Introduction

In cultivated tobacco (Nicotiana tabacum L.) nicotine is the most abundant alkaloid, typically constituting more than 90% of the total alkaloid pool. The remaining proportion of the alkaloid profile is primarily made up of secondary alkaloids, such as nornicotine, anatabine and anabasine (Sisson and Severson 1990). Nornicotine, the most abundant secondary alkaloid, represents about 3–5% of the total alkaloid content (Saitoh, et al. 1985). In most tobacco populations, however, individual plants can be identified in which a large percentage of the nicotine is metabolized into nornicotine in the senescing leaves. Plants that accumulate nicotine as their primary alkaloid are referred as “nonconverters”, while individuals that convert a large portion of their nicotine content to nornicotine during senescence and curing are named “converters”. The transition of tobacco plants from the nonconverter to the converter phenotype is termed “conversion”, a process that can occur in a single generation as nonconverter plants can give rise to converter progeny (Mann et al. 1964). Conversion of tobacco is controlled by an unstable conversion locus which is derived from the N. tomentosiformis progenitor of tobacco (Mann et al. 1964). The molecular identity of the conversion locus has not been unambiguously identified, but indirect evidence suggests that the locus encodes the nicotine N-demethylase gene CYP82E4 (Chakrabarti et al. 2007; Gavilano et al. 2007; Siminszky et al. 2005).

Nicotine is predominantly produced in the roots of tobacco and translocated through the xylem to aerial parts of the plant (Dawson 1942; Hashimoto and Yamada 1994) where it is secreted by trichomes (Thurston et al. 1966). Using reciprocal grafts of converter and nonconverter scions and stocks, (Wernsman and Matzinger 1968) showed that nicotine to nornicotine conversion mainly occurs in the senescing leaves and only low levels of nornicotine are produced in the roots. Similar to nicotine, nornicotine was also detected in trichome secretions (Thurston et al. 1966). Nornicotine is formed through the oxidative N-demethylation of nicotine, a reaction primarily catalyzed by CYP82E4 in the senescing leaves (Gavilano et al. 2006; Siminszky et al. 2005; Xu et al. 2007) and by CYP82E5v2 in the green leaves of tobacco (Gavilano and Siminszky 2007). Transcript accumulation of CYP82E4 is very low in the green leaves of converter and nonconverter tobacco and is sharply upregulated in the senescing leaves of converter tobacco demonstrating that the expression of CYP82E4 is regulated at the level of transcription (Gavilano et al. 2006; Gavilano et al. 2007; Siminszky et al. 2005). The strong induction of CYP82E4 by senescence and the relatively high catalytic competence of the encoded protein were consistent with the senescence-dependent high nornicotine phenotype of converter tobacco suggesting that CYP82E4 plays a key role in nicotine to nornicotine conversion in converter plants.

Beyond the observations that nicotine to nornicotine conversion is induced by senescence, ethephon (Fannin and Bush 1992) and sodium bicarbonate (Shi et al. 2003), little is known about the spatial, temporal, hormonal and stress-mediated regulation of the process. Senescence is a complex physiological process that occurs in response to aging and can be accelerated by various biotic and abiotic stresses, such as drought, darkness and pathogen attack, in addition to the hormones abscisic acid (ABA) and ethylene (reviewed in Lim et al. 2007; Yoshida 2003). Genes whose transcription is upregulated during senescence (i.e. senescence-associated genes, SAGs) can be either directly or indirectly regulated by stress or hormonal treatments. If the response of the SAG is rapid or if it is elicited in the young, green tissues, a direct regulatory role can be inferred for these factors. In contrast, a slow SAG response or a response that is restricted to the older leaves indicates an indirect effect of the inducer on SAG regulation (Weaver et al. 1998).

In this report, we used a fusion construct between the CYP82E4 promoter and the β-glucurodinase (GUS) reporter gene to further investigate the molecular identity of the locus responsible for nicotine conversion. We also examined the tissue- and organ-specific transcriptional regulation of the CYP82E4 promoter, and tested the effects of signaling molecules, such as ethylene, jasmonic acid (JA), salicylic acid, ABA and yeast extract; infection with viruses; and abiotic stresses including drought and wounding on the stimulation of CYP82E4 promoter activity. Our results showed that the CYP82E4 promoter directs reporter gene expression to the aerial parts of the plant, and the activity of the promoter is induced by a senescence-specific pathway in the leaves. They also provide further evidence that the CYP82E4 gene itself represents the unstable converter locus responsible for the spontaneous occurrence of converter plants within nonconverter tobacco populations.

Materials and methods

Isolation of CYP82E4 promoter region

To characterize the sequences immediately upstream of the CYP82E4 reading frame, a 33P-labeled fragment of the CYP82E4 cDNA was used as a hybridization probe to screen a tobacco BAC library generated through the North Carolina State University Tobacco Genome Initiative (http://www.tobaccogenome.org). Diagnostic restriction digests were subsequently used to distinguish BAC clones containing CYP82E4 from clones possessing other P450s of the CYP82E subfamily that cross-hybridized to the probe. Partial DNA sequence information from a BAC encompassing CYP82E4 was obtained by SeqWright DNA Technology Services (http://www.seqwright.com) using a random shotgun sequencing approach. Sequence alignments revealed a contig containing CYP82E4 together with approximately 2.5 kb of 5′ flanking sequence.

