Plant Molecular Biology 39: 657–669, 1999. © 1999 Kluwer Academic Publishers. Printed in the Netherlands.

657

Developmentally regulated patterns of expression directed by poplar PAL promoters in transgenic tobacco and poplar Madoka Gray-Mitsumune1, Elizabeth K. Molitor2, Daniela Cukovic2, John E. Carlson1,3 and Carl J. Douglas2,∗ 1 Biotechnology

Laboratory and Department of Plant Science, University of British Columbia, Vancouver, BC V6T 1Z4, Canada; 2 Department of Botany, University of British Columbia, Vancouver, BC V6T 1Z4, Canada (∗ author for correspondence); 3 present address: School of Forest Resources and the Life Sciences Consortium, The Pennsylvania State University, University Park, PA 16802, USA Received 22 December; accepted in revised form 9 October 1998

Key words: phenylalanine ammonia-lyase, Populus, tobacco, promoter, xylem, lignin

Abstract Phenylalanine ammonia-lyase (PAL) catalyzes the first step in phenylpropanoid metabolism and plays a central role in the biosynthesis of phenylpropanoid compounds. We have previously cloned two PAL genes, PAL1 and PAL2, from a Populus trichocarpa × P. deltoides F1 hybrid. Here, we describe the properties of PAL1 and PAL2 promoters and their expression patterns in transgenic tobacco and poplar. The promoters were 75% identical in the regions sequenced, and each contained two copies of AC-rich putative cis-acting elements that matched a consensus plant myb transcription factor binding site sequence. In transgenic tobacco, PAL1-GUS and PAL2-GUS fusions directed similar patterns of expression in developing primary xylem of leaves, stems, and other organs, and in secondary xylem of stems. Contrary to previously documented patterns of PAL1/2 expression in poplar, no expression of either fusion was detected in epidermal or subepidermal cell layers of young tobacco leaves or stems. In poplar, the PAL2-GUS fusion directed the highest levels of expression in roots and young leaves and stems. In young leaves and stems, high GUS activity was detected in epidermal or subepidermal cells as well as in primary xylem and phloem fibers. GUS activity was low in woody stems, and was weak or absent in developing secondary xylem. The patterns of PAL2-GUS expression in poplar are very similar to those of PAL1/2 mRNA accumulation in poplar. However, the distinct patterns of expression directed by the PAL2 promoter in poplar and tobacco show that PAL2-GUS expression in tobacco does not accurately reflect all aspects of PAL2 expression in poplar.

Introduction Phenylalanine ammonia-lyase (PAL) catalyzes the first committed step in phenylpropanoid metabolism, the deamination of phenylalanine to form cinnamic acid. Cinnamic acid and its derivatives are then used in the biosynthesis of a large number of phenylpropanoid natural products. PAL activity is important in regulating metabolic flux into phenylpropanoid metabolic pathways, and has been shown to be rate-limiting for The nucleotide sequence data reported will appear in the EMBL, GenBank and DDBJ Nucleotide Sequence Databases under the accession number AF038863 and AF038864.

the biosynthesis of phenylpropanoid compounds [2, 19]. In accordance with its central role in regulating entry into phenylpropanoid metabolism, PAL genes are expressed at multiple times and places during plant development, and in response to environmental stimuli such as wounding, pathogen infection, and UV light irradiation [7, 25, 29, 38, 42, 46]. These expression patterns are generally correlated with sites of accumulation of phenylpropanoid products. For example, PAL mRNA accumulation in the vascular systems of plant organs is associated with the deposition of lignin in xylem tracheary elements, expression in petals is associated with the biosynthesis of anthocyanin pigments,

658 and pathogen-activated expression is associated with the biosynthesis of salicylic acid and an array of other phenolic compounds at infection sites [7, 30]. PAL mRNA accumulation is largely controlled by transcriptional regulation of the corresponding genes [6, 22]. PAL promoters fused to the GUS reporter gene have been shown to specify developmentally regulated and stress-activated expression in transgenic plants in developing xylem, pigmented portions of petals, epidermal and sub-epidermal cell layers of vegetative organs, root tips, and in response to wounding and pathogen infection [5, 21, 26, 30, 34, 43]. Although the expression of PAL-GUS fusions has been largely analyzed in heterologous hosts, the ability of PAL promoters to specify patterns of expression that correlate with sites of phenylpropanoid product accumulation can be taken as evidence that the promoters contain all or most of the information required to specify developmentally and environmentally activated patterns of PAL gene expression. PAL promoters and the promoters of related genes encoding other phenylpropanoid enzymes such as cinnamate-4-hydroxylase (C4H) and 4coumarate:CoA ligase (4CL) contain conserved ACrich cis-acting elements termed P and L boxes, or AC elements [3, 14, 15, 16, 27, 28, 44]. These elements serve as binding sites for transcription factors with myb DNA binding domains [13, 11, 39, 40] and play central roles in the control of developmentally regulated PAL expression [14, 15, 24]. The most abundant phenylpropanoid natural product is lignin, which is derived from the polymerization of hydroxycinnamyl alcohols and deposited in the secondary walls of tracheary elements and sclerified parenchyma cells [9, 45]. The biosynthetic commitment to lignin is particularly large in woody species, in which up to 36% of wood (secondary xylem) dry weight is lignin [45]. Since PAL activity is required for the metabolic flow into lignin precursor biosynthesis, the regulation of PAL expression is important in the context of the control of tracheary element differentiation during secondary growth in woody species. Among trees, PAL genes and proteins have been most extensively studied in the genus Populus (poplars and aspens). As in all plants studied to date, PAL is encoded by a gene family in Populus. Two similar P. deltoides PAL genes (PAL1 and PAL2) were isolated from a P. trichocarpa × P. deltoides F1 hybrid (42; E. Molitor and C. Douglas, unpublished). Segregation analysis of PAL1- and PAL2-specific RFLPs in populations of F2 and back-crossed individuals derived from

