R2R3-NaMYB8 regulates the accumulation of phenylpropanoid-polyamine conjugates which are essential for local and systemic defense against insect herbivores in Nicotiana attenuata

Although phenylpropanoid-polyamine conjugates (PPCs) occur ubiquitously in plants, their biological roles remain largely unexplored. The two major PPCs of Nicotiana attenuata plants, caffeoylputrescine (CP) and dicaffeoylspermidine (DCS), increase dramatically in local and systemic tissues after herbivore attack and simulations thereof. We identified Na MYB8, a homolog of Nt MYBJS1, which in BY2 cells regulates PPCs biosynthesis, and silenced its expression by RNAi in N. attenuata (ir-MYB8), to understand the ecological role(s) of PPCs. The regulatory role of NaMYB8 in PPCs biosynthesis was validated by a microarray analysis, which revealed that transcripts of several key biosynthetic genes in shikimate and polyamine metabolism accumulated in a NaMYB8-dependent manner. Wild-type (WT) N. attenuata plants typically contain high levels of PPCs in their reproductive tissues; however, NaMYB8-silenced plants that completely lacked CP and DCS showed no changes in reproductive parameters of the plants. In contrast a defensive role for PPCs was clear; both specialist ( Manduca sexta ) and generalist ( Spodoptera littoralis ) caterpillars feeding on systemically pre-induced young stem leaves performed significantly better on ir-MYB8 plants lacking PPCs compared to WT plants expressing high levels of PPCs. Moreover, the growth of M. sexta caterpillars was significantly reduced when neonates were fed ir-MYB8 leaves sprayed with synthetic CP, corroborating the role of PPCs as direct plant defense. The spatial-temporal accumulation and function of PPCs in N. attenuata are consistent with the predictions of the Optimal Defense Theory: plants preferentially protect their most fitness enhancing and vulnerable parts, young tissues and reproductive organs, to maximize their fitness. 250


INTRODUCTION
In nature, plants are frequently exposed to abiotic and biotic stress factors, including drought, extreme temperatures, high winds, UV-exposure, pathogens and herbivores. These selection pressures enabled plants to refine their constitutive and inducible defenses (Purrington, 2000;Zangerl, 2003;Howe and Jander, 2008;Walling, 2009). A shift from constitutive to inducible defense strategies can be considered a potential cost-saving mechanism whereby plants timely tune the production and accumulation of defenses with the need for the defenses, and thereby forgo the production and opportunity costs they incur when they are not needed, for example ultraviolet (UV) light induces secondary metabolite-flavonoid accumulation to intensify their UV-protective screen (Li et al., 1993;Zhao et al., 2007;Jenkins, 2009;Wang et al., 2009).
Once recognized via specific receptors, stress factors activate phytohormone signaling networks that trigger downstream defense responses in plants. Jasmonic acid (JA) is known to mediate wound and herbivore stress signals in plants that activate local and systemic defenses, and lead to the accumulation of antifeedants and/or ovipositioning deterrents against herbivores. These toxins largely impair insect growth and reduce their survivorship rates, helping plants to diminish further damage (Steppuhn and Baldwin, 2007;Chen, 2008). In a similar manner, salicylic acid (SA) coordinates the elicitation of defenses against invading pathogens, resulting in the accumulation of phytoalexins that limit the spread of pathogens in the plant tissues, but an increasing amount of evidence points to antagonism between SA and JA signaling (Loake and Grant, 2007;Diezel et al., 2009). Therefore, defense-related hormones in plants are engaged in a complex cross-talk that still needs to be fully examined (Koornneef and Pieterse, 2008).
After herbivore attack, the wounds in plants often come in direct contact with herbivores' oral secretions (OS), a potential carrier of herbivore-specific elicitors, which can be recognized by plant cells. Upon perception of these elicitors --e.g. fatty acid-amino acid conjugates (FACs), inceptins or caeliferins --plants trigger herbivore-specific defense responses, which involve large-scale transcriptional, translational and post-translational changes in plants at both local and systemic levels (Halitschke et al., 2001;Gatehouse, 2002;Howe and Jander, 2008). In N. attenuata NaMYB8 could play an important role in plant-herbivore interactions, possibly by regulating PPC levels in response to herbivory.
Following the initial characterization of NaMYB8 expression and the accumulation of CP in N. attenuata, we used NaMYB8 transcription factor as a genetic tool to examine the ecological relevance of OS-induced PPCs accumulation, and significance of PPCs presence in N. attenuata. The plants specifically silenced in expression of NaMYB8 gene were generated through RNAi technology. As expected, the inverted repeat NaMYB8-silenced (referred to as ir-MYB8) plants were unable to accumulate two major PPCs (CP and DCS) in their tissues. A microarray study conducted with ir-MYB8 and WT plants revealed that NaMYB8 protein is required for transcriptional activation of genes involved in phenylpropanoid and polyamine biosynthesis, in addition to several other genes with unknown functions. As both specialist (M. sexta) and generalist (Spodoptera littoralis) caterpillars performed better on ir-MYB8 plants compared to the WT plants, we propose that PPCs and their regulation by NaMYB8 are vital part of direct defense mechanisms used by plants against attacking herbivores.