To isolate a fragment of the upstream regulatory region of CYP82E4 from tobacco genomic DNA, a nested PCR approach was employed. Genomic DNA was isolated from three-month-old tobacco leaves using the Plant DNAzol reagent as described by the manufacturer (Invitrogen Life Technologies, Carlsbad, CA). Due to the complex repeat structure of the CYP82E4 locus, several primers had to be tested to obtain a single amplification product. A single, 3,325 bp amplicon was produced when the CYP82E4 promoter-specific 5′-TGGAAAAGATGGCCGATTAC-3′ forward primer and the CYP82E4 intron-specific 5′-GGGAAGAGCCACACTCCAAT-3′ reverse primer were used in the first PCR reaction, and the nested 5′-GAGGTTAGGCTCGGCATTTA-3′ forward and the 5′-GGGGTATGATGCAGACAACA-3′ reverse primers were added to the second PCR. The first PCR reaction contained 100 ng genomic DNA template, 2 μM each primer, 200 μM each of dNTP, and 1.5 mM MgCl2 in a 50-μl final reaction volume. The conditions for the PCR amplification were as follows: initial denaturing at 95°C for 3 min, followed by 30 cycles of denaturing at 95°C for 30 s, annealing at 55°C for 30 s, extension at 72°C for 90 s. In the nested PCR, the same conditions were used, except 1 μl of the10-fold diluted product of the first PCR was used as a template. The final amplicon was ligated into the pGEM-T Easy T/A cloning vector (Promega Corp., Madison, WI).

Generation of the CYP82E4:GUS and the promoterless GUS constructs

Because the 3,325 bp CYP82E4 cloned fragment contained not only 2,172 bp of sequence corresponding to the CYP82E4 promoter region, but also a portion of the intron and the entire first exon of the gene, an additional PCR was used to extract the 2,172 bp putative regulatory region from the clone. Amplification of the 2.2 kb promoter fragment was achieved by PCR using the primers (5′-GGATCCTATGGATTTTTAGATAAGAAGTTGAGAA-3′ (forward) and 5′-AAGCTTGAGGTTAGGCTCGGCATTTA-3′ (reverse) primers that were included HindIII and BamHI restriction enzyme recognition sites to facilitate ligation into the pBI121 plant expression vector. The DNA sequence of 2.2 kb promoter region was deposited into the GenBank under the accession number EU338375. The CYP82E4:GUS construct was created by replacing the CaMV 35S promoter of the pBI121 vector with the 2.2 kb upstream regulatory region of CYP82E4. The promoterless GUS construct was developed by excising the CaMV 35S promoter from pBI121, filling the 5′ overhangs of the linearized vector with Klenow DNA polymerase and ligating the blunt ends. Consensus sequence analysis of the cis-regulatory elements was performed using the PLACE database (Higo et al. 1999).

Generation of transgenic plants

CYP82E4:GUS and the promoterless GUS constructs were introduced into Agrobacterium tumifaciens strain LBA 4404 and excised leaf tissue of tobacco lines DH98-325-6 and DH98-325-5 were transformed by the method of Horsch et al. (1985). Transgenic plantlets were transplanted to soil and kept in a growth chamber at 24°C with a 16/8 h light/dark cycle for 30 days until transferred into a greenhouse. In the greenhouse, plants were maintained in a 14/10 h light/dark cycle and were fertilized twice a week with Professional All Purpose Plant Food (20-20-20) fertilizer (Spectrum Brands Inc, Madison, WI) until seed maturation. Transgenic T1 generation plants were selected on MS-based phytoagar medium containing 100 mg/l of kanamycin. Kanamycin-resistant seedlings were grown in a growth chamber for 45 days, and thereafter in a greenhouse for an additional 45 days, and maintained as described above until treatment.

Plant treatments

All treatments were administered in the greenhouse with the exception of ethylene and JA treatments that were performed in a growth chamber under conditions already described.

Senescence

Leaves at different developmental stages were directly taken from 3-month-old plants. To artificially induce and accelerate senescence, the 5th leaf from the apex was harvested and cured in the dark in a plastic bag at 24°C until it turned yellow. Chlorophyll measurements were taken with a SPAD-502 chlorophyll meter (Konica Minolta Sensing Inc, Osaka, Japan). A SPAD unit is defined as the ratio of optical density at 650 nm and 940 nm transmitted through the leaf (Chernikova et al. 2000). Flower senescence was measured at different developmental stages defined by Koltunow et al. (1990). According to this system, Stages 1, 10, and 12 represent budding; opening of corolla limb, petal tips pink; and fully open flower with fully expanded, deep pink corolla limb, respectively. We assigned Stage 13 to senesced flowers.