F1 hybrids showed that they are located at unlinked genetic loci [42]. Four PAL genes, palg1, palg2a, palg2b, palg4, were independently identified from the P. sieboldii × P. grandidentata hybrid P. kitakamiensis [35, 36]. The coding region of PAL1/2 and palg1 are 92% identical, suggesting that they are orthologous genes. A possible P. trichocarpa × P. deltoides orthologue of the more divergent palg2a/b genes has been isolated using a PCR approach (C. Hutcheon and C. Douglas, unpublished), indicating that the PAL gene family in this hybrid may be similar to that in P. kitakamiensis. RNA blot analysis showed that PAL1/2 is highly expressed in young leaves and stems of poplar trees, with much lower expression in secondary xylem and mature leaves [42]. In situ hybridization showed that exceptionally high PAL1/2 mRNA levels accumulate in epidermal and subepidermal cell layers in young stems and leaves, with lower levels in primary vascular tissues. The high expression of PAL in epidermal/subepidermal cells correlates with the biosynthesis and accumulation of high levels of phenolic compounds in young Populus leaves and stems [10, 18, 37]. Similar to PAL1/2, palg1 is expressed predominately in young stems and leaves [36], while palg2a and pal2b mRNAs accumulate in older stem internode tissues undergoing secondary growth [36]. PAL activity of the protein encoded by the PAL2 cDNA was confirmed by expression in insect cells [31]. In summary, poplar PAL genes similar to PAL1 and PAL2 are highly expressed in vascular tissues and epidermal/subepidermal tissues of immature organs, while expression of another class of PAL genes, represented by palg2a and palg2b, seem to be more highly expressed in older stem tissues. Certain poplar hybrids are amenable to Agrobacterium tumefaciens-mediated transformation (e.g. [23]). This provides a homologous system for the investigation of regulatory sequences controlling expression of poplar genes. In several studies (e.g. [12, 17, 33]), the expression of heterologous promoterreporter gene fusions has been analyzed in transgenic poplar. To our knowledge, however, there are no published reports characterizing the expression of poplar phenylpropanoid or other promoters in transgenic poplar plants. In order to begin a characterization of poplar PAL regulatory sequences, we cloned poplar PAL1 and PAL2 promoters and fused them to the GUS reporter gene. Analysis of GUS expression patterns in transgenic poplar demonstrated that the PAL2 promoter directs expression specifically

659 in young poplar tissues, consistent with patterns of PAL1/2 mRNA accumulation. However, there were important differences between the PAL-GUS expression patterns in poplar and tobacco, demonstrating apparent species-specific differences in the function of the PAL2 promoter.

Materials and methods Gene constructions A PAL2 promoter fragment was isolated as a 1.7 kb SalI-HindIII restriction fragment from pBS-PAL2, a subclone of the previously described genomic clone λgpopPAL18 [42] which contains a copy of PAL2 (Figure 1). The HindIII site at the 30 end of this fragment is 40 bp upstream of the putative ATG initiation codon. A PAL1 promoter fragment, which lacks this HindIII site, was isolated after amplification by the polymerase chain reaction (PCR). A PAL1-specific PCR primer containing an XbaI site was designed to amplify PAL1 DNA starting at the same relative position as the 30 end of the PAL2 promoter fragment. This primer, along with a vector primer, was used to amplify a 1.1 kb PAL1 promoter fragment from pBS-PAL1, a subclone of the previously described genomic clone λgpopPAL23 [42] which contains a copy of PAL1 (Figure 1). An amplified product was cloned into pBluescript (Stratagene) as a SalI-XbaI fragment, and the sequences of the ends of the fragment confirmed to be identical to that of the original genomic sequence. The central, unsequenced portion of the fragment was then removed as an XhoI-NcoI fragment and replaced with a genomic XhoI-NcoI fragment to guarantee the authenticity of the promoter clone. The PAL1 and PAL2 promoter fragments were cloned as transcriptional fusions upstream of the GUS reporter gene in pBIN 19/GUS, a pBIN 19 [4] derivative containing the GUS gene from pRT99-GUS-JD [41] as an XbaI-EcoRI fragment in the pBIN 19 multi-cloning site (Figure 1). These PAL-GUS fusions, pBG1.1. and pBG 1.7 (Figure 1) were introduced into Agrobacterium tumefaciens strain LBA4404 for transfer into tobacco and poplar. DNA sequencing and primer extension Sequencing of PAL1 and PAL2 promoter fragments was carried out manually using the Pharmacia T7 Sequencing Kit and by automated ABI Prism Cycle Sequencing at the University of British Columbia

NAPS unit. Sequences were analyzed using Genetics Computing Group (GCG) software. For primer extension analysis, a 22 bp oligonucleotide (50 CCCAGCTGAGTGGATCACGTTG-30 ) complementary to PAL1 and PAL2 coding sequence 60 bp downstream of their putative initiation codons was endlabeled with 32 P using polynucleotide kinase. Primer extension was carried out using 10 µg total RNA from an elicitor-treated H11 cell culture [32] and AMV reverse transcriptase (Promega) as described [8]. To estimate the size of primer extension products, sequencing reactions of pBS-PAL1 and pBS-PAL2 using the same primer were run on sequencing gels adjacent to primer extension reaction products. RNA blots RNA was extracted from growth chamber-grown P. trichocarpa × P. deltoides clone H11 plants, or from developing secondary xylem of field-grown H11 plants using Trizol reagent (Gibco) or as described by Allina et al. [1]. RNA blots were prepared and hybridized at high stringency to a PAL7 cDNA probe as described [1]. Agrobacterium-mediated transformation Agrobacterium-mediated transformation of tobacco (Nicotiana tabacum SR1) was performed as described [16]. Poplar transformation was carried out with modifications of the method Leple et al. [23] using a Populus tremula × P. alba INRA clone 717-1B4; [23]) maintained in vitro on half-strength Woody Plant Medium (WPM, Sigma), supplemented with 0.5% agar, 1.5% sucrose, and 200 mg/l L-glutamine. Leaf explants (1–2 cm2 ) or petiole pieces (1 cm) were pre-cultured on WPM medium containing 0.5% agar, 3% sucrose, 200 mg/l L-glutamine, 1 µM BAP and 0.5 µM NAA (SIM1) for 2 days at 25 ◦ C and were co-cultivated with an A. tumefaciens strain containing pBG1.7 as described [23]. After co-cultivation, explants were washed with 1 g/l carbenicillin in WPM, blotted dry onto a sterile filter paper and incubated on SIM1 containing 500 mg/l carbenicillin for 2 weeks at 25 ◦ C in the dark. Explants were then placed on SIM1 containing 500 mg/l carbenicillin and 100 mg/l kanamycin for 2 weeks at 25 ◦ C (16 h light/8 h dark photoperiod). For shoot induction, explants were transferred to SIM2 (WPM medium containing 0.5% agar, 3% sucrose, 200 mg/l L-glutamine and 0.1 µM thidiazuron) containing 500 mg/l carbenicillin and 100 mg/l kanamycin (SIM2 selection

