Sterilized tobacco (N. tabacum cv. Wisconsin 38) seeds were germinated and grown to seedlings under continuous illumination on half-strength Gamborg B5 medium solidified with 2% (w/v) gellan gum and supplemented with 0.3% sucrose at 24°C. Two-week-old plants were transferred to Perlite saturated with half-strength Gamborg B5 medium and grown for another 2 weeks in the greenhouse at 24°C before HT treatment. For HT treatment, the 4-week-old seedlings were placed in a plant growth chamber at 32°C for the indicated time. The plants were placed in a growth chamber at room temperature for the same amount of time as the control. For MeJA treatment, MeJA at different concentrations was sprayed on leaves of the 4-week-old tobacco plants. For saline or heavy-metal treatment, 100 mM NaCl or 30 μM CdCl₂, respectively, was used to water the 4-week-old tobacco seedlings for the indicated times. For wounding treatment, the leaves were wounded with a pattern wheel. After each treatment, the tobacco roots were immediately collected for further molecular analysis and alkaloid measurement. The roots were frozen immediately in liquid nitrogen for later analysis.

A 0.5-g sample from each of the collected tobacco roots was collected and frozen in liquid nitrogen. The frozen samples were lyophilized and then homogenized in 4 ml of 0.1 M H₂SO₄. The homogenate was sonicated for 60 min and centrifuged at 2000 g for 15 min. The resulting supernatant was neutralized by adding 0.4 ml 25% NH₄OH. The mixture was loaded onto an Extrelut-1 column and eluted with 6 ml of chloroform. The eluent was dried at 37°C, and each residue was dissolved in ethanol and analyzed by gas chromatography–mass spectrometry (GC/MS) using a split sampling mode as described in Goossens et al. (2003). The column temperature was held at 100°C for 10 min and then increased to 260°C during a 35-min period, at a gradient of 8°C/min. Signal output was simultaneously monitored for two separate ion pairs for nicotine (m/z 163.2/105.9 and m/z 163.2/80.1) and single ion pairs for anabasine (m/z 162.9/80.1), nornicotine (m/z 149.1/80.1), and nicotine-d3 (m/z 166.3/87.2) during the last 4 min of sample elution. Total elution time monitored was 7.5 min. The stable heavy isotope-labeled Nicotine-d3 (CIJ, MA, US) was used as the internal standard.

To construct binary vectors overexpressing NtMYC2a, total RNA was isolated from tobacco using an RNeasy Mini kit (Qiagen). RNAs were converted to first-strand cDNAs by SuperScriptII Reverse Transcriptase (Invitrogen) with an oligo(dT) primer (Zhang et al., 2012). The coding region of NtMYC2a was amplified using the MYC2a-F and MYC2a-R primers. All primer sequences are provided in Supplementary Table 1. The pOE-6HA binary vector was derived from pGREEN vector with the 35S promoter in the KpnI/XhoI site, and NOS terminator in the NotI/SacI site, and the 6xHA tag fragment were cloned into the XbaI/NotI site. The pOE-6HA vector were digested with BamHI/EcoRI, and the NtMYC2a fragment was inserted into the BamHI/EcoRI site under the control of the 35S promoter using In-fusion Recombination Enzyme (Clontech) to generate the binary pOE-NtMYC2a-6HA construct. The pRI101-GFP vector derived from the p RI101-ON binary vector (Clonetech), and GFP fragment were inserted into the NdeI/EcoRI site. Similarly, the DNA fragment of NtJAZ1 was amplified using the corresponding primers (Supplementary Table 1) and inserted into the EcoRI/HindIII site of the modified pRI101-GFP vector under the control of 35S to generate the pRI101-NtJAZ1-GFP binary construct. To generate the NtJAZ1ΔC fragment, which lacks the JAS domain (containing PIARRNSLTRFLEKRKDRITSTAPYQI) between 181 and 207 site of NtJAZ1, overlapping primers that span the JAS domain were used (see Supplementary Table 1 for the detail primer sequences) to amplify the NtJAZ1 template to produce NtJAZ1ΔC using a QuickChange Site-Directed Mutagenesis kit (Catalog #200521, Stratagene). The NtJAZ1ΔC fragment was then inserted into the EcoRI/HindIII site of the modified pRI101-GFP vector using In-Fusion recombination enzyme (Clontech) to generate the binary pRI101-NtJAZ1ΔC-GFP construct. The constructed binary vector was then transformed into the Agrobacterium tumefaciens strain GV3101. To generate transgenic tobacco, tobacco leaf discs were infected with A. tumefaciens strain GV3101 harboring a binary vector, following a previous method (An, 1985), and the transgenic tobacco seedlings were screened on the selective medium containing 100 mg/L kanamycin sulfate), shoots that were ≥5 cm high were transferred to peat compost and grown to maturity in the greenhouse. Primary transformation lines were propagated up to the T₂ generation by self-pollination.