Ethylene

Plants were placed into air tight, 3.9 l Mason jars fitted with septa. Ethylene gas was injected into the jars to obtain a 100 ppm final concentration, and plants were incubated for either 48 or 96 h (Butt et al. 1998). Control plants were also placed in similar chambers, but without ethylene treatment. After each day the jars were opened to eliminate the build up of CO2 and refilled with ethylene. Ethylene concentrations in the chambers were verified at the beginning and end of the treatment periods by gas chromatography using a Buck Scientific gas chromatograph (model 910; Buck Scientific Inc, East Norwalk, CT) equipped with a flame ionization detector (155°C) and an alumina column (125°C, nitrogen flow rate: 1 ml/min). For alkaloid and fluorometric GUS activity analyses, samples were taken from the 2nd and the 5th leaves from the apex.

Wounding

Mature green leaves (3rd from the apex) were pierced multiple times with a scalpel (Feather Disposable Scalpel No.10, Feather Safety Razor Co. LTD, Japan) and unwounded plants were used as a control. Samples were taken 48 and 72 h after treatment.

Drought

Drought treatment involved withholding water for six days; regularly watered plants were used as controls. At the time of sample collection, drought-treated plants showed visible wilting, whereas control plants appeared normal.

ABA

A solution containing 0.1 mM ABA [(±)-abscisic acid, Sigma-Aldrich, St Louis, MO] and 0.02% Tween-20 or 0.02% Tween-20 only was sprayed on the ABA-treated and control plants, respectively, until the solution ran off the foliage. Samples were taken 24 h after treatment (Weaver et al. 1998).

Jasmonic acid

Jasmonic acid was dissolved in small amount of ethanol and then diluted to 100 μM using distilled water. Plants were sprayed with jasmonic acid solution until the solution ran off the leaves. After spraying, the plants were transferred to an enclosed chamber, and samples were taken 48 h later (Kimura et al. 2001). Control plants were sprayed with water only.

Salicylic acid

Leaf discs excised from the 3rd leaf from the apex were floated on a solution containing 5 mM and 20 mM salicylic acid dissolved in 0.1 M phosphate buffer (pH 7.0). Samples were taken 24 h after treatment (Yin et al. 1997). Control leaves were floated on 0.1 M phosphate buffer.

Yeast extract

Leaf discs excised from the 3rd leaf from the apex were placed into Petri dishes lined with blotting papers that were soaked in 1% and 5% sterile yeast extract, and the leaf discs were incubated for 8 or 21 h. Control leaves were placed on blotting paper soaked with water (Pasquali et al. 1999).

Tobacco mosaic virus infection

Because tobacco cultivar DH98-325-6 carries the tobacco mosaic virus (TMV) resistance gene (termed N gene), TMV infection does not produce systemic infections at temperatures below 28°C, but rather, it induces the appearance of localized, necrotic spots, also known as hypersensitive response (Whitham et al. 1994). The TMV resistance trait is temperature dependent, because the N gene-mediated hypersensitive response is suppressed by elevated temperatures (above 28°C) and becomes reactivated once the plants are returned to permissive (below 28°C) conditions (Samuel 1931). Thus, the temperature-dependent function of the N gene product provides a facile means of infecting a large number of cells in TMV-resistant plants. Exposing TMV-infected plants to high temperatures and then returning them to a cooler environment causes a systemic viral infection and the widespread occurrence of necrotic lesions. To induce TMV infections in a large area of the inoculated leaf, we took advantage of the temperature sensitivity of the TMV-elicited hypersensitive response. TMV-infected leaf tissues of the TMV-susceptible TN86 tobacco cultivar were homogenized in 5 mM phosphate buffer, pH 7.0, and the transgenic DH98-325-6 plants were inoculated with the leaf extract using carborundum as an abrasive. Infected plants were incubated at 32°C for three days for systemic infection and then transferred to 22°C for nine hours to induce hypersensitive response. Control plants were rubbed with carborundum and kept under the same temperature regime as the TMV infected plants (Yang et al. 1999).

Histochemical and fluorometric GUS activity assays and alkaloid analysis

For the histochemical localization of GUS activity, tissues were incubated in GUS staining solution (100 mM phosphate buffer, pH 7.0; 2 mM X-Gluc (5-bromo-4-chloro-3-indolyl-beta-d-glucuronic acid, cyclohexylammonium salt); 10 mM EDTA and 0.1% Triton-X) at 37°C overnight, and destained with 95% ethanol. To determine GUS activity using the MUG (4-methylumbelliferyl β-d-glucuronide) fluorescent assay (MUG assay), 150 mg plant tissue was homogenized in 650 μl GUS extraction buffer (50 mM NaHPO4, pH 7.0; 10 mM DTT; 10 mM EDTA; 0.1% sarcosyl; 0.1% Triton-X and 1% PVPP) and centrifuged for 10 min at 16,000 rcf. The supernatant was transferred in a fresh Eppendorf tube, centrifuged for 5 min at 16,000 rcf and the supernatant was passed through a 650 μl Sephadex G-25 column. Ten μl plant extract was mixed with 130 μl assay buffer (1 mM MUG in the GUS extraction buffer) and incubated for 15 min. To arrest the reaction, a 10 μl aliquot of the mixture was mixed with 190 μl stop buffer (0.2 M Na2CO3). Conversion of 4-MUG to 4-MU (4-methyl umbelliferone) was measured using an excitation wavelength of 365 nm, an emission wavelength of 445 nm in a SpectraMax M2 microplate reader (Molecular Devices Corporation, Sunnyvale, CA). Protein concentration was determined using the Bio-Rad protein assay according to the manufacturer’s instructions (Bio-Rad Laboratories, Hercules, CA). Alkaloid analysis of tobacco leaves was performed by gas chromatography as previously described (Gavilano et al. 2006). Nicotine conversion was calculated as [%nicotine/(%nicotine + %nornicotine)] × 100, where alkaloid concentration is expressed as percentage of leaf dry weight.