Isolation and expression of NtMYBJS1 homolog in N. attenuata
The transcripts of NtMYBJS1 gene accumulated 3h after treatment of tobacco BY-2 cell cultures with methyl jasmonate (MJ; Gális et al., 2006). We therefore used a 1h W+OS-elicited cDNA pool --a time point known to be associated with accumulation of many JA-responsive genes --to clone a homolog of NtMYBJS1 from N. attenuata. The cloned sequence, which was designated NaMYB8 in this study (Supplemental Fig . S1A; GenBank Acc. GU451752), occurs as a single copy gene in N. attenuata's genome (Supplemental Fig. S2C).
We then examined transcriptional response of NaMYB8 gene to wounding and herbivore cues: a fully-expanded rosette leaf was wounded with a pattern wheel and either water (W+W) --mimicking mechanical wounding --or M. sexta's OS (W+OS) --representing simulated herbivory --were applied to the wounds. Biotic stresses are known to trigger transcriptional responses that typically follow one of the expression profiles: (i) rapid and transient up-regulation of transcripts encoding primary regulators, and (ii) steadily increasing but long-lasting transcriptional responses of genes encoding defensive metabolites [e.g. trypsin proteinase inhibitors (TPIs)] or enzymes involved in their biosynthesis (Zhao et al., 2005). Interestingly, NaMYB8 7 gene showed a mixed pattern of its transcript regulation in the local leaves, including rapid transient increase of transcripts 45 min after W+W-and W+OS-elicitations (Fig. 1A,inset), followed by stably elevated levels of NaMYB8 transcripts in the W+OS-elicited leaves for additional 1-2 days (Fig. 1A). Because NaMYB8 transcripts responded differentially to the presence of M. sexta's OS, we assumed that it could play an important role in the inducible plant defenses employed by plants against attacking herbivores. The NaMYB8 transcript levels in unwounded control leaves did not change significantly over 2 days (0-50h; Fig. 1A), excluding the possibility that NaMYB8 transcript accumulation is controlled by circadian rhythm.

Transient and stable silencing of NaMYB8 expression in planta
In order to confirm the role of NaMYB8 transcription factor in plant-insect

Transcriptional targets of NaMYB8
In order to correlate the observed metabolic changes in NaMYB8-silenced lines with the NaMYB8 transcript accumulation in N. attenuata, we used a custom oligonucleotide microarray (Biochip ver. 4), spotted specifically with herbivory-  Table S1). These genes predominantly included sequences from phenylpropanoid metabolism, such as phenylalanine ammonia lyase (PAL) and 4-coumaroyl-CoA ligase (4CL), and genes encoding enzymes involved in the synthesis of polyamines (Supplemental Table S1); we assume that these genes represent direct targets of NaMYB8 transcriptional activity in N. attenuata. The transcripts of the key biosynthetic gene involved in putrescine biosynthesis, ornithine decarboxylase (ODC), were also less abundant in ir-MYB8 leaves, but only at 1.8-fold-change level, and therefore below our arbitrarily-selected www.plantphysiol.org on August 13, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved. threshold (2-fold). In contrast, NaMYB8 silencing significantly influenced the accumulation of spermidine synthase transcripts (~3.9-fold reduction) which are required for DCS biosynthesis from spermidine.
Only few genes were actually up-regulated in NaMYB8-silenced plants compared to the WT plants (Supplemental Table S1), demonstrating that NaMYB8 generally functions as positive transcriptional regulator in N. attenuata.