The pGREENII0800-LUC vector were derived from pGREEN vector, and is widely used in the analysis of promoter transcription factor interactions, such vector contained the LUC reporter marker in the NcoI/XbaI site, and a REN gene under the control of a 35S promoter in the BglI/KpnI site for estimating the extent of transient expression (Hellens et al., 2005). To perform a transient transcriptional activity assay using the luciferase reporter, the 755-bp NtPMT1a promoter was amplified from genomic DNA and inserted into pGREENII0800-LUC to generate the pNtPMT1a: LUC reporter construct. To generate the mutated pNtPMT1a: LUC reporter construct, the mutated site in the promoter region of NtPMT1a was generated using the QuickChange Site-Directed Mutagenesis kit. The coding region of NtMYC2a was amplified and inserted into the pGREEN62-SK vector under the control of the 35S promoter to generate the effector construct. Root tissue was collected from 4-week-old tobacco plants. Root cell protoplasts were isolated, and polyethylene glycol (PEG)–mediated transient transformation was performed as described (Sheen, 2001). In brief, for each transformation, 10 μg reporter construct and 10 μg effector construct were introduced into 200 μL of protoplast solution (~2 × 10⁶ cells). Transformed protoplasts were resuspended in 4 ml liquid MS medium with 0.4 M sucrose and incubated at 24°C in darkness overnight or subjected to heat shock at 32°C for 30 min and then incubated at 24°C in darkness overnight. We firstly obtained the protoplast coexpressing the pNtPMT1a: LUC reporter construct and NtMYC2a or empty effector constructs by PEG-mediated transformation, and then co-transformed the JAZ1 or JAZ1ΔC effector construct into the protoplast with pNtPMT1a: LUC/NtMYC2a to obtained the protoplast with triple constructs (pNtPMT1a: LUC/NtMYC2a/JAZ1 or pNtPMT1a: LUC/NtMYC2a/JAZ1ΔC). For the luciferase activity assay, samples were lysed in a buffer containing 100 mM potassium phosphate (pH 7.5) and 1 mM DTT and were cleared by centrifugation at 12,000 g for 5 min. Then, 15 μL of sample was used to measure the luciferase and Renilla luciferase (REN) activities with the dual-luciferase reporter assay system (Promega). Data are presented as the ratio of luciferase activity relative to the control REN activity.

Proteins were isolated with extraction buffer containing 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1 mM PMSF, 1 × Complete Protease Inhibitor Cocktail (Roche), 5% glycerol, 1 mM EDTA, and 1 mM DTT. The protein concentration was determined with the Bradford assay (Bio-Rad). The samples were mixed with 2 × SDS sample buffer and boiled for 3 min and then separated on a 10% SDS-polyacrylamide gel. The membrane transfer and protein gel blot assays were performed as described (Wang et al., 2009; Zhang et al., 2013). Subsequent immunoblotting assays were performed with anti-GFP (Clontech) and anti-Actin (Agrisera) at dilutions of 1:3000 and 1:2000, respectively.

