Genetic Factors for Enhancement of Nicotine Levels in Cultivated Tobacco

Nicotine has practical applications relating to smoking cessation devices and alternative nicotine products. Genetic manipulation for increasing nicotine content in cultivated tobacco (Nicotiana tabacum L.) may be of value for industrial purposes, including the possibility of enhancing the efficiency of nicotine extraction. Biotechnological approaches have been evaluated in connection with this objective, but field-based results are few. Here, we report characterization of two genes encoding basic-helix-loop-helix (bHLH) transcription factors (TFs), NtMYC2a and NtMYC2b from tobacco. Overexpression of NtMYC2a increased leaf nicotine levels in T1 transgenic lines approximately 2.3-fold in greenhouse-grown plants of tobacco cultivar ‘NC 95′. Subsequent field testing of T2 and T3 generations of transgenic NtMYC2a overexpression lines showed nicotine concentrations were 76% and 58% higher than control lines, respectively. These results demonstrated that the increased nicotine trait was stably inherited to the T2 and T3 generations, indicating the important role that NtMYC2a plays in regulating nicotine accumulation in N. tabacum and the great potential of NtMYC2a overexpression in tobacco plants for industrial nicotine production. Collected data in this study also indicated a negative feedback inhibition of nicotine biosynthesis. Further enhancement of nicotine accumulation in tobacco leaf may require modification of the processes of nicotine transport and deposition.

methylputrescine oxidase (MPO) which catalyzes the deamination of methylputrescine. The pyridine ring is derived from nicotinic acid, an intermediate of the pyridine nucleotide cycle that is regulated by an entry enzyme of the cycle, quinolinic acid phosphoribosyltransferase (QPT) 11 . The nicotine synthase (NS), which catalyzes the final condensation step between nicotinic acid and methylpyrrolinium cation to form nicotine, has not been elucidated, although two candidate genes (A622 and NtBBL) had been identified 10,12 .
Several attempts have been made to increase nicotine levels in tobacco plants using biotechnological methods. Because of the central role that PMT plays in nicotine biosynthesis 13 , NtPMT1a was the first nicotine synthetic gene to be overexpressed in N. sylvestris, a progenitor species of tobacco, under control of the CaMV 35S promoter. Overexpression in greenhouse-grown plants was found to increase PMT transcripts levels 4 to 8 fold, with nicotine levels increased by about 40% 14 . In a separate report, overexpression of NtPMT1a or NtQPT2 individually or simultaneously was shown to enhance leaf nicotine levels in T 0 transgenic presumably greenhouse-grown plants of N. tabacum cv. K326 15 . Wang 16 obtained contradictory results in field-grown tobacco plants, however, and failed to observe increased nicotine contents in transgenic plants overexpressing NtPMT1a and/or NtQPT2. Recently, genes encoding for transcription factors were shown to positively regulate nicotine levels. NbbHLH1 and NbbHLH2 were the first two MYC2 related transcription factors characterized in the genus Nicotiana. Overexpression of either gene significantly increased nicotine levels in N. benthamiana, a small wild Nicotiana species. However, the reported nicotine level in fresh leaf of NbbHLH2 overexpression plants was only about 0.025% (0.25 mg/g) 17 , which is roughly ten times lower than nicotine levels typically observed for field-grown tobacco cultivars. In addition, several ERFs (ethylene response factors), another important class of transcription factors, were shown to positively regulate nicotine biosynthesis 18,19 , but their effects on enhancement of nicotine content in cultivated tobacco plants have not been demonstrated.
In another approach, overexpression of the allene oxide cyclase gene, a gene controlling a key step in jasmonate formation, was reported to cause a 4.8 fold increase in nicotine content in transgenic greenhouse-grown T 0 N. tabacum cv. Petit Havana 20 , an early-flowering tobacco type with very low leaf number 21 .
Here, we describe two NtMYC2 genes, NtMYC2a and NtMYC2b, isolated from a yeast one hybrid library screening using the promoter of NtQPT2 as bait and demonstrate that overexpression of either gene can effectively increase nicotine accumulation in greenhouse-grown T 0 and T 1 generation tobacco plants of a standard tobacco cultivar. We further demonstrate stable transmission of the increased nicotine trait through the T 2 and T 3 generations of NtMYC2a overexpression lines. Results from field tests of NtMYC2a overexpression lines suggests that transgenic overexpression of this transcription factor may be useful as a novel source to increase the nicotine content in cultivated tobacco. In addition, we observed evidence of a negative feedback loop of nicotine to regulate expression of its synthetic genes in tobacco root. Strategies to further increase nicotine levels in tobacco are discussed.