Results

Generation of transgenic CYP82E4:GUS tobacco lines

To investigate the gene expression pattern conferred by the putative CYP82E4 promoter, we transformed tobacco with a fusion construct between the 2.2 kb 5′ flanking region of the tobacco CYP82E4 and the GUS reporter gene. The construct was introduced into the full-sib tobacco cultivars DH98-325-6 (converter) and DH98-325-5 (nonconverter), and 17 and 13 independently transformed plants were regenerated, respectively. Histochemical assay of GUS activity revealed that upon senescence, eight, five and four plants in the DH98-325-6 background and seven, two and four plants in the DH98-325-5 background exhibited high, intermediate and low GUS activity, respectively (data not shown). To determine whether the CYP82E4 promoter confers the same expression pattern in independently transformed transgenic lines, a small-scale, preliminary experiment was conducted using the eight highly-expressing DH98-325-6 primary transformants. The plants were exposed to all the experimental treatments used in the study (see "Material and Methods") and GUS activity was measured by histochemical staining. The results showed identical pattern of GUS expression in all eight transgenic lines demonstrating that the positional effect frequently associated with independent transgene integration events did not influence the spatial and developmental regulation of the CYP82E4 promoter. As a result, we used the T1 progeny of the CYP82E4:GUS_59 line, which exhibited high levels of GUS expression in the DH98-325-6 converter background, for all the experiments unless otherwise indicated.

Expression of the CYP82E4 promoter is differentially regulated in various organs

Leaves

To confirm that the transcriptional control of the CYP82E4:GUS reporter gene construct recapitulates the senescence-inducible regulation of the native CYP82E4 gene, we measured GUS activity in tobacco leaves at five developmental stages ranging from young leaves with dark green color to mature, artificially cured leaves that reached complete senescence (Fig. 1). The chlorophyll content of the leaves was measured with a SPAD meter and the degree of yellowing was correlated with GUS activity. GUS activity, as measured by the MUG assay, was equally low in the green or slightly yellow leaves of the CYP82E4:GUS and negative control (promoterless GUS; data not shown) plants, but GUS activity rapidly increased once the leaves reached the later stages senescence (Fig. 1). In addition to the experiments that evaluated the expression pattern conferred by the CYP82E4 promoter in naturally senescent leaves, we also measured GUS activity in detached leaves that were artificially cured in the dark. The results showed that the expression of the CYP82E4 promoter was strongly induced in both the naturally and artificially senesced leaves, but GUS activities reached higher levels in detached leaves cured in the dark (Fig. 1). Senescence-inducible GUS activity was confirmed by histochemical staining of green and senescent leaf tissues. Similar to the MUG assays, the results showed little or no GUS activity in the tissues of the green leaf, while high levels of GUS expression were apparent once the leaves turned yellow (Fig. 2b, c) indicating that the expression directed by the CYP82E4 promoter is tightly correlated with the progression of leaf senescence. A more detailed analysis of the tissue-specific localization of GUS activity revealed that the CYP82E4 promoter was expressed in the epidermis and both the palisade parenchyma (Fig. 2e) and the spongy mesophyll (Fig. 2f), and strong GUS activity was apparent in the trichomes (Fig. 2d). Similar to the lamina, GUS activity was weak in the petiole of a green leaf (Fig. 2g) and strongly upregulated in that of the senescing leaf (Fig. 2h).

Fig. 1
figure 1

Senescence-induced activation of GUS activity in the leaves of 3-month-old DH98-325-6 tobacco line carrying the CYP82E4:GUS construct. An illustration of the leaf colors at the different stages of development is shown in Fig. 2A. Stages 1–4 and 5 represent stages of natural and artificially-induced, complete senescence, respectively. Each data point represents four independent lines measured in two independent replications. Error bars denote standard error of the mean