ChIP experiments were performed as described (Ouyang et al., 2011) using root tissue from 4-week-old transgenic seedlings that had been subjected to 24 h of HT treatment at 32°C or to control conditions. In brief, 5 g of root tissue was cross-linked in 1X PBS buffer (25 mL) with 1% formaldehyde on ice under vacuum; the vacuum was released after 15 min and reapplied for another 14 min. Fixation was stopped by adding glycine to a final concentration of 0.125 M, and the fixed sample was washed three times with MC buffer and ground to a powder in liquid nitrogen, followed by nuclear isolation by adding the frozen powder to 30 mL of M1 buffer. The homogenate was filtered through four layers of Miracloth prior to nuclei isolation. Nuclear-enriched extracts were resuspended in 5 mL lysis buffer (50 mM HEPES, pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% deoxycholate, 0.1% SDS, 1 mM phenylmethylsulfonyl fluoride, and 10 mM sodium butyrate) containing a plant proteinase inhibitor cocktail (Sigma-Aldrich), followed by sonication for 5 min on medium power in 1.5 mL sonic buffer using a Bioruptor UCD-200 (Diangenode) and centrifugation at 500 g to remove starch granules. The chromatin solution was sonicated for 5 min on medium power five times to create ~300-bp average fragment sizes, as estimated by agarose gel electrophoresis. Antibody against HA and the IgG (Sigma-Aldrich) control were used for immunoprecipitation. The precipitated DNA was recovered using a QIAquick PCR purification kit (Qiagen) and analyzed by real-time quantitative PCR using the appropriate DNA primers (Supplementary Table 1) and SYBR Green Real-Time PCR Master Mix (Applied Biosystems).

The transgenic NtJAZ1–green fluorescent protein (GFP) plants were treated with or without HT for the indicated periods, and the GFP fluorescence in the root tip was observed under a confocal laser scanning microscope (Olympus). GFP excitation was produced with the 488-nm line of an argon ion laser, and emission was detected between 505 and 530 nm (Tsien and Waggoner, 1995).

JA and JA-Ile were quantified using a high-performance liquid chromatography (HPLC)-MS/MS approach as reported (Vadassery et al., 2012). Frozen leaf tissue (~200 mg) was ground in liquid nitrogen. Phytohormones were extracted with 1 ml of ethyl acetate spiked with internal standards (100 ng of [²H₆]-JA and 20 ng each of JA-[¹³C₆]-Ile, [²H₆]-JA and JA-[¹³C₆]-Ile purchased from OlChemIm Ltd, Czech Republic). After extraction by vortexing for 10 min, the organic phase was obtained by centrifugation at 1600 g for 15 min at 4°C. Samples were evaporated almost to dryness in a vacuum concentrator (Eppendorf, Hamburg, Germany) under reduced pressure at 30°C. Leaf samples were then diluted in 600 μl of 70:30 (v/v) methanol/water. Analysis was performed with a Shimadzu 8040 HPLC-MS/MS system. Chromatography was performed on a Shimadzu FPLC system (Shimadzu). Separation was achieved on a Shimadzu pack XR-ODS column (2.0 mm i.d., 75 mm; Shimadzu). Mobile phases consisted of 0.05% [(v/v)] formic acid and 5 mM ammonium formate in water (A) and methanol (B). The elution profile was as follows: 0–0.01 min, 20% B; 0.01–8.0 min, 20–95% B; 8.0–8.1 min, 20% B. The mobile phase flow rate was 0.3 mL min⁻¹. The column temperature was maintained at 50°C. An ESI tandem mass spectrometer (Shimadzu) was operated in the negative ionization mode. The instrument parameters were optimized by infusion experiments with pure standards, where available. Molecular ions [M-H]⁻ at mass-to-charge ratio (m/z) 209.1 and 322.2, generated from endogenous JA and JA-Ile, were fragmented. Multiple reaction monitoring was used to monitor analyte parent ion → product ion: Mass-to-charge ratio m/z 209.1 → 59.0 (CE, −14 V; Q1 pre bias, −12V; Q3 pre bias, −22V) for JA; m/z 215.1 → 62.0 (CE, −10 V; Q1 pre bias, −12V; Q3 pre bias, −24V) for [²H₆]-JA; m/z 322.2 → 130.1 (CE, −21 V; Q1 pre bias, −21V; Q3 pre bias, −23V) for the JA-Ile conjugate; and m/z 328.2 → 130.1 (CE, −22 V; Q1 pre bias, −25V; Q3 pre bias, −23V) for the JA-[¹³C₆]Ile conjugate.