660

Figure 1. Structure of poplar PAL1 and PAL2 genes and generation of PAL-GUS gene fusions. Horizontal arrows indicate locations of PCR primers; vertical dashed arrows indicate cloning steps; bent arrows indicate transcription start sites. Open boxes indicate λ vector sequences. B, BamHI; R, EcoRI; S, SalI; X, XhoI; Xba, XbaI.

medium) and incubated at 25 ◦ C under a 16 h light/8 h dark photoperiod; explants were transferred to fresh SIM2 selection medium every fourth week. To induce roots, shoots were removed, transferred to RIM (half strength of WPM containing 0.5% agar, 1.5% sucrose and 200 mg/l L-glutamine) containing 500 mg/l carbenicillin and 100 mg/l kanamycin and incubated at 25 ◦ C at 16 h light/8 h dark photoperiod. Rooted transgenic plants were maintained in vitro on RIM containing carbenicillin (500 mg/l) and kanamycin (100 mg/l). Putative transgenic lines were screened for GUS activity as described below; GUS-positive lines were maintained in vitro, transferred to soil when ca. 5 cm in height, and maintained in a growth chamber at 22 ◦ C under a 16 h light/8 h dark photoperiod. GUS assays Histochemical staining of growth chamber-grown transgenic tobacco plants and 2-month old growth chamber-grown transgenic poplar plants was performed as described by Jefferson et al. [20]. Leaves and stems were hand-sectioned and fixed for 30 min in 0.5% paraformaldehyde in 0.1 M sodium phosphate

(pH 7.0) before staining with 1.5 mM 5-bromo-4chloro-3-indol-1-glucuronide (X-Gluc). After staining, chlorophyll was removed with ethanol, ethanol was replaced stepwise with 50% glycerol, and sections were mounted for observation under a light miscroscope. To quantify GUS activity, tissues were homogenized in extraction buffer (10 mM (β-mercaptoethanol), 10 mM Na2 EDTA, 0.1% v/v Triton X-100, 50 mM sodium phosphate buffer, pH 7.0) with a pinch of sand, followed by centrifugation to remove debris. An aliquot of the supernatant was added to the GUS reaction mixture (1 mM 4-methylumbelliferyl glucuronide (MUG), 20% v/v methanol in extraction buffer) and incubated at 37 ◦ C. The reaction was stopped by adding 0.2 M Na2 CO3 . Enzymatic conversion of MUG to 4-methylumbelliferone (4-MU) was measured using a fluorometer (365 nm excitation and 455 nm emission) and GUS activity was expressed as the concentration of 4-MU released per minute per mg protein. Protein concentrations were determined using a Bradford Protein Assay Kit (BioRad, Mississauga, Ontario) and bovine serum albumin as a standard.

661 Lignin staining Stems of transgenic poplar were hand-sectioned and fixed as described above. Fixed tissue sections were stained for lignin using phloroglucinol/HCl (5% in an ethanol/HCl mix 9:1) and immediately observed.

Results Characterization of PAL1 and PAL2 promoters Genomic clones containing P. deltoides alleles of PAL1 and PAL2, isolated from a P. trichocarpa × P. deltoides genomic library, have been previously described [42]. We subcloned restriction fragments containing the PAL1 and PAL2 genes, including ca. 1.1 kb and 1.7 kb of 50 -flanking DNA, respectively, into plasmid vectors to facilitate analysis of the respective promoters (Figure 1). About 600 bp of PAL1 and PAL2 DNA upstream of the putative ATG initiation codons was sequenced. The PAL1 and PAL2 transcription start sites used in elicitor-treated tissue culture cells were determined by primer extension (data not shown), but we cannot exclude the possibility that slightly different start sites are used in other organs. These data are summarized in Figure 2. Starting 20 bp downstream of its transcription start site, the PAL2 sequence was identical to that of the PAL cDNA clone PAL7, isolated from a P. trichocarpa × P. deltoides clone H11 cDNA library [42]. This confirms that the PAL7 cDNA is a product of the PAL2 gene. Upstream of their transcription start sites, the PAL1 and PAL2 promoter sequences were ca. 75% identical when several gaps were introduced to maximize alignment; within 300 bp of their transcription start sites, the promoters showed higher levels of similarity. Within this region the promoters contained putative TATA boxes and two AC elements, cis-activing elements conserved in the promoters of PAL and other phenylpropanoid genes (Figure 2). The sequences of these AC elements were 100% identical to the binding site of a myb-related transcription factor, the product of the maize P gene (13) and very similar to the sequences of AC elements within PAL and other phenylpropanoid gene promoters (Table 1). Expression of PAL1-GUS and PAL2-GUS fusions in transgenic tobacco Transcriptional fusions of the 1.1 kb PAL1 promoter and the 1.7 kb PAL2 promoter to the uidA (GUS)

Table 1. Sequences of conserved AC cis-elements in poplar PAL1 and PAL2 genes and related phenylpropanoid genes. Sequence