Young systemic leaves accumulate high levels of CP after simulated herbivory
Resource allocation and spatial-temporal accumulation of defense-related metabolites varies within plant tissues and often reflects the degree and frequency of stresses that plants have to face during their development (Boege and Marquis, 2005;Boege et al., 2007). We therefore analyzed NaMYB8-dependent CP accumulation in plant ontogeny, namely at rosette, early elongated, elongated, flowering and mature stages of the development, using 3 day W+OS-elicited (stressed) and unelicited (control) WT plants. Even though CP accumulation followed a complex developmental pattern (Fig. 3), several general trends could be recognized. For example, the high constitutive levels of CP in the vegetative tissues at rosette and early elongated stages clearly shifted towards reproductive tissues after flowering and capsule development. In the mature plants, almost no CP could be detected in the leaves. Interestingly, while CP levels always increased in the local leaves following W+OS elicitation, CP accumulated even more in the systemically induced young stem leaves of these plants, even at flowering stage when it could be barely detected in the vegetative plant parts (Fig. 3).
We examined the transcript profile of NaMYB8 in WT plants at an elongated stage of development to address whether the systemic accumulation of CP could be due to metabolite mobilization to the systemic leaves or the transmission of systemic signal from the locally W+OS-induced leaves to systemic ones, which would require NaMYB8 to mediate the up-regulation of downstream biosynthetic genes. The accumulation of CP coarsely correlated with the NaMYB8 transcript abundances in both control and OS-elicited leaves (Fig. 3, inset), suggesting that CP accumulation in the distal leaves is most likely subject to systemic activation of NaMYB8 gene expression and hence, transcriptional activity of the NaMYB8 protein.
As both local and systemic accumulation of CP and DCS was completely abolished in ir-MYB8 plants (Fig. 5A), we propose that NaMYB8 gene serves as an universal master regulator for CP and DCS biosynthesis in N. attenuata plants.
However, in contrast to a strong inducible character of CP accumulation in the leaves ( Fig. 5A), no significant increase in DCS accumulation was observed in the systemically induced young stem leaves of WT plants (Fig. 5A), while locally W+OS-induced leaves still showed a small but significant increase in DCS levels ( Fig. 5A). In addition, the constitutive levels of DCS were usually higher compared to the constitutive levels of CP. This suggests that other cis-acting regulatory elements most probably contribute to the regulation of CP and DCS-synthase genes, which additionally modifies the rate-limiting NaMYB8 transcriptional activity.
Alternatively, a variation in substrate availability of putrescine and spermidine could be contributing to the differential spatial-temporal accumulation of CP and DCS in N. attenuata plants (Paschalidis and Roubelakis-Angelakis, 2005).

Examination of putative role of CP and DCS in plant development
The preferential accumulation of CP in the young leaves and reproductive tissues of N. attenuata suggested that CP might play multiple defensive and/or developmental roles in N. attenuata. Several previous reports have associated high CP levels with flower initiation and bud tissue development in tobacco (Martin-Tanguy, 1985;Martin-Tanguy, 1997;Balint et al., 1987), indirectly proposing a role of CP in flower development. Moreover, the accumulation of phenylpropanoids, polyamines and their conjugates in the reproductive tissues has been previously associated with pollen fertility and floral initiation in plants (Wada et al., 1994;Imai et al., 2004;Kasukabe et al., 2004;Paschalidis and Roubelakis-Angelakis, 2005;Fellenberg et al., 2008Fellenberg et al., , 2009Grienenberger et al., 2009;Matsuno et al., 2009). P ir-MYB8-818 = 0.510; P ir-MYB8-810 = 0.999). We further quantified the seed mass in two seed capsules located on uppermost lateral branch, one located nearest (T0) and one farthest (T1) from the branching point (see Fig. 4E for details); the seed mass was not statistically different between ir-MYB8 and WT plants ( We next focused on an alternative hypothesis that CP and DCS could be involved in direct defenses targeted against attacking herbivores, accounting for the high levels of PPC accumulation in reproductive plant parts.