To investigate the effects of different environmental factors on nicotine biosynthesis in roots from 4-week-old hydroponic tobacco seedlings, we monitored the effects of various stresses, such as HT, wounding, salinity, and heavy metal cadmium treatment, on different alkaloid content, including nicotine, nornicotine, and anatabine. We found that HT, wounding and salinity treatments increased the contents of nicotine, nornicotine in tobacco roots, but cadmium treatment only increased the contents of nicotine and nornicotine, but not anatabine (Figure 1A). Among these treatments, HT treatment resulted in the greatest increase in these alkaloid content (Figure 1A). In a time course experiment, HT treatment caused a gradual increase in nicotine, nornicotine, and anatabine, which still sustained high level after 9 days of treatment (Figure 1B). Given that NtPMT1 is the key enzyme in the nicotine biosynthesis pathway (Shoji and Hashimoto, 2011; Zhang et al., 2012), we evaluated the effect of HT on NtPMT1 accumulation in tobacco roots using NtPMT1 antibody. We found that the level of NtPMT1 increased after 1 day of HT treatment and remained high for at least 9 days of HT treatment (Figure 1C).

NtMYC2a acts as the key transcriptional factor to regulate nicotine synthesis in tobacco (Zhang et al., 2012). To understand the role of NtMYC2a in nicotine synthesis under different stresses, we determined the transcriptional level of NtMYC2a after 12 h of stress treatment. As shown in Figure 2A, we found that all of these treatments, including wounding, HT, and saline or heavy metal exposure, could increase the transcriptional level of NtMYC2a in tobacco root tissue. Among these treatments, the effect of HT was substantially higher. We also measured the time course effect of HT treatment on the transcription of NtMYC2a in tobacco root tissue. HT treatment activated expression of NtMYC2a after 6 h and had a maximal effect after 24 h (Figure 2B). The sustained activation of NtMYC2a transcripts even after 72 h of HT treatment (Figure 2B) suggested that NtMYC2a may play a role in HT-induced nicotine biosynthesis.

It is reported that NtMYC2a binds to the core G-box motif [5′-CACGT(T/G)-3′] as well as to the G-box-related motif sequences [CA(C/T)(G/A)TT or CA(C/T)(G/A)TG] (Shoji and Hashimoto, 2011). By searching the promoter region, we found a series of NtMYC2a-recognized G-box or G-box-related motifs in promoters of genes associated with JA biosynthesis, including NtLOX, NtAOS, NtAOC, and NtOPR (Figure 3A; see Supplementary File 1 for detailed information). These results hint that NtMYC2a may bind these motif to activate their expression. To further investigate the role of NtMYC2a in regulating the transcription of JA biosynthesis genes, we generated transgenic tobacco expressing 6HA-fused NtMYC2a driven by a 35S promoter (NtMYC2a-6HA lines). Quantitative RT-PCR analysis showed that three transgenic tobacco lines all expressed high levels of NtMYC2a compared with Actin as the control (Supplementary Figure 1A). HA immunoblots of these lines showed that this protein was well expressed within root tissues (Supplementary Figure 1B). Based on complementary ChIP-PCR studies, NTMYC2a strongly interacted with selected components of NtAOS, NtAOC, and NtOPR promoters, such as the G-motif 5′-CACGT(T/G)-3′, but showed less interaction with fragments with other G-motifs, such as CATAT(T/G) in the promoter of NtLOX (Figure 3B). To test this possibility, we measured changes in NtMYC2a-induced gene expression in tobacco root tissue over a time course of HT treatment. Indeed the expression of these four genes was increased after 6 h of HT treatment, and these levels were sustained for the entire 96 h of HT treatment (Figure 3C). Mean wile, we found the higher transcriptional levels of NtLOX, NtAOS, NtAOC, and NtOPR in the transgenic NtMYC2a-6HA lines compared those in the control wild type lines (Supplementary Figure 2). These data are consistent with a previous study (Shoji and Hashimoto, 2011) showing that NtMYC2a binds to different G-motifs with varying affinities and indicate that the G-motif may be important for the binding of NtMYC2a to regulate the expression of JA synthesis–related genes under HT stress.