Plant/Gene/Elementa

Reference

ACCAACCA ACCAACCA ACCTACCA ACCTACCA T A ACCAACCC ACAAACCC ACCTACCA ACCTACCA ACCAACCC ACCTAACT ACCAACCC ACCAACCC

poplar PAL1 AC-I (−264) poplar PAL2 AC-I (−250) poplar PAL1 AC-II (−86) poplar PAL2 AC-II (−76) maize P (Myb) core binding site parsley PAL1 P Box parsley PAL1 L Box bean PAL1 AC-I bean PAL1 AC-II bean PAL1 AC-III parsley 4CL1 FP4 parsley 4CL1 FP8

this paper this paper this paper this paper 13 27 27 14 14 14 16 16

a The positions of poplar PAL elements are given with respect

to transcription start sites.

reporter gene were generated in the pBIN19 binary vector and designated pBG1.1 and pBG1.7, respectively (Figure 1). Following transformation of tobacco with Agrobacterium strains harboring the PAL-GUS fusions, ten pBG1.1 and eight pBG1.7 transformants were analyzed. Segregation of kanamycin resistance in the progeny of self-pollinated primary transformants suggested that the number of transgene loci ranged from one to three or greater in these transformants (data not shown). Histochemical assay of GUS activity in the primary transformants and F1 progeny showed that GUS activity was consistently localized to the xylem in the vascular systems of stems, leaves, roots, and flowers, and that there were no apparent differences in the expression patterns directed by the promoters. All lines were examined for GUS expression using the histochemical assay, and representative results showing PAL1-GUS (pBG1.1) and PAL2-GUS (pBG1.7) expression in the xylem of leaf and stem tissues are shown in Figure 3. As previously observed with other phenylpropanoid promoters [5, 16, 26, 43], PAL1GUS AND PAL2-GUS were expressed in the primary xylem of leaves (Figure 3A, B) and stems (not shown), and in ray parenchyma cells of secondary xylem (Figure 3C, D). In developing leaves and stems, there was a notable lack of expression in epidermal or subepidermal cell layers, even in the most highly expressing lines (Figure 3). This is in contrast to the high levels of PAL1/2 expression in these cell layers in young poplar leaves and stems [42]. GUS expression was also lack-

662

Figure 2. Promoter sequences of PAL1 and PAL2. Alignment of the PAL1 and PAL2 sequences is shown, with gaps (dots) introduced to maximize alignment. Bases corresponding to the putative ATG initiation codons, TATA boxes, transcription start sites, and the start of identity with the PAL7 cDNA are indicated in bold. The transcription start sites are also indicated by a bent arrow. Shaded boxes show the locations of two AC elements in each sequence. Numbering of bases is with respect to the transcription start sites.

ing in root tips, and was not reproducibly observed in non-vascular tissue of any floral organs except the surface of ovules (data not shown). Thus, both promoters directed expression primarily to the primary and secondary xylem of tobacco. Expression of the PAL2-GUS fusion in transgenic Populus The GUS expression patterns specified by the PAL1 and PAL2 promoters in tobacco leaves and stems differed from the pattern of PAL1/2 expression in poplar, where high levels of PAL1/2 mRNA accumulate in the epidermal/subepidermal cells of leaves and stems [42]. This discrepancy could be due to the absence of critical cis-acting elements in the PAL1 and PAL2 promoters as isolated, or because certain PAL1/2 cisacting elements do not function properly in the het-

erologous tobacco host. In order to clarify this point, we investigated the activity of the expression of the poplar PAL2-GUS fusion in transgenic poplar. After Agrobacterium-mediated transformation, 24 GUS Populus tremula × P. alba plants transgenic for pBG1.7 (PAL2-GUS) were obtained. The GUS activity in leaves in vitro grown plants 5 to 7 cm tall ranged from 0.16 to 26.24 nmol/min per mg protein (Figure 4A). The relative levels of GUS activity in leaves, stems, and roots of four of these lines examined were similar, with highest activities in roots, followed by stems and leaves (Figure 4B). GUS activities were next assayed in six growth chamber-grown transgenic lines two months after transfer to soil, at which time the plants were ca. 40 cm tall. GUS activities in stem and leaf samples were measured and were expressed relative to the activity in the upper one cm of stem tissue from each line.

663

Figure 3. Histochemical localization of GUS activity transgenic tobacco plants transformed with PAL1-GUS and PAL2-GUS fusions. A. Young leaf cross-section of a plant transgenic for pBG1.1. B. Young leaf cross section of a plant transgenic for pBG1.7. C. Cross-section through a mature stem of a plant transgenic for pBG1.1. D. Cross-section through a mature stem of a plant transgenic for pBG1.7.

This facilitated comparison of the relative GUS expression levels at different developmental stages using data from all six lines. Similar to in vitro grown plants, GUS activity in all lines was highest in roots (data not shown), followed by stems and leaves. Figure 5 shows that expression in leaves and stems varied according to developmental age in all lines. GUS activity was highest in immature leaves and stems within 5 cm of the shoot apex and consistently decreased in the transgenic lines as leaves and stems matured. These levels were several-fold lower in fully expanded leaves (20 cm from the apex) and in stems undergoing significant amounts of secondary growth (30–40 cm from the apex) compared to the youngest stems and leaves. The levels of GUS expression directed by the PAL2 in different poplar organs are consistent with the previously reported preferential accumulation of PAL1/2 mRNA in young leaves and stems of poplar, relative to older organs [42]. In order to confirm that these GUS activity levels accurately reflect to sites of maximal PAL1/2 expression, we hybridized a blot containing RNA from various poplar organs with a PAL2 cDNA (PAL7) probe. Figure 6 shows that, as previously reported, PAL1/2 RNA was relatively abundant in young stems, young leaves, and elicitor-treated cells, and was barely detectable in secondary xylem and old leaves. In addition, this blot showed that PAL1/2 RNA was relatively abundant in roots, where the PAL2 pro-

Figure 4. Expression of the PAL2-GUS fusion gene in in vitro grown transgenic poplar. Transgenic poplar plants grown in vitro were harvested when they were 5–7 cm tall. Samples were extracted and assayed for GUS activity as described in Materials and methods. A. Variation in GUS activity among transgenic lines. Leaves (1–3 cm in length) from the individual transformants indicated were assayed for GUS activity. GUS activity of line 6.1 (26.24 nmol MU/min per mg protein) is indicated at the top of the bar. B. GUS activity in different organs of selected transgenic lines. Activity was measured in roots (open bars), stems (solid bars) and leaves (stippled bars).