Results
Cloning of NtMYC2a and NtMYC2b genes from tobacco. The yeast one-hybrid technique was used in this study to identify transcription factors that bind to the promoter sequence of NtQPT2. A total of 1.6 million yeast clones were screened. Analysis of sequencing data revealed that a single clone carried a cDNA encoding a bHLH transcription factor, but the 5′ terminal sequence was missing. The full length cDNA sequence was obtained using the Rapid Amplification of cDNA Ends (RACE) technique. During the process, another highly homologous bHLH transcription factor was also identified.
The full-lengths of the two cDNAs were found to be 2214 and 2391 bp encoding for proteins of 659 and 658 amino acids, respectively. The two cDNAs shared 95% identity to each other at the nucleotide sequence level, and BLAST analysis to the NCBI database showed these two transcription factors to be 99.9% (1979/1980 nt) and 100% (1977/1977 nt) identical to the coding sequences of NtMYC2a (GenBank No. GQ859160) and NtMYC2b (GenBank No. GQ859161). The single nucleotide difference in NtMYC2a could be caused by error in sequencing or PCR, or by cultivar difference. The two isolated TF genes were thus designated as NtMYC2a and NtMYC2b.
A phylogenetic tree based upon amino acid sequences was constructed using the maximum likelihood method in MEGA5.2 22 to infer the evolutionary relationships of both NtMYC2a and NtMYC2b with other bHLH transcription factors. Several closely related transcription factors from the BLAST result were also included. As expected, NtMYC2a and NtMYC2b are closely related to each other, most closely related to NbbHLH2 and NaMYC2, and more distant to NtMYC1a, NtMYC1b, NbbHLH1 and StMYC2 17,23-25 (Fig. 1).
NtMYC2a and NtMYC2b differentially enhanced nicotine levels and affected expression of nicotine synthetic genes. To investigate the effect of the NtMYC2a and NtMYC2b TF genes on the biosynthesis of nicotine, we evaluated transgenic tobacco plants over-expressing these genes. The coding sequences of NtMYC2a and NtMYC2b were placed under the control of the CaMV 35S promoter to enhance the expression of the genes. In total, seven 35S:NtMYC2a and nine 35S:NtMYC2b T 0 plants were generated and initially tested.
Northern analysis was performed to test the expression levels of NtMYC2a/b in T 0 greenhouse-grown plants. The results indicated that two NtMYC2a transgenic plants (AOE-3 and AOE-6) and six NtMYC2b Scientific RepoRts | 5:17360 | DOI: 10.1038/srep17360 transgenic plants ( 11,13,14,16 and 17) clearly exhibited increased expression of these genes compared to the wild type and vector control plants (Fig. 2). Leaf nicotine levels of these plants were quantified. Three 35S:NtMYC2b plants (BOE-10, 16 and 17) accumulated higher nicotine levels (approximately 40% over the vector control) while the other three plants (BOE-11, 13 and 14) produced nicotine levels that were similar to, or lower than, the controls. Two 35S:NtMYC2a T 0 plants (AOE-3 and AOE-6) exhibited much higher nicotine levels as compared to the controls (123% and 150% higher, respectively, than the vector control) (Fig. 3). Based on this initial data, we selected two 35S:NtMYC2a and two 35S:NtMYC2b overexpression plants (AOE-3, AOE-6, BOE-16 and BOE-17) to study the effects of overexpression of these two TFs on the expression of genes associated with the nicotine biosynthetic pathway using quantitative real-time reverse transcription-PCR (qRT-PCR). Overexpression of NtMYC2b did not substantially alter the mRNA levels of the nicotine synthetic genes except for a moderate reduction in the expression of the NtPMT, NtA622 and NtBBL genes (Fig. 4a). However, the expression of all the evaluated nicotine synthetic genes was remarkably reduced by overexpression of NtMYC2a. We extended the transcript quantification of the seven nicotine synthetic genes to plants of T 1 generation of those overexpressing lines. Two plants from each line and a total of four plants for each NtMYC2 gene construct were analyzed. The results largely confirmed observations for the T 0 plants. The expression of the nicotine synthetic genes was much suppressed in NtMYC2a overexpression lines as compared to vector control (Fig. 4b). No obvious growth or developmental differences were noted between the transgenic and wild-type plants.