HT treatment induced a gradual increase in both JA and JA-Ile in tobacco roots, and the levels of JA and JA-Ile gradually increased after HT treatment and remained high for at least 9 days of treatment (Figure 4A). We also found that the application of exogenous MeJA treatment increased the nicotine content in tobacco seedlings (Figure 4B). These data correlated with the above finding that HT increased the transcripts of genes associated with JA biosynthesis including NtLOX, NtAOS, NtAOC, and NtOPR (Figure 3C). Diethyldithiocarbamic acid (DIECA) JA biosynthesis in plants. Application of DIECA partly suppressed HT-induced nicotine biosynthesis (Supplementary Figure 3), which further supports the role of JA signaling in HT-induced nicotine biosynthesis.

JA promotes the degradation of NtJAZ1, which also participates in nicotine biosynthesis (Shoji et al., 2008). As HT induced JA accumulation, it is possible that HT also triggers the degradation of NtJAZ1 in tobacco roots. To test this hypothesis, we generated transgenic tobacco plants that constitutively express a GFP fusion of NtJAZ1 driven by the cauliflower mosaic virus (CaMV) 35S promoter (NtJAZ1-GFP lines). Western blotting using GFP antibody showed a higher level of NtJAZ1-GFP in the root tissue of transgenic NtJAZ1-GFP lines as compared with the non-transgenic control line (Supplementary Figure 4), confirming that a higher level of NtJAZ1-GFP is present in the transgenic NtJAZ1-GFP lines. Obvious GFP fluorescence was also detected in the root tips of transgenic NtJAZ1-GFP plants (Figure 5A). The degradation of NtJAZ1 was observed by monitoring the GFP signal in NtJAZ1-GFP lines. The intensity of GFP fluorescence in the roots of NtJAZ1-GFP plants decreased gradually during HT treatment (Figure 5A, middle panel) in contrast these lines under normal condition as the control (Figure 5A, upper panel). Consistent with this, immunoblotting analysis using anti-GFP demonstrated that HT obviously induced NtJAZ1-GFP degradation in the tobacco root tissue compared with those line under normal condition as the control (Figure 5B).

Proteasome-mediated ubiquitination modulates JAZ degradation in Arabidopsis (Chung et al., 2008; Yan et al., 2013). We thus investigated whether NtJAZ1 degradation under HT treatment also depends on the proteasome-mediated ubiquitination pathway. We treated a NtJAZ1-GFP line with MG132, a specific inhibitor of the 26S proteasome, to suppress the proteasome-mediated ubiquitination pathway, and found that HT did not efficiently reduce GFP fluorescence in the root tips of the MG132-treated NtJAZ1-GFP plants (Figure 5A, bottom panel), which suggests a role for the proteasome-mediated ubiquitination pathway in HT-induced NtJAZ1 protein degradation. Application of exogenous MeJA also accelerated NtJAZ1-GFP degradation in the tobacco root tissue (Figure 5C).