Full size image
Fig. 2
figure 2

Leaf and flower phenotypes at different stages of development and histochemical localization of GUS activity in various organs. Pictures show tissues of the CYP82E4:GUS_59 line, unless otherwise noted. (a) From left to right, leaf phenotypes at four stages of natural senescence and at artificially-induced, complete senescence. GUS activity in the (b) green leaf (c) senescing leaf (d) trichomes; (e, f) cross sections of senescing leaves; (g) cross section of petioles of a young, green and (h) a senescing leaf; (I) stem cross section of a young, a (j) mature CYP82E4:GUS_59 plant and (k) a young, promoterless GUS-transformed plant; (l) flowers at stage 1, (m) stage 10, (n) stage 13; (o) stamens and pistil. (p) Illustration of the stages of flowering, from left to right, stages 1, 10, 12 and 13 are shown. Abbreviations: c, cortex; ca, cambium; cu, cuticle; e:epidermis; m, mesophyll; p, palisade parenchyma; pi, pith; v, vascular bundle; x, xylem

Full size image

Stem

GUS histochemical analysis revealed that CYP82E4 promoter regulated gene expression is directed to the pith and cortex regions of the stem (Fig. 2i–k). In contrast to the leaf and flower where the CYP82E4 promoter was preferentially expressed in senescing tissues, CYP82E4 promoter-regulated GUS activity was observed throughout the pith and cortex of both young, green (Fig. 2i) and mature, senescing stems (Fig. 2j).

Roots

GUS activity was measured in fresh and stored root tissues collected from young (2-month-old) and mature (flowering) plants grown under the conditions already described (see “Generation of the transgenic plants” section). Regardless of the plant’s age from which the root was harvested or the length of root storage, GUS activity in the roots was very low (Fig. 3).

Fig. 3
figure 3

Fluorometric analysis of GUS activity in the root of CYP82E4:GUS-transformed DH98-325-6 plants at different developmental stages. GUS activity was measured in the roots of 2-month-old, young (Y) and flowering, mature (M) plants after 0 (0 d), 10 (10 d) and 20 (20 d) days of post-harvest storage. Bars represent the GUS activity mean of two independent lines, four independent replications each. GUS activity in the cured leaves (CL) and stage 13 flowers (FL-13) of the same transgenic lines from which the roots were harvested are shown as comparisons. Error bars denote standard error of the mean

Full size image

Flowers

Expression of the CYP82E4 promoter in the flowers was also induced by senescence. GUS activity measurements taken at four different stages of flower development (F1, F10, F12 and F13) showed CYP82E4 promoter activity was low at stages F-1 and F-10, intermediate at stage F-12 and sharply increased to high levels at stage F-13 (Fig. 4). Histochemical staining of the flowers also demonstrated that the expression of the CYP82E4 promoter was regulated by a senescence-dependent pathway in these tissues. Negligible and low levels of staining was observed at stages F-1 and F-10, respectively; however, GUS activity dramatically increased in the senescing tissues at stage F-13 (Fig. 2l–n). In the flower, the highest levels of GUS activity were localized in the sepal, petal, stigma and the anther (Fig. 2o). Collectively, these results indicate that the CYP82E4 promoter is preferentially activated in the aerial parts of the plant where it induces gene expression in the senescing tissues of the leaves and flowers and in the young and senescing tissues of the stem.

Fig. 4
figure 4

Senescence-induced activation of GUS activity in the flowers of the tobacco CYP82E4:GUS_59 line. An illustration of the flowers at the different developmental stages is shown in Fig. 2P. Plants transformed with the promoterless GUS construct were used as controls. The values show the mean of three independent replicates, error bars represent standard error of the mean

Full size image

Expression of the CYP82E4 promoter in converter and nonconverter tobacco

The conversion of tobacco from the nonconverter to the converter phenotype is controlled by an unstable mutation targeting a yet unidentified conversion locus that mediates the transcriptional suppression and activation of the CYP82E4 gene in nonconverter and converter plants, respectively. To investigate whether the converter locus encodes CYP82E4 per se, as opposed to an upstream regulator of the gene, we compared expression patterns of the CYP82E4:GUS construct in converter and nonconverter tobacco. Functional expression of GUS in converter tobacco and the lack of GUS expression in nonconverter tissues would indicate that the converter locus is a distinct gene that regulates CYP82E4 in trans. Alternatively, detection of strong GUS expression in the senescing leaves of both converter and nonconverter plants would demonstrate that the transcriptional activators of CYP82E4 are functional regardless of the nornicotine phenotype of the genetic background and suggest that the unstable mutation targets the CYP82E4 locus. The results of our experiments were consistent with the second scenario. High levels of GUS activity were observed in the senescing leaves of converter and nonconverter plants alike (Fig. 5), providing additional evidence that the CYP82E4 gene likely constitutes the conversion locus.

Fig. 5
figure 5

Fluorometric analysis of GUS activity in the green and artificially cured leaf of CYP82E4:GUS-transformed DH98-325-5 (nonconverter) and DH98-325-6 (converter) tobacco. Bars represent the mean GUS activity of five independent lines measured in two independent replications. Error bars denote standard error of the mean

Full size image

Expression of the CYP82E4 in leaves is senescence-specific

Expression analysis of the CYP82E4 promoter

GUS activity measurements demonstrated that the CYP82E4 promoter confers senescence-induced gene expression in the leaf. Because senescence is controlled by a complex regulatory network that integrates stress responses and includes several signaling molecules and hormones, we sought to determine whether the hormones and stresses that accelerate the onset of senescence exert a direct effect on the expression of the CYP82E4. The plant hormones and signaling molecules employed in the study were ABA, ethylene, JA, salicylic acid, and yeast extract; and the stresses included drought, wounding and TMV infection.