Inheritance of high nicotine content in T 1 NtMYC2a and NtMYC2b overexpression lines. To further confirm the functions of both NtMYC2 genes and to determine the potential of inheritance of the elevated nicotine trait, T 1 generation families were produced by self-pollination of AOE-3, AOE-6,  BOE-16, BOE-17 and vector control T 0 plants. T 1 plants were grown in greenhouse for approximately two months and PCR was used to identify transgenic segregants. Leaf nicotine levels were determined in non-topped plants and plants ten days after topping. The results (Fig. 5) showed that non-topped NtMYC2a overexpression (AOE-3 and AOE-6) T 1 generation plants had nicotine levels that were 134% higher than that of the vector control, while NtMYC2b overexpression plants (BOE-16 and BOE-17) exhibited 35% increase in nicotine content. Similar increases were observed in the topping treatment. Results were highly consistent with those observed in the T 0 plants (Fig. 3), indicating inheritance of high-nicotine trait in these transgenic plants.
In both the topped and non-topped treatments, nicotine levels of NtMYC2a overexpression plants were shown to be significantly greater than those of NtMYC2b overexpression plants (Student's t-test, P < 0.0001), indicating these two closely-related paralogs function differentially to affect nicotine biosynthesis in tobacco.   to evaluate the possible effects of overexpression of these genes on overall plant performance, we carried out field experiments to compare vector control, two non-segregating NtMYC2a overexpression T 2 lines (AOE-3-19, AOE-6-6) and two non-segregating NtMYC2b overexpression T 2 lines (BOE-16-37, BOE-17-12). The field-grown transgenic lines exhibited no obvious phenotypic differences from vector control parental line, even though nicotine levels in AOE-3-19 and AOE-6-6 were significantly greater than the vector control (76% and 48% higher, for lines AOE-3-19 and AOE-6-6, respectively) (Student's t-test, P < 0.05). The two NtMYC2b overexpression lines did not exhibit a significant increase in nicotine levels (Fig. 6).
Because the 35S:NtMYC2a construct exhibited a much greater impact on nicotine levels, we continued to evaluate the T 3 generation of these NtMYC2a overexpression lines in the field. Again, very significant increases in nicotine levels were observed (Fig. 7) (58% and 56% higher, for lines AOE-3-19   and AOE-6-6, respectively). However, the cured leaf yields of these lines were reduced by 23% and 26%, respectively, as compared to the control. Although yields were decreased by about one fourth in both lines, the total nicotine production per hectare for the overexpression lines was still increased by 22% and 14%, respectively.

The expression of nicotine synthetic genes was inhibited by nicotine. Previous studies showed
that high levels of nicotine can be toxic to tobacco nicotine-producing root cells, and maintenance of a relatively low cytoplasmic nicotine concentration is necessary for normal root cell functions 26,27 . Since the expression of most nicotine synthetic genes in NtMYC2a overexpression lines was down regulated, we further tested whether nicotine has a role in regulating the expression of its own synthetic genes. qRT-PCR results showed that the expression levels of all nicotine synthetic genes were repressed to about one-half two hours after root treatment of tobacco seedlings with 0.4 mM exogenously applied nicotine (Fig. 8).

Discussion
Previous reports have indicated that expression of the major genes involved in nicotine biosynthesis was coordinately regulated to some extent by MeJA treatment [28][29][30] . These findings prompted us to hypothesize that a major or master transcription factor related to the MeJA signaling pathway might exist to co-regulate the nicotine synthetic genes for nicotine formation. Since overexpression of two key genes in the nicotine biosynthesis pathway (NtPMT1a and NtQPT2), individually or simultaneously, did not increase the nicotine content in cultivated tobacco, this leads to the suggestion that other key gene(s) might exist in the pathway and their expression might also need to be altered to substantially increase the nicotine accumulation in the tobacco plant 16 . In the current study, we attempted to identify a putative master transcription factor with the hope that this master transcription factor would regulate all the nicotine synthetic genes, and that manipulation of the expression of this gene would help us achieve the goal of generating transgenic tobacco plants with elevated leaf nicotine accumulation.