Lack of CP and DCS accumulation makes plants susceptible to herbivores
To test whether CP accumulation is part of plant defense activated against attacking herbivores, we assessed the performance of M. sexta (N. attenuata specialist) and S. littoralis (N. attenuata generalist) caterpillars on WT and NaMYB8silenced plants. It is known that the elicitors present in herbivore's OS --for instance, the FAC N-linolenoyl-L-glutamic acid (C18:3-Glu) --are responsible for activation of defense mechanisms targeted against herbivores in N. attenuata (Halitschke et al., 2003;Giri et al., 2006); C18:3-Glu elicitor was found predominantly in OS of both N. attenuata generalist and specialist herbivores (Diezel et al., 2009). We therefore analyzed the CP and DCS accumulation in both W+C18:3-Glu-induced local rosette and systemically induced young stem leaves of N. attenuata. This treatment enhanced both local and systemic accumulation of CP in the plants, while the levels of DCS did not changed significantly (Fig. 5B). Because the overall pattern of CP accumulation in W+FAC-treated leaves was comparable to W+OS-elicited leaves, we therefore used the standardized FAC treatment to maximize the accumulation of CP in the systemically induced young leaves in following herbivory bioassays (see Fig. 3 and related paragraph).
To precondition the plants, we treated rosette leaves of early elongated WT and ir-MYB8-810 N. attenuata, one at the time, with W+C18:3-Glu, 1-2 days before placing the caterpillars onto the young, systemically pre-induced leaves. This experimental setup was specifically designed to mimic the initial feeding of the To conduct the actual bioassays, freshly hatched M. sexta (specialist herbivore) neonates were placed directly on to the FAC pre-induced stem leaves. Due to high sensitivity of S. littoralis (generalist herbivore) to N. attenuata's defenses, hatched neonates had to be first pre-reared on artificial diet for 6 days. To eliminate the effect of artificial diet present in caterpillar's gut, a short feeding on WT or ir-MYB8 leaves was then carried out with the larvae and subsequently, pre-weighed caterpillars were transferred to WT and ir-MYB8 young stem leaves, and their body mass gain was recorded every day. Both M. sexta (unpaired t test, P = 0.004) and S. littoralis (unpaired t test, P < 0.001) caterpillars performed better on ir-MYB8-810 plants compared to the WT plants (Fig. 5C). A similar trend in caterpillar performance was previously observed in M. sexta neonates that were allowed to feed directly on NaMYB8-VIGS and EV-VIGS plants (Supplemental Fig. S5, Repeatedmeasures ANOVA; F 1, 28 = 6.265; P = 0.018). These results provided strong evidence that PPCs function as indispensable defensive metabolites targeted against leafchewing herbivores in N. attenuata plants.