moter fusion directed high levels of GUS expression (Figure 4). Thus, the organ-specific GUS expression patterns specified by the PAL2 promoter closely mirror sites of PAL2 mRNA accumulation. To determine the cell and tissue-specific patterns of expression directed by the PAL2 promoter, GUS activity was assayed histochemically in the 2-month old growth chamber-grown lines at the same locations where it was assayed quantitatively. The sites of sampling for histochemical assays are indicated on Figure 5, and representative results from line 9.2 are shown in Figure 7. Similar staining patterns were ob-

664

Figure 5. Expression of the PAL2-GUS fusion gene in soil-grown transgenic poplar. Leaves and stem segments at different distances from apex of transgenic lines 1.1, 1.2, 4.5.1, 6.2, and 8.2 ca. 40 cm in height (center photograph) were harvested and assayed for GUS activity. GUS activities in stems (a) and GUS activities in leaves (b) are shown relative to activity in the highest expressing sample, the youngest stem segment (<1 cm from apex). Values are the mean of 6 transgenic lines ±SE. Letters in a and b refer to locations from which stem and leaf tissue was taken for histochemical assay of GUS activity, and refer to panels shown in Figure 6.

served in lines 1.1 and 5.1. Cross-sections through the youngest leaves at the shoot apex revealed very high GUS activity of the PAL2 promoter in epidermal and/or subepidermal layer of leaf midrib, as well as in the developing leaf blade (Figure 7A). Vascular tissues were not apparent in stained sections at this stage. Expanding young leaves one cm from the shoot apex showed more distinct histochemical staining in epidermal/subepidermal cells of the midrib, strong staining in the developing xylem and phloem of the midrib, and strong staining in the developing leaf blade (Figure 7B). Weaker staining was detected in parenchyma cells of the midrib. The first fully expanded young leaves (20 cm from the apex) exhibited a staining pattern similar to that in expanding young leaves, except that staining was generally weaker and restricted to vascular tissues and epidermal/subepidermal cells (data not shown). In petioles, strong GUS activity was observed in epidermal/subepidermal cells and was associated with developing xylem vessels and phloem fibers (Figure 7D). Histochemical staining of similar petiole sections with phloroglucinol to show sites of lignin deposition indicated that the xylem vessels were

heavily lignified, while large amounts of lignin were not yet detectable in phloem fibers (Figure 7C). Phloroglucinol-stained stem sections taken 1,5 and 30 cm from the apex for lignin are shown in Figure 7E, G, H, and similar sections stained for GUS activity are shown in Figure 7F, H, J. In stems one cm from the apex, primary xylem was just beginning to form in vascular bundles, which could be recognized by the presence of scattered lignified vessels (Figure 7E). At this stage, strong GUS activity was detected in primary xylem and phloem, and in the ring of interfasicular cambium (Figure 7F). Staining was also obvious in epidermal/subepidermal cells. Cross-sections of stems at 5 cm from apex were characterized by distinct bundles of developing, lignified primary xylem and a ring of differentiating phloem fiber cells (Figure 7G, H). In these sections, GUS activity was specific to vascular tissues, with little or no activity in epidermal/subepidermal cells. The most intense staining was observed in cells adjacent to phloem fibers, the interfasicular cambium, and in parenchyma cells near primary xylem vessels. At 30 cm from the apex, secondary xylem occupied the major part of the stem, and cells in the

665 secondary xylem and phloem fibers were highly lignified (Figure 7I). Relative to younger stems, staining for GUS activity at this stage was weak (Figure 7J). Histochemically detectable GUS activity was present in parenchyma cells of the primary xylem and in phloem fiber cells (Figure 7J). Very weak or no activity was detected in ray parenchyma cells of the secondary xylem and in the cambium.

Discussion The GUS reporter gene system has been very useful in elucidating patterns of spatial and temporal expression directed by phenylpropanoid gene promoters at the cell and tissue level. Fusions of the GUS reporter gene to PAL [5, 21, 26, 34, 43], cinnamate-4-hydroxylase (C4H; [3], 4CL [16, 38]) and cinnamyl alcohol dehydrogenase (CAD; [12,17]) genes have shown that the sites of promoter activity are generally consistent with expected sites of phenylpropanoid product accumulation (e.g. lignin deposition in xylem; anthocyanin accumulation in flower petals). In only a few cases, however, has the cell and tissue type-specific expression of such promoter-GUS fusions been shown to correlate with the cell and tissue type-specific accumulation of mRNA for the corresponding genes [17, 38]. Furthermore, with few exceptions (e.g. [3, 34]), these studies have involved transfer of promoter constructs into heterologous hosts (tobacco, Arabidopsis, poplar, potato). There are surprisingly few studies in which developmentally regulated expression directed by phenylpropanoid gene promoter-GUS fusions has been analyzed in homologous hosts, and few, if any, studies in which such expression patterns have also been compared to expression patterns in a heterologous host. In this study, we analyzed the expression of poplar PAL-GUS fusions in both tobacco and poplar. To the best of our knowledge, this is the first example of the functional test of a poplar phenylpropanoid promoter in transgenic poplar. In poplar, the PAL2-GUS fusion specified organ-specific patterns of expression similar to patterns of PAL1/2 mRNA localization (Figure 6; [42]). Similar to PAL1/2, mRNA levels, GUS activity was highest in stem tissue near the shoot apex and in roots, somewhat lower in young leaves at the shoot apex, and much lower in older leaves and older stems undergoing secondary growth (Figures 4 and 5). The cell type-specific expression of the GUS fusion was quite similar to that observed by in situ

Figure 6. RNA blot analysis of PAL1/2 mRNA accumulation in poplar. Total RNA was extracted from the organs indicated, or from control (eli−) and elicitor-treated (eli+) poplar cell cultures, separated on a formaldehyde agarose gel, transferred to a nylon membrane, and hybridized to a poplar PAL7 cDNA probe. Equal amounts of RNA were loaded in each lane, as judged by ethidium bromide staining of the gel.