After 48 or 96 h of treatment with ethylene, the 5th (older) leaf of tobacco showed more visible yellowing than the 2nd (younger) leaf of the treated plants and all the leaves of the control plants that were exposed to air (Fig. 6c). In the yellowing 5th leaves of the ethylene-treated plants, CYP82E4 promoter-mediated GUS activity was upregulated compared to the air treated control, in contrast to the green 2nd leaf of the same plant where the rate of GUS activity was not significantly different from that of the controls (Fig. 6a). Similar results were obtained from the experiments in which tobacco leaves were infected with TMV. Relative to the mock-treated control, the hypersensitive response elicited by TMV infection did not cause visible signs of senescence 15 h after the heat treatment, but the TMV-treated leaves were more yellow than the control 96 h after the heat shock (Fig. 7c). Accordingly, GUS activity was not significantly different between the TMV-treated and control leaves 15 h after heat shock, whereas the leaves measured 96 h after heat shock showed greatly enhanced levels of both GUS activity and nicotine conversion (Fig. 7a, b). These results indicate that ethylene and infection by TMV can enhance the expression of the CYP82E4 promoter, but the stimulation can only be detected once the leaves begin to senesce suggesting that the inducing effect of these factors on the expression of CYP82E4 is associated with their ability to accelerate senescence.

Fig. 6
figure 6

GUS activity (a), nicotine conversion (b), and chlorophyll content (c) at different developmental stages of the leaves 48 and 96 h after ethylene treatment. Bars represent the mean values of six individuals of the CYP82E4:GUS_59 line measured in two independent experiments. Error bars denote standard error of the mean. Abbreviations: L2, 2nd leaf from the apex (younger leaf); L5, 5th leaf from the apex (older leaf)

Full size image
Fig. 7
figure 7

GUS activity (a), nicotine conversion (b), and chlorophyll content (c) in the leaf 15 h and 96 h after the heat shock-induced systemic TMV infection. Bars represent the mean values of measurements taken from the treated, nonsenescing leaf (3rd leaf from the apex) of six CYP82E4:GUS_59 individuals in two independent experiments. Mock-treated leaves of the same position were used as controls. Error bars denote standard error of the mean. Abbreviation: TMV, tobacco mosaic virus

Full size image

Because resistance traits to several pathogens including Phytophthora parasitica var. Nicotiana, Pseudomonas syringae pv. tabaci, and Thielaviopsis basicola have been incorporated into the elite tobacco cultivars DH98-325-5 and DH98-325-6, evaluation of the effects of common tobacco pathogens on the expression of the CYP82E4 promoter was not feasible. To simulate pathogen attack, we treated the green leaves with yeast extract, a complex fungal preparation that induces the production of phytoalexins, compounds that serve as key components in the plant’s defense against phytopathogenic microorganisms (Szabo et al. 1999). GUS activity in both the yeast extract-treated and untreated control leaves was very low, comparable to the levels obtained from leaves transformed with the promoterless GUS construct (Fig. 8a). Similarly, the pattern of gene expression from the CYP82E4 promoter in green leaf tissue did not increase significantly in response to other treatments, such as wounding, drought, salicylic acid, ABA and JA (Fig. 8a). Of all treatments tested, senescence was the only factor to which increased CYP82E4 promoter expression could be consistently correlated. These results therefore suggest that the expression of the CYP82E4 promoter in tobacco leaves is senescence-specific.

Fig. 8
figure 8

GUS activity (a) and nicotine conversion (b) in the leaf following various stress and signaling molecule treatments. Bars represent the mean values of measurements taken from the treated, nonsenescing leaf (3rd leaf from the apex) of six CYP82E4:GUS_59 individuals in two independent experiments. Mock-treated leaves located at the same position were used as controls. Error bars denote standard error of the mean. Abbreviations: ABA, abscisic acid; C, control; D, drought; H, hour; Proml, promoterless; SA, salicylic acid; W, wounding; YE, yeast extract

Full size image

Alkaloid analysis

Previous studies suggested that a strong positive correlation exists between the rates of CYP82E4 transcription and nicotine conversion in the senescing leaves of tobacco (Gavilano et al. 2007; Siminszky et al. 2005). To further investigate the effects of various signaling molecules and stresses in CYP82E4 expression, we measured the changes in nornicotine content of the leaves in response to the same biological factors already described. The results showed that the patterns of the CYP82E4 promoter driven GUS expression and nornicotine production were very similar. Relative to the air-treated controls, nicotine conversion in the 5th leaf, but not the 2nd leaf, was significantly induced following 48 or 96 h of ethylene treatment (Fig. 6b). Similar to GUS activity, nicotine conversion did not change significantly in the TMV-infected versus control leaves 15 h after heat shock, but it was significantly upregulated compared to the control in the infected leaves 96 h after the heat treatment (Fig. 7b). Also in agreement with the results of the GUS activity assay, the remaining treatments did not significantly induce nornicotine production (Fig. 8b), confirming the notion that senescence is a major trigger of nornicotine production in converter tobacco.