Exogenous application of synthetic CP impairs growth of M. sexta caterpillars
We further tested the role of CP as plant specific-defensive metabolite against herbivores by spraying physiologically relevant concentration of synthetic CP on ir-MYB8 leaves, known to be deficient in accumulation of CP and DCS. We first examined the turnover and stability of the exogenously applied CP on the leaf surface, and found that CP sprayed on ir-MYB8-810 leaves remained stable for at least two days after application. This shows that CP was neither degraded nor it was mobilized to other plant parts. After the spray, we recovered about 130 μg CP per g fresh mass (FM), which was just below the CP concentrations accumulated in the leaves after W+OS-elicitation (~170 μg CP per g FM; Fig. 6A). In the control treatment, no CP was found in the water-sprayed leaves of ir-MYB8-810 plants ( 14 6A). We then clip-caged two neonates on CP-or water-sprayed ir-MYB8-810 leaves and examined the growth of the caterpillars. The caterpillars fed for 4 days on CPsprayed leaves showed less mass gain compared to caterpillars fed on water-sprayed (control) ir-MYB8-810 leaves ( Fig. 6B; unpaired t test, P = 0.011), further highlighting the role of CP as direct defense metabolite in response to herbivory in N. attenuata plants.
Previously, the accumulation of several PPCs in N. attenuata leaves in response to herbivory has been reported (Kessler and Baldwin, 2004;Paschold et al., 2007) but no conclusive experimental evidence exists for the role of PPCs in plant-herbivore interactions. Here, we show that CP and DCS accumulation in N. attenuata, regulated by NaMYB8 transcription factor, plays an important role in plant defense against leaf-chewing herbivores.

JA signaling is required for CP accumulation
The accumulation of CP was previously shown to be strongly induced by further examined in ir-COI1 N. attenuata plants impaired in JA-Ile perception due to non-functional SCF coi1 protein complex. ir-COI1 plants showed strongly reduced CP levels (Paschold et al., 2007), similar to tomato jai-1 mutants defective in JA perception, which failed to accumulate CP in the flowers and in the MeJA-treated leaves (Chen et al., 2006). All together, these observations provide a strong link between JA signaling and CP accumulation; however, the actual regulatory mechanisms initiated after JA perception, particularly those downstream of putative MYC2 protein regulation in N. attenuata, leading to PPCs' biosynthesis in plants, such as PAL (phenylpropanoid biosynthesis) or ODC/ADC (polyamines) remain elusive.
The direct connection between MYB transcriptional activity and the regulation of PPC biosynthesis was first reported by Gális et al. (2006) and Shinya et al. (2007). Ectopically expressed R2R3-MYB transcription factors --NtMYBJS1 spectrum of elicitors that could induce various MYB transcription factors, and enhance the accumulation of PPCs, suggests that these metabolites could be mediating a broad range of resistances to fungi, necrotrophic pathogens and herbivores.

NaMYB8: OS-responsive regulator of CP and DCS accumulation
NaMYB8 transcripts accumulated rapidly in N. attenuata rosette leaves after wounding and returned to the pre-elicited levels within three hours after treatment ( Fig. 1A, inset); however, the M. sexta's OS-elicited rosette leaves showed delayed reinstatement of transcripts to their basal levels at late hours after induction, reflecting that NaMYB8 expression discriminates between herbivory-associated damage and simple mechanical wounding. Similar trends in transcript accumulation are typically found in genes involved in defense triggered by M. sexta's OSelicitation, for instance, genes encoding TPIs (Halitschke et al., 2001). Previously, we showed that NaMYB8 gene is also induced by UV-B in the glasshouse and by cumulative stress conditions in the natural environment of N. attenuata, both following an RdR2-dependent induction pattern (Pandey et al. 2008). It suggests that apart from regulating N. attenuata's response to herbivores, the NaMYB8 gene may also be regulated by various abiotic stresses, namely high levels of UV-B in the natural environment.
We examined the transcription regulatory activity of NaMYB8 by performing microarray analysis with OS-elicited ir-MYB8-818 leaves hybridized against identically elicited WT leaves. This analysis confirmed that NaMYB8 gene specifically activates transcription of genes involved in phenylpropanoid and polyamine biosynthesis (Supplemental Table S1). Similar transcriptional regulatory pattern of genes involved in phenylpropanoid pathway was observed in BY-2 tobacco cell cultures ectopically expressing NtMYBJS1 gene, using an independent tobacco microarray system (Gális et al., 2006). Interestingly, several novel NaMYB8controlled genes with a potential role in PPC biosynthesis have been identified in the current microarray experiment, and the functional characterization of these genes is currently in progress. Remarkably, one of the NaMYB8-regulated genes encodes a functional DCS synthase in N. attenuata (N. Onkokesung; manuscript in preparation). proof that NaMYB8 transcription factor is essential for CP and DCS biosynthesis and accumulation in N. attenuata plants (Fig. 7), similar to the role of PAP1 gene, an AtMYB75 transcriptional regulator that governs anthocyanin biosynthesis in A.

