¯on inner linings containing 1 mL of MeSA (Sigma Chemical Co., St. Louis, Mo., USA; lot 71H0485), and punctured the liner once with an embroidery needle. A preliminary experiment was performed to determine the release rate from the vials by successively weighing the vials for 3 d prior to the start of the experiment. Three vials were chosen to create a range of MeSA rates: low (23 lg h)1), medium (127 lg h)1), and high (182 lg h)1). Release rates are signi®cantly greater than amounts reported by Shulaev et al. (1997) for the release of MeSA from a 5-week-old TMV-inoculated tobacco plant. On Day 0, the release devices were placed by the air inlets and six uninoculated plants were placed into each chamber; a control chamber had no MeSA-releasing device. Vials were reweighed periodically throughout the experiment to determine the amount of MeSA released. Experiment 4. Eect of TMV inoculation on caterpillar consumption. To determine if an impaired wound-induced defense response aects the amount of plant material consumed by a nicotinetolerant herbivore, tobacco hornworm larvae were placed on plants that were mock-inoculated, unwounded (C); mock-inoculated, wounded (W); TMV-inoculated, unwounded (V); or TMV-inoculated, wounded (VW). Tobacco hornworm eggs (Carolina Biological Supply, Burlington, N.C., USA) were hatched on untreated N. tabacum leaves in plastic rearing chambers placed within a growth chamber at 18 °C and 16:8 h light:dark cycles. One day after hatching, larvae weighing between 5.0 and 7.2 mg were used for the experiment. On Day 0, plants were inoculated as in Experiment 1 and on Day 4, plants were wounded and a single 24 h-old larva was placed on a leaf at one phytotactic node above the youngest inoculated or wounded leaf. On Day 8, a single leaf (1 leaf node above that on which the caterpillar was feeding and therefore not inoculated or wounded) was harvested for nicotine analysis. As leaves at dierent leaf nodes can vary in nicotine content (Ohnmeiss et al. 1997), care was taken to collect the leaf from the same phytotactic position on each plant. On Day 11, caterpillars were removed from the plants and all the leaves and their damaged areas were traced. The tracings were photocopied and digitized to calculate the amount of leaf area and the percentage of tissue consumed by the larvae. Quanti®cation of MeSA emissions. The amount of volatile MeSA inside and exiting the growth chambers housing the uninoculated and TMV-inoculated plants and the MeSA-releasing devices were sampled with traps placed inside the chamber at the inlet (next to the MeSA sources for Experiment 3), and outside the chamber at the outlet of the fan (Fig. 1A). Traps contained 40 mg of 80/100 mesh Super Q (Alltech Associates, Deer®eld, Ill., USA) secured in a 6-cm hollow glass rod (2 mm i.d., 4 mm o.d.) with glass wool on both ends. Prior to collection, 1 lg of [13C]MeSA in 20 lL of CH2Cl2 was added as an internal standard to each trap. The [13C]MeSA was synthesized from the methylation of salicylic acid
89
(Sigma; Lot 28F3421) with diazald-n-methyl-13C (Sigma; Lot 29,598±1). Air was drawn through the traps at 150 mL min)1 by two Gast pumps (model DAA-V114-GB; Gast Mfg., Benton Harbor, Mich., USA) connected to a 15-port manifold, which allowed volatiles to be collected from all chambers simultaneously. Volatiles were trapped for 6 h, from 10 a.m. to 4 p.m., on Day 3 of the experiment for Environments 1 and 2 and on Day 2 for Environment 3. After collection, traps were immediately eluted with 300 lL of CH2Cl2 and analyzed by GC-MS. One microliter of each sample was injected into a 230 °C injection port with an automatic sampler (Model 7673; Hewlett-Packard, Kennett Square, Penn., USA) using a 0.5-min splitless injection; volatiles were separated on a DB-Wax column (3 m long, 0.25 mm i.d.) maintained at 50 °C for 1 min and subsequently heated at 10 °C min)1 to 230 °C, with a continuous He ¯ow of 0.6 mL min)1, as determined by an electronic pressure control unit on the gas chromatograph (Hewlett-Packard 5890 series II). Methyl salicylic acid was detected with a mass spectrometer (Hewlett-Packard 5971) operated in the selective-ion-monitoring (SIMS) acquisition mode using 152 and 153 m/z as target ions for MeSA and [13C]MeSA, respectively. Both target ions had a retention time of 13.10 min and the contribution of unlabeled MeSA (m/z = 152) in the internal standard (m/z) was corrected by comparison with standards. The amounts of MeSA trapped in the two traps were summed to determine the amount of MeSA released within each growth chamber. Wound-induced changes in leaf JA and whole-plant nicotine. Two fully-expanded leaves were wounded with two revolutions of a fabric-pattern wheel (Dritz, Spartanburg, S.C., USA) on each side and parallel to the mid-rib within a 6-cm2 window cut from a piece of plastic. The wounding protocol produced 50 punctures per 6 cm2 on each wounded leaf (Fig. 1B). The leaves to be wounded had been previously inoculated (mock or virus) or, on uninoculated plants, comparable leaves were chosen. Figure 1C depicts the plants wounded for JA analysis in each environment. After wounding, plants were returned to their original positions