Discussion

Nornicotine production in converter tobacco is primarily catalyzed by the senescence-inducible nicotine N-demethylase gene CYP82E4 (Gavilano et al. 2006; Siminszky et al. 2005; Xu et al. 2007). To provide new insights into the regulation of CYP82E4, we created a fusion construct between the 2.2 kb upstream regulatory region of CYP82E4 and the GUS reporter gene and analyzed the temporal and spatial patterns of gene expression conferred by the promoter. The results indicated that the CYP82E4 promoter is differentially regulated in various plant organs. Substantial levels of promoter activity were observed in the green, young stems; in contrast, significant GUS activity was only evident in the leaves and flowers during senescence (Figs. 1, 2, 4), and very low GUS expression was detected in the roots regardless of the plant’s age or the length of storage in the dark prior to analysis (Fig. 3). In the stem, expression from the CYP82E4 promoter was localized to the cortex and pith regions, but no GUS activity was observed in the xylem (Fig. 2i, j). The lack of CYP82E4 promoter activity in the root and the xylem is consistent with the observations that nicotine is produced in the root and transported through the xylem to the leaves without undergoing N-demethylation at the site of its production or en route to the leaves (Wernsman and Matzinger 1968).

The expression pattern of the CYP82E4 promoter in the tissues of the young, green stem stands in contrast to that observed in other aerial organs of tobacco, such as the leaves and flowers, where CYP82E4 promoter activity was only detected in the senescing tissues. A possible explanation for this discrepancy could be provided based on our current understanding that the loss of the content of the conductive cells of the vessels and tracheids, a process that occurs during the formation of xylem tracheary elements, involves programmed cell death (reviewed in Fukuda 2000). According to this hypothesis, the expression of the CYP82E4 promoter in young stem may also be related to senescence, as the vessel and tracheid tissues undergo programmed cell death during tracheary element differentiation. This theory, however, is not supported by the relatively uniform expression pattern of the CYP82E4 promoter in the cortex and pith regions of the stem (Fig. 2i) or the work of other investigators who found no indications for the existence of common mechanistic elements in programmed cell death that occurs during leaf senescence and the vacuole collapse-dependent programmed cell death that takes place during tracheary element differentiation (reviewed in Fukuda 2000; Ye and Droste 1996; Ye and Varner 1996). These two lines of evidence, therefore, strongly suggest that the expression of the CYP82E4 promoter in the stem tissues is not senescence-dependent. Interestingly, the expression patterns of CYP82E4 and the Brassica napus LSC54 gene that encodes a metallothionein (Buchanan-Wollaston 1994) are similar in that the expression of both genes is induced by senescence in the leaves, and they are also expressed in the nonsenescent stem and inflorescence meristem, respectively.

To enhance our understanding about the regulatory network that controls nicotine conversion, we tested the effects of senescence, plant hormones, and biotic and abiotic stresses on the rate of nornicotine production and the expression of the CYP82E4 promoter in tobacco leaves. Of all the treatments tested, leaf senescence elicited the largest increase in CYP82E4 promoter expression (Fig. 1). Because previous studies have suggested that artificial senescence upregulates more genes than natural senescence (Becker and Apel 1993; Weaver et al. 1998), we examined the effects of artificial air-curing and natural senescence on nornicotine production and CYP82E4 promoter driven GUS expression. Although sharp upregulation of CYP82E4 expression and nicotine conversion were observed in response to both treatments, the stimulating effect of artificial senescence was greater (Fig. 1). These results are consistent with the findings of Weaver et al. (1998) who reported that dark detachment-induced senescence was the most effective treatment of the factors tested to induce the transcript accumulation of ten out of eleven senescence-associated genes. Expression of the reporter gene was strong in the senescing leaves of both the converter and nonconverter tobacco backgrounds confirming our previously proposed model (Gavilano et al. 2007) that the unstable mutation that controls tobacco conversion from the nonconverter to the converter phenotype directly alters the CYP82E4 locus rather than an upstream regulator of the gene (Fig. 5). Because sequence analysis of CYP82E4 in several different converter and nonconverter backgrounds revealed no polymorphisms (including the promoter region described here), an epigenetic mechanism of gene activation is likely.