hybridization in poplar [42]: in young leaves, petioles, and stems, GUS activity was prevalent in epidermal or subepidermal cell layers as well as in developing vascular tissues, while relatively little GUS expression was detected histochemically in stems undergoing secondary growth. In young leaves <1 cm in length, very high levels of GUS activity may have led to overstaining of leaf tissue in Figure 7A; nevertheless, GUS activity in the mid-rib seemed to be associated with epidermal or subepidermal cell layers, and the very high expression in the developing blades of these young leaves is consistent with high levels of mRNA observed in this tissue by in situ hybridization [42]. The lack of strong PAL2-GUS activity in secondary xylem is also consistent with the in situ hybridization data. In particular, GUS activity was clearly weak or absent in ray parenchyma cells of the secondary xylem, where CAD-GUS expression has been detected in hand-sectioned poplar stems [12]. These data show that the PAL2 promoter contains information required to specify developmentally regulated patterns of PAL1/2 expression in poplar previously defined by RNA blots and in situ hybridization. Two aspects of this expression are distinct from that specified by other phenylpropanoid promoter-GUS fusions in heterologous hosts such as tobacco: very high levels of expression in epidermal/subepidermal cells in young leaves and stems, and low expression in ray parenchyma cells of secondary xylem. In tobacco, the pattern of expression specified by the PAL-GUS fusions appeared to be different from that in poplar. In particular, GUS activity was not observed in epidermal or subepidermal cells of developing leaves and stems of any of the 18 transgenic tobacco lines examined,

666

Figure 7. Histochemical assay of PAL2-GUS activity in transgenic poplar. Cross-sections from different tissues of 2-month old soil-grown transgenic poplars were taken from the locations indicated in Figure 5 (letters a, b, e–j in Figure 5 correspond to panels A, B, E–J). A, B, D, F, H, J, sections stained for GUS activity; C, E, G, I, sections stained for lignin using phloroglucinol. A. Cross section through a young leaf 0.5 cm from apex. B. Cross section of an expanding young leaf 1 cm from apex. C, D. Cross sections of petioles of an expanding leaf. E, F. Cross sections of stems one cm from apex. G, H. Cross sections of stems 5 cm from apex; (I, J) Cross sections of stems 30 cm from apex. e, epidermal/subepidermal cells; ic, interfasicular cambium; pf, phloem fibers; px, primary xylem parenchyma cells. Bars = 200 µM.

667 and GUS activity was present not only in developing primary xylem of leaves and stems but also in ray parenchyma cells of secondary xylem of these lines (Figure 3). Thus, the poplar PAL gene fusions are regulated differently in the two hosts, and the pattern of PAL-GUS activity observed in tobacco does not faithfully reflect all aspects of PAL1/2 expression in poplar. The lack of detectable epidermal/subepidermal GUS activity in tobacco could be explained by overall low expression of the PAL-GUS transgenes. However, epidermal expression was never observed in young leaves of tobacco lines that exhibited strong xylem-specific expression. In contrast, epidermal/subepidermal GUS activity was observed in young leaves of all poplar lines tested. Furthermore, the GUS activity levels in the seedlings of more highly expressing PAL2-GUS tobacco lines (data not shown) were similar to those observed in tobacco lines transgenic for other PALGUS fusions [14, 15, 24]. Thus, the poplar PAL1 and PAL2 promoters appear capable of directing cell typespecific expression in tobacco, but not in epidermal cells of young leaves. The differences in expression specified by the PAL2 promoter in poplar and tobacco could be explained by host-specific differences in the distribution and/or activity of transcription factors that regulate expression from this promoter. Cellular signals that regulate PAL1/2 promoter activity in epidermal/subepidermal cells of poplar young leaves and stems may be absent in similar tobacco cell types. The specialized expression of PAL1/2 genes in epidermal/subepidermal cells of young poplar tissues is probably an adaptation associated with very high phenylpropanoid biosynthetic activity in these tissues [10, 18, 37]. This specialization is likely to rely on the activity of transcription factors present in these cell types in poplar but absent or inactive in tobacco, where equivalent phenylpropanoid metabolic activity is lacking. Conversely, transcription factors capable of activating expression of the PAL2 promoter may be absent or inactive in ray parenchyma cells of poplar secondary xylem, but present and capable of activating PAL1/2 expression in the ray parenchyma cells of tobacco secondary xylem. The different relative activity of the PAL2 promoter in primary and secondary xylem of poplar (high in primary xylem, low in secondary xylem; Figure 7) is consistent with the pattern of expression of the Populus kitakamiensis palg1 gene [35, 36], which is similar to PAL1/2 genes in promoter and coding sequences. These similarities in sequences and ex-

pression patterns suggest that palg1 and PAL1/2 are functional equivalents. Two divergent P. kitakamiensis PAL genes, palg2a and palg2b (75% nucleic acid identity to coding regions of palg1 and PAL1/2) appear to be more highly expressed in stems undergoing secondary growth [35, 36], and lack significant similarity to the PAL1 and PAL2 promoters. The palg2a/b-like sequences in P. trichocarpa × P. deltoides hybrids (C. Hutcheon and C. Douglas, unpublished data) may represent PAL genes in this hybrid that are more highly expressed in stems undergoing secondary growth. The strong PAL2 promoter activity in poplar primary xylem relative to quite weak activity in secondary xylem indicates that the signals controlling phenylpropanoid biosynthetic activity during development of primary and secondary xylem are at least partially distinct. Apparently, signals that activate PAL2 expression in the primary xylem of poplar young stems and leaves are absent in developing secondary xylem, where this gene is largely silent. Conversely, expression of different PAL gene family members with divergent promoters (palg2a/b) seems to be more prevalent in stems where secondary xylem is differentiating [35, 36], again indicating that different signals for PAL gene activation are generated in differentiating secondary xylem. This suggests that there may be variation in the regulatory circuits that activate phenylpropanoid gene expression during the lignification phase of xylem differentiation in woody plants such as poplar. The PAL promoters analyzed here are similar to those of several other PAL genes that have been examined in that they contain copies of conserved AC elements [14, 15, 28, 35, 36], also called P and L boxes [27]. Since such elements have been shown to be binding sites for myb transcription factors [11, 13, 39, 40], it is likely transcriptional regulators of this class are involved in controlling PAL expression in poplar. Thus, different myb proteins, or interactions of a given myb protein with different additional regulatory proteins, may regulate expression of different PAL genes at different stages of development (e.g. primary xylem vs. secondary xylem) in poplar. In summary, our data show that the poplar PAL2 promoter contains the information required to specify the patterns of developmentally regulated expression characteristic of this gene in poplar. The discrepancy between the patterns of expression directed by the promoter in tobacco and poplar indicates that there are species-specific differences in the way the promoter functions in these two plants. Future charac-