in the small chambers. All plants used for JA analysis were harvested at 90 min after wounding, the time when JA values were highest in leaves after the same wounding protocol had been used on N. sylvestris plants (Baldwin et al. 1997; Ohnmeiss et al. 1997). The damaged tissue was harvested and similar-sized unwounded areas were harvested from two undamaged leaves on the same plants. Wounded and unwounded leaf areas from virus-inoculated plants were selected so as to avoid viral lesions. In N. sylvestris, unwounded leaves can serve as control samples for wounded leaves on the same plant for the measurement of wound-induced JA production because, at 90 min after wounding, the JA induction has not spread systemically to other unwounded leaves (Ohnmeiss et al. 1997). The leaf material from each plant, paired by wounding treatment, was weighed, ¯ash-frozen in liquid N2 and stored at )80 °C until JA or nicotine analysis. A mock-inoculated plant was removed from all Environment 1 chambers so that all environments had four plants per chamber after the harvest of plants to be analyzed for JA on Day 4. Plants to be analyzed for nicotine were wounded on ®ve leaves with four revolutions of the fabric-pattern wheel on each side and parallel to the mid-rib and along the length of the mid-rib. For Experiment 2, one uninoculated plant in Environments 1 and 2 and one virus-inoculated plant in Environment 2 were wounded (W-nic; Fig 1C). Wounded and control plants (Nic in Fig. 1C) were harvested on Day 8, 4 d after wounding, when whole-plant nicotine concentrations typically attain their highest values (Baldwin and Schmelz 1996). In Experiment 1, necrotic tissue was removed from plants prior to harvest. Nicotine and JA analysis. Whole-plant and single-leaf nicotine levels were determined by HPLC analysis of methanol:water extracts of homogenized, freeze-dried plants, as described in Baldwin and Schmelz (1994). Jasmonic acid in fresh-frozen shoot
90
C.A. Preston et al.: TMV inoculation prevents wound-induced defenses
Fig. 1. A Schematic of the experimental design. Growth chambers housed six plants and had uni-directional ¯ow generated by a fan attached on the outside of each chamber. Two traps containing Super Q adsorbant were placed inside the chamber for the quanti®cation of volatile MeSA levels (see ®g.). B For quanti®cation of wound-induced JA, plants were damaged on two fully expanded leaves with a fabric-pattern wheel in a limited wound window. After 90 min, the wound window was harvested for JA analysis. As a control, a window from each of two fully expanded undamaged leaves on the same plant was harvested simultaneously. C To determine if volatilization of MeSA from TMV-inoculated tobacco plants could inhibit the wound-induced JA and nicotine responses of neighboring uninoculated tobacco plants, six plants, uninoculated/mockinoculated (clear circles) or virus-inoculated (shaded circles), were placed in the chambers to create eight replicate control environments without inoculated plants (Environment 1) and eight replicate environments with infected plants near uninoculated plants (Environment 2). To examine the eects of high levels of airborne MeSA, an MeSA-releasing device was placed inside the chamber near to the air entrance, upwind of six uninoculated tobacco plants (Environment 3). Plants analyzed for JA (JA in circle) were wounded and harvested according to B. Plants used for nicotine analysis were undamaged (Nic) or damaged (W-nic) with four revolutions of the pattern-tracing wheel on each side of the mid-rib on the ®ve oldest leaves. D The timeline for the experiment: Day 0, plants were inoculated and placed into the small growth chambers. Plants were left undisturbed in the chambers for 4 d while systemic increases of SA occurred in response to the TMV-inoculation (Enyedi et al. 1992; Shulaev et al. 1995). Day 4, plants for JA and nicotine analysis were wounded (W-JA, W-nic). Plants for JA analysis were harvested 90 min after wounding Day 8, plants for nicotine analysis were harvested tissue was measured by GC-MS with [1,2-13C]JA as an internal standard following methods described in Baldwin et al. (1997).
Results
Statistical analysis. Paired and unpaired t-tests, one-way and twoway analysis of variance (ANOVA), Fisher's LSD tests, and posthoc tests with Bonferroni adjustments of signi®cance levels were used to analyze nicotine and JA values; whole-plant and single-leaf nicotine concentrations were arcsine-transformed before analysis to normalize the data. All data were analyzed using the Statview 4.5 statistical package (Abacus Concepts, Inc., Berkeley, Calif., USA) and user-de®ned contrasts on nicotine values were performed using the MGLH module of the Systat statistical package (Evanston, Ill., USA).
Experiment 1. Can TMV-inoculated plants respond to MJ treatments? A one-way ANOVA among all treatments not treated with MJ (C-0, V-0, VW-0; Fig. 2A,B) determined that whole-plant nicotine concentrations, irrespective of the inclusion of tissues with virus lesions, were signi®cantly dierent among all treatments (F3,13 = 15.428 and F2,12 = 5.938, respectively; P < 0.017). Wounding of uninoculated plants (W-0) signi®cantly (P values