Although some transcription factors have been reported to positively regulate nicotine biosynthesis, none of them has been reported to increase the nicotine level in a field-grown, cultivated tobacco variety 17,19,24,[31][32][33] .
To use a plant-based system for secondary metabolite production, greenhouse experiments can provide valuable initial information. Field testing is necessary, however, to determine the effect of a given technology because this is where crop plants are ultimately grown. Positive results in greenhouse environments may not always translate into similar results in actual field growing environments. In this study, we used tobacco cultivar NC95 to evaluate the potential of overexpression of NtMYC2a or NtMYC2b gene to increase the nicotine accumulation in field-grown tobacco.
Constitutive overexpression of NtMYC2a and NtMYC2b was shown to significantly increase leaf nicotine levels in N. tabacum in T 0 and T 1 generation non-topped or topped greenhouse-grown plants (Figs 3 and 5). Consistent with these results, the T 2 and T 3 generations of the NtMYC2a overexpression lines exhibited increased nicotine concentrations in field testing, indicating that the high nicotine trait can be stably inherited (Figs 6 and 7). Results from the current study clearly show that NtMYC2a is more effective for increasing nicotine content than NtMYC2b.
From an industrial perspective, both nicotine concentration and leaf yield are important for nicotine production. In the present study, we observed that the leaf yield of both T 3 NtMYC2a overexpression lines decreased by approximately one-fourth. Indeed, there are reports of negative genetic correlations between yield and nicotine content in non-transgenic populations of N. tabacum 34 . The negative correlation could be caused by competition for plant resources necessary for both nicotine biosynthesis and tobacco plant growth such as nitrogen. If this is the case, increased nitrogen fertilizer application might partially alleviate the decreases in leaf yields. In our experiments, although leaf yields were decreased, the total nicotine yield per hectare in an AOE3 line was still 22% higher than would have otherwise been achieved using the non-transgenic tobacco control line.
It was expected that the transcript levels of nicotine synthetic genes would increase in the NtMYC2b and NtMYC2a overexpression T 0 or T 1 plants since nicotine levels were significantly increased. Our results indicated, however, that overexpression of NtMYC2b in transgenic tobacco plants did not lead to obvious increases in transcript levels of nicotine synthetic genes in root tissues and the expression of all nicotine synthetic genes was remarkably suppressed in NtMYC2a overexpression lines (Fig. 4). These observations are similar to a previous report that overexpression of NbbHLH2 reduced the transcript levels of most nicotine synthetic genes while the nicotine levels were significantly increased in MeJA-treated transgenic plants 17 . Since nicotine was shown to retard tobacco root growth and so to be toxic to the tobacco root cell at relative high concentration 26 , and exogenously supplied nicotine reduced the activity of PMT in tobacco root at the time of decapitation 35 , we speculated that high level of nicotine may inhibit the expression of its own synthetic gene(s) to maintain it a reasonable level in tobacco root to ensure normal root cellular functions. To gain more insights, we performed a nicotine feeding experiment and the data (Fig. 8) showed that nicotine did down regulate the expression of its biosynthetic genes. Thus, a plausible explanation of down regulation of nicotine synthetic genes' expression in NtMYC2a overexpression plants is that, high levels of nicotine produced in the roots of transgenic plants may trigger negative feedback regulation on the expression of its synthetic genes. In a similar case, flavonol-dependent feedback inhibition of transcription of several flavonol synthetic genes appears to prevent the accumulation of toxic non-glycosylated flavonols in Arabidopsis thaliana 36 .
To achieve further increases in nicotine accumulation in tobacco leaves, strategies might be pursued to target the process of uploading of nicotine into the xylem and/or translocation and accumulation to leaf vacuoles. Recent research has revealed two vacuole-localized transporters responsible for the sequestering of nicotine into the vacuole of root and leaf (MATE1/2, NtJAT1) and one plasma membrane transporter (NUP1) responsible for importing nicotine from the apoplast into the cytoplasm of tobacco root cells 5,26,37 . However, transporters responsible for loading of nicotine into the xylem in the root, root to shoot transportation, and unloading of nicotine into leaf cells remain unknown. Characterization of those transporters might be helpful to further enhance nicotine levels in tobacco leaves.