668 terization of the transcription factors responsible for directing poplar PAL expression to differentiating primary or secondary xylem will aid in elucidating regulatory mechanisms underlying these developmental processes.

Acknowledgements We are grateful to Nancy Mah and Janis Pereira for expert technical assistance and to the Institut National de la Recherche Agronomique, France, for supplying hybrid poplar clone INRA 717-1B4. This work was supported by Natural Sciences and Engineering Research Council of Canada (NSERC) grants to J.E.C. and C.J.D., and by a Science Council of British Columbia GREAT graduate fellowship to M.G.-M.

12.

13.

14.

15.

16.

17.

18.

References 19. 1. Allina SM, Pri-Hadash A, Theilmann DA, Ellis BE, and Douglas CJ: 4-coumarate:enzyme A ligase in hybrid poplar. Properties of native enzymes, cDNA cloning, and analysis of recombinant enzymes. Plant Physiol 116: 743–754 (1998). 2. Bate NJ, Orr J, Ni W, Meromi A, Nadler-Hassar T, Doerner PW, Dixon RA, Lamb CJ, Elkind Y: Quantitative relationship between phenylalanine ammonia-lyase levels and phenylpropanoid accumulation in transgenic tobacco identifies a rate-determining step in natural product synthesis. Proc Natl Acad Sci USA 9108: 7608–7612 (1994). 3. Bell-Lelong DA, Cusumano JC, Meyer K, Chapple C: Cinnamate-4-hydroxylase expression in Arabidopsis: regulation in response to development and the environment. Plant Physiol 113: 729–738 (1997). 4. Bevan M: Binary Agrobacterium vectors for plant transformation. Nucl Acids Res 12: 8711–8721 (1984). 5. Bevan M, Shufflebottom D, Edwards K, Jefferson R, Schuch W: Tissue- and cell-specific activity of a phenylalanine ammonia-lyase promoter in transgenic plants. EMBO J 8: 1899–1906 (1989). 6. Chappell J, Hahlbrock K: Transcription of plant defence genes in response to light and elicitor. Nature 311: 76–78 (1984). 7. Dixon RA, Paiva NL: Stress-induced phenylpropanoid metabolism. Plant Cell 7: 1085–1097 (1995). 8. Douglas CJ, Hoffmann H, Schulz W, Hahlbrock K: Structure and elicitor or u.v.-light-stimulated expression of two 4coumarate:CoA ligase genes in parsley. EMBO J6: 1189–1195 (1987). 9. Douglas, CJ: Phenylpropanoid metabolism and lignin biosynthesis: from weeds to trees. Trends Plant Sci 1: 171–178 (1996). 10. English S, Greenaway W, Whatley FR: Analysis of phenolics of Populus trichocarpa by GC-MS. Phytochemistry 30: 531– 533 (1991). 11. Feldbruegge M, Sprenger M, Hahlbrock K, Weisshaar B: Pcmyb1, a novel plant protein containing a DNA-binding domain with one Myb repeat, interacts in vivo with a lightregulatory promoter unit. Plant J 11: 1079–1093 (1997).

20. 21.

22.

23.

24.

25.

26.

27.

28.

Feuillet C, Lauvergeat V, Deswarte C, Pilate G, Boudet A, Grima-Pettenati J: Tissue- and cell-specific expression of a cinnamyl alcohol dehydrogenase promoter in transgenic poplar plants. Plant Mol Biol 27: 651–667 (1995). Grotewold E, Drummond B, Bowen B, Peterson T: The mybhomologous P gene controls phlobene pigmentation in maize floral organs by directly activating a flavonoid biosynthetic gene subunit. Cell 76: 543–553 (1994). Hatton D, Sablowski R, Yung M H, Smith C, Schuch W, Bevan M: Two classes of cis sequences contribute to tissue-specific expression of a PAL2 promoter in transgenic tobacco. Plant J 7: 859–876 (1995). Hatton D, Smith C, Bevan M: Tissue-specific expression of the PAL3 promoter requires the interaction of two conserved cis sequences. Plant Mol Biol 31: 393–397 (1996). Hauffe KD, Paszkowski U, Schulze-Lefert P, Hahlbrock K, Dangl L, Douglas CJ: A parsley 4CL-1 promoter fragment specifies complex expression patterns in transgenic tobacco. Plant Cell 3: 435–443 (1991). Hawkins S, Samaj J, Lauvergeat V, Boudet A, Grima-Pettenati J: Cinnamyl alcohol dehydrogenase: identification of new sites of promoter activity in transgenic poplar. Plant Physiol 113: 321–325 (1997). Hegnauer R: Chemotaxonomie der Pflanzen, vol 6, pp. 241– 259. Birkhäuaser-Verlag (1973). Howles PA, Sewalt VJH, Paiva NL, Elkind Y, Bate NJ, Lamb C., Dixon RA: Over-expression of L-phenylalanine ammonialyase in transgenic tobacco plants reveals control points for flux into phenylpropanoid biosynthesis. Plant Physiol 112: 1617–1624 (1996). Jefferson RA: Assaying chimeric genes in plants: The GUS gene fusion system. Plant Mol Biol Rep 5: 387–405 (1987). Kawamata S, Shimoharai K, Imura Y Ozaki M Ichinose Y, Shiraishi T, Kunoh H, Yamada T: Temporal and spatial pattern of expression of the pea phenylalanine ammonia-lyase gene1 promoter in transgenic tobacco. Plant Cell Physiol 38: 792– 803 (1997). Lawton M, Lamb C: Transcriptional activation of plant defense genes by fungal elicitor, wounding and infection. Mol Cell Biol 7: 335–341 (1987). Leple J, Brasileiro A, Michel M, Delmotte F, JoLianin L: Transgenic poplars: expression of chimeric genes using different constructs. Plant Cell Rep 11: 137–141 (1992). Leyva A, Liang X, Pintor-Toro JA, Dixon RA, Lamb, C J: cis-element combinations determine phenylalanine ammonialyase gene tissue-specific expression patterns. Plant Cell 4: 26–271 (1992). Liang XW, Dron M, Cramer CL, Dixon RA, Lamb CJ: Differential regulation of phenylalanine ammonia-lyase genes during plant development and by environmental cues. J Biol Chem 264: 14486–1442 (1989). Liang XW, Dron M, Schmid J, Dixon RA, Lamb CJ: Developmental and environmental regulation of a phenylalanine ammonia-lyase-beta-glucuronidase gene fusion in transgenic tobacco plants. Proc Natl Acad Sci USA 86: 9284–9288 (1989). Logemann E, Parniske M, Hahlbrock K: Modes of expression and common structural features of the complete phenylalanine ammonia-lyase gene family in parsley. Proc Natl Acad Sci USA 92: 5905–5909 (1995). Lois R, Dietrich A, Hahlbrock K, Schulz W: A phenylalanine ammonia-lyase gene from parsley: structure, regulation and identification of elicitor and light responsive cis-acting elements. EMBO J 8: 1641–1648 (1989).