Besides natural senescence, only exposure to ethylene and the hypersensitive response to TMV infection induced nornicotine production and the expression of GUS from the CYP82E4 promoter in the leaves (Figs. 6 and 7). However, both of these treatments accelerated leaf senescence, and increases in nicotine conversion and GUS expression were only detected in the visibly yellowing leaves of the ethylene-treated plants or in the yellowing TMV-treated lamina. Induction of nornicotine production or GUS activity was not apparent in the ethylene- or TMV-treated green leaves (Figs. 6 and 7). The observations that ethylene and the TMV infection-elicited hypersensitive response promote leaf senescence are in agreement with a large body of evidence that demonstrate the important roles of these factors in the induction of senescence-associated programmed cell death (reviewed in Buchanan-Wollaston 1997; Lim et al. 2007). The findings that CYP82E4 expression was only stimulated in the older, yellowing leaves indicate that effects of ethylene and TMV infection on CYP82E4 expression are either indirectly exerted through downstream components of the ethylene- and TMV-induced senescence signaling pathway, or that the effects of these agents are direct but require the action of additional factors uniquely present in older leaves (Butt et al. 1998). The existence of such age-specific factors that control ethylene-induced senescence is supported by previously reported data demonstrating that the senescence-promoting effect of ethylene is much greater on the older than the younger leaves of the plant (Grbic and Bleecker 1995).

Nicotine conversion and CYP82E4 promoter-driven GUS expression analyses revealed that a variety of chemical and stress treatments generated similar response profiles of both variables (Figs. 6–8). These results suggest that the 2.2 kb CYP82E4 promoter region used in the reporter construct contains all the cis-acting regulatory elements necessary for the authentic recapitulation of the regulatory attributes of the native CYP82E4 gene. The results also showed that the very low levels of CYP82E4 expression in the green leaves was not affected by the remaining treatments used in the study, such as wounding, drought, salicylic acid, ABA, yeast extract and JA (Fig. 8) confirming the notion that regulation of CYP82E4 is senescence-specific in these tissues. In contrast to the induction pattern of nornicotine production, the synthesis of nicotine has been shown to be triggered by wounding and JA treatments in several Nicotiana species (Baldwin and Ohnmeiss 1993; Baldwin et al. 1994; Halitschke and Baldwin 2003) suggesting that nicotine and nornicotine production are controlled by at least partially non-overlapping signal transduction pathways in Nicotiana. The observation that several cis-acting ABA and drought response motifs (e.g. drought response elements, MYB and MYC recognition sequences) are located in the CYP82E4 promoter (data not shown), yet induction of CYP82E4 by ABA and drought was not evident in our experiments reaffirm the widely reported finding that only a small portion of transcription factor binding sites influence transcription (reviewed in Latchman 1998; Weinzierl 1999). Alternatively, some of these motifs are likely to lie in regions of the 2.2 kb fragment that are not part of the functional promoter. For the same reasons, other binding sites present in the CYP82E4 promoter that have been shown to play important roles in regulating gene expression during senescence (e.g. W-boxes; Eulgem et al. 2000; Miao et al. 2004) potentially, but not necessarily, represent key regulatory elements for the senescence-inducible expression of CYP82E4. Further work will be required to define the specific bounds of the CYP82E4 promoter and the relevance of the specific motifs located therein.

Based on the classification system proposed by Gan and Amasino (1997), SAGs that are exclusively expressed during senescence, and those whose expression is low in the nonsenescent tissues and increases in response to senescence are termed Class I and Class II SAGs, respectively. According to its regulatory properties in the leaves, CYP82E4 belongs with the group of Class I SAGs. The relatively few reports published on the isolation of Class I SAGs suggest that the number of genes that display the regulatory attributes of Class I SAGs is low in the plant genome. Examples of Class I SAGs include the cysteine proteinase SAG12 isolated from Arabidopsis (Lohman et al. 1994; Weaver et al. 1998), and NtCP1 a putative cysteine proteinase derived from tobacco (Beyene et al. 2006). Strong, leaf senescence-specific promoters can be employed as powerful tools in genetic engineering senescence-related processes. For example, Gan and Amasino (1995) showed that delaying the onset of programmed cell death by overexpressing a rate-limiting cytokinin biosynthetic gene under the transcriptional control of the SAG12 promoter extended the photosynthetically active lifespan of the leaf. Other potentially valuable applications of senescence-specific promoters include the plant-based production of high-value natural products and pharmaceuticals that exhibit phytotoxic properties. Restricting the production of these compounds to the senescing leaves can prevent injury to the tissues during the plants’ growth period.

While our data strongly suggest that the expression of CYP82E4 is senescence-specific, the function of the CYP82E4 protein in senescence or in the physiological processes linked to senescence remains unclear. The functions of the SAG-encoded proteins are commonly associated with biochemical processes related to cellular degradation, nutrient translocation, protection from senescence-induced cell damage and pathogen response (reviewed in Buchanan-Wollaston 1997). Although formally not tested, visible differences are not apparent in the morphology, growth and development of converter versus nonconverter tobacco suggesting that nornicotine production has no major direct role in senescence or any other developmental processes. Given the differential insecticidal properties of nicotine and nornicotine (Riah et al. 1997; Siegler and Bowen 1946; Yamamoto et al. 1968), one possibility is that senescence-specific conversion of nicotine to nornicotine facilitates the defense of the structurally weakened, senescing leaf tissue against herbivore attack. Ultimately, more research is needed on the differential toxicity of nicotine alkaloids in various insects to elucidate the individual roles of nicotine and nornicotine in plant defense and senescence.