In summary, we isolated two bHLH transcription factor genes, NtMYC2a and NtMYC2b, in this study. Overexpression of NtMYC2a was shown to substantially increase nicotine accumulation in the cultivated tobacco variety NC95. The high nicotine trait was stably inherited to the T 1 , T 2 and T 3 generations, and field tests showed that nicotine concentrations reached as high as 6.74% and 5.23% dry weight in tobacco leaves in the T 2 and T 3 generations of NtMYC2a overexpression lines, levels that were significantly higher than the controls. Together with the leaf yield data of T 3 lines, as high as a 22% increase in total nicotine production per hectare was achieved in NtMYC2a overexpression lines, thus indicating a great potential of NtMYC2a overexpression in tobacco plants for industrial nicotine production.

Yeast one-hybrid experiments for cloning transcription factors.
A yeast one-hybrid system, Matchmaker TM One-Hybrid Library Construction and Screening Kit (Clontech, Mountain View, CA, USA), was employed to screen for transcription factors that bind to the promoter of the tobacco NtQPT2 gene. The promoter, 1034 bp in length, was cleaved from construct pTobRD2-PMTOX provided by 22nd Century, LLC (Buffalo, NY, USA), and inserted upstream of the GAL4 minimal promoter in vector pHIS2.1 to form the bait construct pTobHis.
Total RNA was extracted from roots of two-month-old greenhouse-grown tobacco plants (cv. NC95) 0.5 h after topping and was used for cDNA preparation. Screening of positive colonies was performed according to the kit manual. Sequences of the cDNA inserts of positive colonies were used for BLAST analysis with the NCBI GenBank database to identify transcription factors.

Recovery of full-length cDNAs of the TFs.
A GeneRacer kit (Invitrogen, Carlsbad, CA, USA) was used to obtain the missing 5′ sequences of the cDNAs following manufacturer's instructions. The gene-specific primer used in RACE for NtMYC2 was NtMYC2GSP: 5′ -ACACATTTGGTACAACAGCTCTAAGTGC-3′ . The primers designed to obtain the full-length coding sequences of NtMYC2a and NtMYC2b are listed in Supplementary Table S1.
Generation of transgenic tobacco plants for overexpression of the NtMYC2 genes. Binary vector pBI121 38 was used as the backbone to produce NtMYC2a or NtMYC2b overexpression constructs. The coding sequences of the NtMYC2 genes were cleaved from their cloning vectors (pCR-BluntII-TOPO or pCR4-TOPO) and inserted into pBI121 to replace the GUS gene.
Transgenic plants were produced from cv. NC95 by using standard Agrobacterium tumefaciens-based leaf disc transformation 39 . Initial primary transformants (T 0 ) plants were grown in the greenhouse. T 0 plants were self-pollinated to produce T 1 generation plants that were tested the presence of transgene using Scientific RepoRts | 5:17360 | DOI: 10.1038/srep17360 vector specific 5′ primer (GTGGCTCCTACAAATGCCATCA) and gene-specific 3′ primers (CAAACAC AAGTTGCTTACTGG for NtMYC2a and CAAACACAAGTTGCTTACTAC for NtMYC2b). T 1 plants were self-pollinated to produce the T 2 generation, and non-segregating T 2 lines were selected on MS medium containing 100 mg/L kanamycin (total 60 seeds, all germinated into healthy seedlings). Plants from non-segregating T 2 lines were self-pollinated to produce T 3 generation plants.
For T 0 and transgenic T 1 plants grown in greenhouse without topping, leaf samples were collected just before flowering. For transgenic T 1 plants with topping, leaf samples were collected ten days after topping (plants were topped immediately prior to flowering). From each plant, all leaves were collected, dried at 60 °C for three days and ground for nicotine quantification. house for about two months until just before flowering. Around 700 mg root tissue from each plant was used for RNA isolation. Total RNA was extracted by TRIzol (Invitrogen) according to the manufacturer's manual. The extracted RNA was dissolved in DEPC treated water and quantified by Nanodrop (ND-1000, Thermo Fisher Scientific, Waltham, MA, USA). Ten μ g RNA was separated on 1% agarose gel in MOPS buffer. The gel was stained with EtBr and the image was taken under UV. Separated RNA on the gel was blotted onto the Hybond-N+ nylon membrane (Amersham, Arlington Heights, IL, USA).