669 29. Lois R, Hahlbrock K: Differential wound activation of members of the phenylalanine ammonia-lyase and 4coumarate:CoA ligase gene families in various organs of parsley plants. Z Naturforsch C 47: 90–94 (1992). 30. Mauch-Mani B, Slusarenko A: Production of salicylic acid precursors is a major function of phenylalanine ammonialyase in the resistance of Arabidopsis to Peronospora parasitica. Plant Cell 8: 203–212 ( 1996). 31. Mckegney G, Butland S, Theilmann D, Ellis B: Expression of poplar phenylalanine ammonia-lyase in insect cell cultures. Phytochemistry 41: 1259–1263 (1996). 32. Moniz de Sá M, Subramaniam R, Williams FE, Douglas CJ: Rapid activation of phenylpropanoid metabolism in elicitor treated hybrid poplar (Populus trichocarpa Torr and Gray × Populus deltoides Marsh) suspension-cultured cells. Plant Physiol 98: 728–737 (1992). 33. Nilsson O, Little CHA, Sandberg G, Olsson O: Expression of two heterologous promoters. Agrobacterium rhizogenes RolC and cauliflower mosaic virus 35S, in the stem of transgenic hybrid aspen plants during the annual cycle of growth and dormancy. Plant Mol Biol 31: 887-895 (1996). 34. Ohl S, Hedrick SA, Chory J, Lamb CJ: Functional properties of a phenylalanine ammonia-lyase promoter from Arabidopsis. Plant Cell 2: 837–848 (1990). 35. Osakabe Y, Osakabe K, Kawai S, Katayama Y, Morohoshi, N: Characterization of the structure and determination of mRNA levels of the phenylalanine ammonia-lyase gene family from Populus kitakamiensis. Plant Mol Biol 28:1133–1141 (1995). 36. Osakabe Y, Osakabe K, Kawai S., Katayama Y, Morohoshi N: Structure and tissue-specific expression of genes for phenylalanine ammonia-lyase from a hybrid aspen. Plant Sci 105: 217–226 (1995). 37. Pearl IA, Darling SF: Phenolic extractives of leaves of Populus balsamifera and of P. trichocarpa. Phytochemistry 10: 2844– 2847 (1971)

38.

39.

40.

41.

42.

43.

44.

45. 46.

Reinold S, Hahlbrock K: Biphasic temporal and spatial induction patterns of defense-related mRNAs and proteins in fungus-infected parsley leaves. Plant Physiol 112: 131 (1996). Sablowski RWM, Baulcombe DC, Bevan M: Expression of a flower-specific Myb protein in leaf cells using a viral vector causes ectopic activation of a target promoter. Proc Natl Acad Sci USA 92: 6901–6905 (1995). Sablowski RW, Moyano E, Culianez-Macia FA, Schuch W Martin C, Bevan M: A flower-specific Myb protein activates transcription of phenylpropanoid biosynthetic genes. EMBO J 13: 128–137 (1994). Schulze-Lefert P, Dangl JL, Becker-Andre A, Hahlbrock K, Schulz W: Inducible in vivo footprints define sequences necessary for UV-light activation of the parsley chalcone synthase gene. EMBO J 8: 651–656 (1989). Subramaniam R, Reinold S, Molitor EK, Douglas, CJ: Structure, inheritance, and expression of hybrid poplar (Populus trichocarpa × Populus deltoides) phenylalanine ammonia-lyase genes. Plant Physiol 102: 71–83 (1993). Shufflebottom D, Edwards K, Schuch W, Bevan M: Transcription of two members of a gene family encoding phenylalanine ammonia-lyase leads to remarkably different cell specificities and induction patterns. Plant J 3: 835–845 (1993). Wanner LA, Li G, Ware D, Somssich IE, Davis KR: The phenylalanine ammonia-lyase gene family in Arabidopsis thaliana. Plant Mol Biol 27: 327–338 (1995). Whetten R, Sederoff R: Lignin biosynthesis. Plant Cell 7: 1001–1013 (1995). Wu SC, Hahlbrock K: In situ localization of phenylpropanoidrelated gene expression in different tissues of light-grown and dark-grown parsley seedlings. Z Naturforsch C 47: 591–600 (1992).