CP accumulates in the young vegetative and reproductive tissues
The ecological role of metabolites can be often deduced from their spatialtemporal accumulation pattern. CP and DCS accumulated in the shoot apices, young leaves, and female reproductive organs during flower induction and development in tobacco (Cabanne et al., 1981;Martin-Tanguy, 1985;Edreva et al., 2007), and a similar pattern of PPC accumulation was also observed in some Araceae species (Ponchet et al., 1982). The accumulation of CP in reproductive organs is widely reported in the literature, which is generally linked to plant development. However, when the distribution of PPCs was examined across seven species belonging to In the search for ecologically meaningful interpretations of CP accumulation patterns, we decided to treat the plants with simulated herbivory (Halitschke et al., 2001) at five developmental stages and compared the CP accumulation with untreated plants. Interestingly, both constitutive and inducible levels of CP were higher in the young stem leaves compared to the mature rosette leaves (Fig. 3, 5A). In agreement with previous studies (Edreva et al., 2007), CP was most abundant in the buds at the early elongated stage but these levels progressively declined as plants aged (Fig. 3).
At mature stage, plants retained high levels of CP almost exclusively in the reproductive organs, flowers and capsules. However, when ectopically silencing the ability of plants to accumulate PPCs, we found no aberrant morphological changes associated with this novel trait. In summary, the CP accumulation shifted primarily from photosynthetically active young leaves during vegetative growth to flower buds, flowers and seed capsules at maturity, as well as it was significantly stimulated by simulated herbivory treatments (Fig. 3).

CP and DSC are indispensable for plant defense against herbivores
In our follow up hypothesis, CP could be a direct defensive metabolite that accumulates upon herbivory and against herbivores. According to the Optimal Defense Theory, plants tend to allocate more defense-associated metabolites to the valuable plant parts during development --photosynthetically active tissues, meristems and reproductive tissues (flowers and seeds) --to protect these organs from stress factors, including herbivores (McKey, 1974;Ohnmeiss and Baldwin, 2000;Stamp, 2003). A shift in CP accumulation, consistent with the theory, was indeed Further studies with ir-MYB8 plants will enable us to understand the role of CP and DCS in defense against pathogens, as well as the results from our microarray analysis shall be used to identify novel genes responsible for CP and DCS biosynthesis.

Plant growth conditions in the glasshouse
Nicotiana attenuata Torr. Ex S. Watson (22 nd inbred generation) seeds, originally collected from a native population from a field site located in Utah, USA, were used for all described experiments, including transformation and generation of transgenic lines. The seeds were germinated on sterile Gamborg B5 medium (Sigma, http://www.sigmaaldrich.com) after 1h treatment with diluted smoke (House of Herbs) and 1μM GA3 (www.carl-roth.de). Ten days after germination, seedlings were transferred into Teku pots containing peat-based substrate and after additional 10-12 days, the plantlets were transplanted into individual 1L pots with the same substrate. In the glasshouse, plants were grown at 24-26°C, relative humidity ~ 55%, and supplemented with light from 400-and 600-W sodium lamps (Philips Sun-T Agro; http://www.nam.lighting.philips.com) for 16h.

Virus induced gene silencing (VIGS)
A VIGS system based on the tobacco mosaic rattle virus was used as described in Saedler and Baldwin (2004). Three-week-old N. attenuata plants were inoculated with Agrobacterium tumefaciens and pTVMYB8 plasmid --carrying a fragment of NaMYB8 (NaMYB8-VIGS) or pTV00 --carrying empty vector construct as control (EV-VIGS).