NtMYC2a can bind to the G-motif within the promoter of NtPMT2 to activate NtPMT2 expression (Shoji and Hashimoto, 2011). We also searched on G-box in the NtPMT1 (Supplementary File 2). To further investigate the effect of HT on the binding capability of NtMYC2a to NtPMT1 in vivo, we generated a construct in which the luciferase gene is driven by the NtPMT1 promoter and co-transformed this construct into tobacco root protoplasts along with the NtMYC2a-6HA construct by polyethylene glycol-mediated co-transformation. We observed significant luminescence in tobacco root protoplasts co-expressing pNtPMT1:LUC and NtMYC2a-6HA (Figures 6A,B); however, the luminescence intensity increased after the protoplasts were subjected to HT at 32°C for 30 min. Luminescence was observed in the two controls, root protoplasts co-transformed with pNtPMT1:LUC with the empty effector vector and protoplasts coexpressing NtMYC2a-6HA and pNtPMT1-M1:LUC, which harbors a mutation (CACGCA) in the G-box motif of NtPMT1 (Supplementary Figure 5).

We also performed ChIP assays using the transgenic MYC2a-6HA line and anti-GFP Notably, our assays showed that, among the G-box–containing (F2) and non-G-box–containing (F1 and F3) regions, only the G-box–containing region was greatly enriched by anti-GFP in the ChIP assays (Figures 6C,D, Supplementary File 2). These results demonstrate that NtMYC2a specifically associates with the G-box region in the promoter of NtPMT1.

NtJAZ1 can sequester NtMYC2a under normal conditions (Zhang et al., 2012). In addition, JA treatment can promote NtJAZ1 degradation to release free NtMYC2a and subsequently activate the transcription of NtPMT1 and nicotine synthesis (Shoji et al., 2008). To investigate whether HT treatment also releases NtMYC2a from the NtMYC2a-NtJAZ1 complex to activate nicotine synthesis, we generated a construct containing GFP fused to NtJAZ1 (NtJAZ1-GFP) or to a version of NtJAZ1 that lacks the JAS motif (NtJAZ1ΔC-GFP; Figure 7A). NtJAZ1ΔC is resistant to proteasome-mediated degradation (Shoji et al., 2008). We also generated a transgenic line expressing NtJAZ1ΔC driven by the 35S promoter (NtJAZ1ΔC-GFP). Western blotting indicated a high level of NtJAZ1ΔC-GFP accumulation in these transgenic lines (Supplementary Figure 4). HT treatment obviously induced the degradation of NtJAZ1-GFP in the root tissue of transgenic NtJAZ1-GFP lines but did not induce the degradation of NtJAZ1ΔC-GFP in the root tissue of transgenic NtJAZ1ΔC-GFP lines (Figures 7A,B). Consistence with it, we found that co-expression of NtJAZ1ΔC also repressed NtMYC2a-induced pNtPMT1-LUC activity at 24°C. In contrast, HT treatment abolished the inhibitory effect of NtJAZ1 on NtMYC2a-induced pNtPMT1-LUC activity but did not significantly inhibit the effect of NtJAZ1ΔC on NtMYC2a-induced pNtPMT1-LUC activity (Figures 6A,B), which suggest the possibility that HT accelerated the nicotine biosynthesis through releasing the inhibitory effect of NtJAZ1 on NtMYC2a function.

The root tissue of NtJAZ1-GFP and NtJAZ1ΔC-GFP seedlings exhibited lower nicotine biosynthesis under normal conditions at 24°C compared with that of the non-transgenic tobacco. HT treatment at 32°C for 5 days obviously increased nicotine biosynthesis in the root tissue of the non-transgenic line. HT also increased the nicotine content in the NtJAZ1-GFP lines, in contrast, HT treatment did not significantly induce nicotine biosynthesis in the NtJAZ1ΔC-GFP lines as compared with that in the NtJAZ1-GFP or non-transgenic control lines (Figure 8). These data indicate that NtJAZ1ΔC, as the dominant-negative form of NtJAZ1, suppresses HT-induced nicotine biosynthesis after HT treatment.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.