The primers designed from the coding sequence of NtMYC2b for generating the probes for northern hybridization were: NtMYC2a/bNOR F: 5′ -GAAGTAACGGATACTGAATGG-3′ NtMYC2a/bNOR R: 5′ -ATCCTTGTGTTTGCTGAGAAT-3′ The probe is a 505 bp fragment which shares 94% nt identity to the corresponding region of NtMYC2a and thus used to test the transcript level of both NtMYC2 genes (NtMYC2a/b).
qRT-PCR analysis of nicotine synthetic genes expression. Root tissue was collected from T 0 and T 1 transgenic plants just before flowering. RNA was isolated from root tissue using TRIzol reagent (Invitrogen) according to the manufacturer's manual and first strand cDNA was synthesized using SuperScript ™ III Reverse Transcriptase (Invitrogen) with an oligo (dT) primer. qRT-PCR was performed using FastStart Universal SYBR Green Master (Rox) (Roche, Mannheim, Germany) on ABI7900 (Applied Biosystems, Foster City, CA, USA). The N. tabacum actin gene was used as a control for normalization. Three technical replicates were performed using the following PCR program: 50 °C for 1 min; 95 °C for 15 sec; 40 cycles of 95 °C for 15 sec followed by 63 °C for 1 min using the primers for nicotine synthetic genes listed in Supplementary Table S2. For T 0 plants, two transformation events per gene construct were analyzed. For T 1 plants, two transgene positive plants from each T 1 line and total four plants for each NtMYC2 gene overexpression construct were analyzed. For nicotine-treated tobacco seedlings, three plants from each time point were analyzed. The means and standard errors are presented.
Field tests of T 2 and T 3 generations of the NtMYC2 overexpression lines. Two NtMYC2a and two NtMYC2b non-segregating overexpression lines (T 2 generation) were compared for nicotine levels in a field experiment relative to a vector control. The experimental design was a randomized complete block design with three replications and each plot consisting of 15 plants. Plants were topped immediately prior to flowering. Ten days after topping, the first priming was harvested. The following priming was harvested ten days after the previous one. All the leaves at the stalk position #10 (from the top to the bottom) from the second priming from each replicate were dried, ground, and analyzed for nicotine contents using methodologies described below.
Similar field tests were performed to examine the performance of two T 3 NtMYC2a overexpression lines in a replicated experiment using a randomized complete block design with four replications. Plots consisted of single 20-plant rows. Leaves were harvested in three separate harvests (primings) and flue-cured. Each priming was weighed to generate yield data. Fifty-gram cured leaf samples were prepared for each plot by compositing cured leaf from each priming on a weighted-mean basis for nicotine quantification.
Quantification of nicotine. Each sample was prepared by placing 0.2000 ± 0.0010 g of dried ground tobacco leaves into a 50 mL Erlenmeyer flask, and 2 mL of 2N NaOH solution was added to each flask and swirled to moisten the tobacco. After 15 min of rest, 10 mL of methyl tertiary butyl ether (MTBE) containing 0.1062 g/mL of quinoline was added to the flask. The flasks were placed on a shaker for 2.5 hrs. After shaking the flasks were allowed to sit overnight to separate. Approximately 1 mL of the top MTBE layer was transferred into a vial. GC analysis was conducted using a split injection (40:1) on an Agilent HP 6890 GC-FID (Agilent Technologies, Santa Clara, CA) using a 30 meter DB-5 MS column (0.53 mm ID and 1.5 μ m film thickness). The carrier gas was helium at a linear velocity of approximately 38 cm/sec. The injector and detector were both set at 250 °C. The analysis consists of a temperature program from Scientific RepoRts | 5:17360 | DOI: 10.1038/srep17360 110 °C initially held for 0.5 min followed by a ramp to 280 °C at a rate of 25 °C/min where the final temperature was held for 20 min. Data were collected and analyzed using Agilent Chemstation software. A multipoint internal standard calibration table was constructed for nicotine. Statistical analysis. Two-tailed Student's t-tests were performed for comparisons of any two groups within each experiment. Probability values of < 0.05 were considered statistically significant.