Generation and characterization of ir-MYB8 transgenic lines
Stably transformed ir-MYB8 lines were generated by introducing an ir construct containing a 309-bp NaMYB8 gene fragment and a gene for hygromycin resistance (hptII) as a screening marker in a pSOL8 transformation vector (Supplemental Figure S2A) into N. attenuata plants as described by Bubner et al. (2006). A. tumefaciens-mediated plant transformation was followed as described by Krügel et al. (2002). Transformed lines, each containing a single insertion of the hptII marker gene as determined by Southern hybridization, were further screened by segregation analysis of T2 seedlings for their hygromycin resistance to obtain homozygous transformed lines. Quantitative real-time PCR (RT-qPCR) was used to quantify the transcript accumulation of NaMYB8 gene and two independently transformed homozygous diploid lines with efficiently silenced expression of NaMYB8 gene, ir-MYB8-810 and ir-MYB8-818, were selected for all subsequent experiments.

CCGGATCGGACGATTGCG-5') were PCR-amplified and used as probe for
Southern hybridizations to confirm the single-insertion of hptII gene in ir-MYB8 transformed lines. The DNA probes were labeled with α -32 P using Rediprime TM II DNA labeling system (Amersham Biosciences; http://www.amershambiosciences.com).

Expression analysis by RT-qPCR
To analyze NaMYB8 gene expression, total RNA was extracted from fully A 2 µg of total RNA was reverse-transcribed using oligo-dT (Fermentas; http://www.fermentas.com) as primer and Superscript II reverse transcriptase (Invitrogen). All RT-qPCR assays were performed with cDNA corresponding to 100 ng RNA before reverse transcription and gene-specific primers using qPCR core kit for SYBR Green I (Eurogentec GmbH; http://www.eurogentec.com), following the manufacturer's instructions.
To determine the NaMYB8 genes transcript levels in the ir-MYB8 lines, genespecific primers were designed outside the region used for making ir silencing construct. All gene-specific primers were designed with Primer3 online available software (http://frodo.wi.mit.edu/primer3). For all RT-qPCR analyses, unless stated differently, cDNA from five replicate biological samples was used and an assay was carried out on a Stratagene Mx3005P TM real-time PCR system (http://www.stratagene.com). Relative gene expression was calculated using a 10-fold dilution series of cDNA containing NaMYB8 as well as the elongation factor-1α house keeping gene from N. tabacum (EF1-α; Acc. D63396) as an endogenous reference.
To determine whether herbivore attack elicits NaMYB8 transcription, 5 fully

Secondary metabolite analysis
The secondary metabolite analysis was carried out on (+1) leaves after M. sexta caterpillar feeding on WT/ir-MYB8 and EV/NaMYB8-VIGS plants for four days, using high-performance liquid chromatography (HPLC) as described by Keinänen et al. (2001). MeOH prepared with 0.5% acetic acid water. The sample in FastPrep tubes was homogenized on a FastPrep homogenizer (Thermo Electron) for 45 sec and then centrifuged for 12 min at 13000 rpm. The supernatant was transferred into 1.5-mL Eppendorf tubes, centrifuged, and finally transferred to a glass vial where it was analyzed by an Agilent-HPLC 1100 series (http://www.chem.agilent.com). The ODS Inertsil C-18 column (3 µm, 150 x 4.6 mm i.d.) was attached to a Phenomenex Security Guard C18 pre-column (http://www.phenomenex.com). The solvents were (A) 0.25% H3PO4 in water and (B) acetonitrile. The elution system was as follows: 0-6 min, 0-12% of B; 6-10 min, 12-18% of B; 10-30 min, 18-58% of B. The flow rate was 1 mL/min, the injection volume was 10 µL, and the column oven was set at 24 °C. The nicotine eluted at retention time 1.83 min was detected at 254nm; CP, CGA, DCS eluted at retention times 8.3 min, 12.3 min, 12.7 min respectively were detected at 320nm; rutin eluted at 16.4 min was detected at 360 nm and diterpene glycosides eluting at 24-26 min were detected at 210nm.

Measuring ethylene accumulation
At least three replicate measurements were used to quantify ethylene production in WT and NaMYB8 transgenic N. attenuata plants. Three leaves were www.plantphysiol.org on August 13, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved. treated either with W+W or W+OS, in the case of EV/NaMYB8-VIGS; or W+OS in the case of WT/ir-MYB8-818/ir-MYB8-810; or plants were left untreated (controls).
Leaves were cut from stems, immediately sealed in a three-neck 250-mL round bottom flasks and kept in the glasshouse under light conditions for 5 h. The headspace of the flasks was flushed into a photoacoustic laser spectrometer with hydrocarbonfree clean air, and the ethylene concentration was quantified by comparing ethylene peak areas with peak areas generated by a standard ethylene gas as previously described by Von Dahl et al. (2007).

Herbivore performance
The growth performance of M. sexta (N. attenuata specialist) and Spodoptera littoralis (N. attenuata generalist) caterpillars was examined using WT and ir-MYB8silenced plants. Due to the high sensitivity of the generalist herbivore to N. attenuata defenses, S. littoralis neonates were first reared on artificial diet for 6 days and then placed on WT or ir-MYB8 leaves for one day to get rid of artificial diet present in the caterpillars' guts, and subsequently, the pre-weighed S. littoralis caterpillars were placed on the plants. Freshly hatched N. attenuata specialist M. sexta neonates, were placed directly on WT/ir-MYB8-810 stem leaves. Both generalist and specialist herbivores were allowed to feed only on systemically pre-induced stem leaves of plants, which had their single rosette leaf elicited with FAC (18:3-Glu; 0.07 nmol/μL in 0.02% (v/v) Tween-20/water; 20 μ L per leaf) every 4 th day from the start of the experiment. S. littoralis caterpillar mass was recorded daily over 5 days, while the M. sexta caterpillar mass was recorded on 4, 7 and 11 day of feeding.

CP synthesis
CP was synthesized in the laboratory as described in (Hu and Hesse, 1996).

STATISTICAL ANALYSIS
All statistical analyses were performed with SPSS software (http://www.spss.com).

SUPPLEMENTAL MATERIALS LEGENDS Supplemental Figure S1
Deduced NaMYB8 protein sequence aligned with its homolog from N. tabacum (NtMYBJS1).
Sequence of NaMYB8 gene was originally obtained using PCR and cDNA template from N. attenuata, and primers designed according to NtMYBJS1 coding sequence.
3'-end of NaMYB8 gene was obtained by 3'RACE, and sequence was finally verified M. sexta neonates were placed on the stem leaf of either NaMYB8-VIGS or EV-VIGS plants and were allowed to feed without restricting the movement of caterpillars. Initial M. sexta's caterpillars mass was recorded on 4th day of feeding, followed by recording mass every second day of the experiment (P < 0.05 (*); n = 15).

Supplemental Table S1
Silencing NaMYB8 suppresses the accumulation of CP and DCS by down-regulating the expression of genes involved in phenylpropanoid and polyamine metabolism.        Silencing NaMYB8 in N. attenuata makes plants vulnerable to insect herbivores.
(A) Mean (± SE) of CP and DCS accumulation in locally W+OS-induced rosette leaves and systemically induced young stem leaves in N. attenuata analyzed 3 days after induction (n = 4); control plants remained unelicited. Neither CP nor DCS were detected (nd) in ir-MYB8 plants. (B) A pattern wheel-wounded rosette leaf of WT plants was treated either with OS-specific elicitor C18:3-glutamic acid dissolved in 0.02% tween or was treated with 0.02% tween (mock treatment), and both local and systemic accumulation of CP and DCS were analyzed as above. (C) Mean (± SE) mass gained by N. attenuata generalist herbivore (S. littoralis) and specialist herbivore (M sexta) caterpillars when fed on systemically preinduced young stem leaves of ir-MYB8-810 and WT plants, whose rosette leaves were elicited with C18:3-glutamic acid (FAC). A single rosette leaf was elicited every 4th day to enhance the CP accumulation in the young stem leaves. Asterisks represent significantly different growth responses of herbivores that fed WT and homozygous ir-MYB8-810 plants at specific time points at P < 0.01 (**) and P < 0.001 (***; n = 16; FM, fresh mass).
www.plantphysiol.org on August 13, 2017 -Published by Downloaded from Copyright © 2010 American Society of Plant Biologists. All rights reserved.