Introduction

Metallothioneins (MTs) are described as gene-encoded low-molecular-weight cysteine-rich proteins that bind cations of transition metals (Kägi 1991). They are distributed in all eukaryotes, including animals, fungi and plants, and some prokaryotes. Although the mammalian MTs represent the most intensively studied group, their primary functions remain unclear (Vasak and Hasler 2000). In angiosperms, MTs have been studied in detail; however, there is little information about the regulation and functions of such proteins in conifers. The first gene-encoded plant MT was discovered in wheat by Lane et al. (1987) and its translated product displayed a pattern of cysteine distribution analogous to that of rat MT (Kawashima et al. 1992). Since then, genes and cDNAs encoding proteins with sequence similarity to MTs have been isolated from several plant species but only wheat Ec protein and proteins MT1, MT2, and MT3 from Arabidopsis (Murphy et al. 1997) have been purified, so the term MT-like proteins is preferred in plants.

MT-like proteins can be categorized into three classes based on the pattern of cysteine distribution (Robinson et al. 1993; Rauser 1999). Class-I MTs are those containing two cysteine-rich domains as seen in animal MTs, and class-II MTs carry an additional cysteine-rich domain within the internal region. Class-III MTs are atypical, enzymatically synthesized polypeptides based on repeating units of γ-glutamylcysteine (Cobbett 2000). Class-I MTs are further subdivided into four types (Yu et al. 1998) distinctive in the characteristics of their cysteine-containing motifs.

It is clear that each MT-like gene type exhibits characteristic development- and tissue-specific expression patterns. The expression of class-II MT genes, such as the wheat and maize EcMT, is restricted to immature embryos (Kawashima et al. 1992; White and Rivin 1995; Reynolds and Crawford 1996). Class-I MT-like transcripts have been detected primarily in roots (de Miranda et al. 1990; Evans et al. 1990; Hsieh et al. 1996; Laplaze et al. 2002), senescent leaves (Buchanan-Wollaston and Ainsworth 1997; Foley et al. 1997), stems, leaves and flowers (Snowden and Gardner 1993; Foley and Singh 1994; Coupe et al. 1995; Zhou and Goldsbrough 1995; Choi et al. 1996; Whitelaw et al. 1997; Chang et al. 2004). Transcripts of type-3 MT-like genes have also been detected in fruits, and show differential expression during fruit development (Ledger and Gardner 1994; Lam and Abu Baker 1996; Reid and Ross 1997; Siti Nor Akmar et al. 2002). Some class-I MT genes show programmed expression during embryogenesis. Transcripts of barley pZE40, rice Ose712 (both type 2) and white spruce EMB30 (type 3) genes are temporally expressed during embryo maturation (Smith et al. 1992; Dong and Dunstan 1996; Chen and Chen, 1997).

Direct evidence for cis-acting sequences required for plant MT-like gene expression has been derived from promoter studies in transgenic plants. Fordham-Skelton et al. (1997) investigated the pattern of GUS expression mediated by the progressive 5′-deletion of the PsMTA gene promoter in transgenic Arabidopsis and showed that the −583/−285 region was responsible for β-glucuronidase (GUS) expression in roots. The 300-bp sequence contained three copies of ethylene-responsive elements (EREs; ATTCAA). Interestingly, at least one ERE has been identified in all plant MT-like gene promoters isolated so far, including barley B22E (Klemsdal et al. 1991) and tomato LeMTB (Whitelaw et al. 1997).

The MT transcripts accumulate in tissues undergoing senescence (Buchanan-Wollaston 1994; Hsieh et al. 1995; Thomas and de Villiers 1996) and ripening (Ledger and Gardner 1994; Reid and Ross 1997), known sites of ethylene action. However, the molecular basis of the activation of MT-like gene expression by ethylene in plants remains unanswered. A root-specific element (RSE) has been identified in MT-like gene promoters from pea (Fordham-Skelton et al. 1997), maize (de Framond 1991) and oil palm (Siti Nor Akmar et al. 2002). An imperfect abscisic acid (ABA)-responsive element (ABRE) has been identified in plant MT-like gene promoters (Evans et al. 1990; Klemsdal et al. 1991; Kawashima et al. 1992; Zhou and Goldsbrough 1995; Whitelaw et al. 1997), but is absent in the promoter of the non-functional Arabidopsis MT1b gene (Zhou and Goldsbrough 1995). The presence of the ABRE is consistent with the up-regulation of the Ec gene by ABA in germinating wheat embryos (Kawashima et al. 1992) and developing pollen embryoids (Reynolds and Crawford 1996). Plant MT-like gene expression is also regulated by different stresses, including infection with tobacco mosaic virus in tobacco (Choi et al. 1996) and Phytophthora infestans in potato (Birch et al. 1999).

We reported earlier that Douglas-fir MT (PmMT) cDNA encodes a novel cysteine- and serine-rich MT, indicating a new subtype or prototype MT from which other plant MTs may have evolved. High expression of PmMT transcript in somatic embryos (SE)/zygotic embryos (ZE) and regulation by ABA, osmoticum, and metals is intriguing (Chatthai et al. 1997). To further understand the regulation of MT-like genes in conifers, corresponding genomic sequences have been isolated from a genomic library of Douglas-fir using MT cDNA probes. To study the role of conserved regions, several deletions of the gPmMTa were fused to the GUS gene and analyzed in Douglas-fir SE/ZE. To evaluate the potential of the gPmMT promoter for biotechnological application, the gPmMTa–gusA fusions (full length and truncated) were introduced into tobacco via Agrobacterium tumefaciens-mediated transformation, and GUS activity was examined in transgenic tobacco seeds and plants.

Materials and methods

Screening of a Douglas-fir genomic library

A Douglas-fir (Pseudotsuga menziesii [Mirb] Franco) genomic library was constructed from Sau3A I genomic DNA fragments ranging from 9 to 20 kb, according to the protocol for the lambda EMBL3/BamHI vector kit (Stratagene, La Jolla, CA, USA). Douglas-fir genomic DNA was isolated from spring-flush needles according to De Verno et al. (1989) and purified by ultracentrifugation through a CsCl gradient. A titer for the stored Douglas-fir genomic library was 6×1010 pfu·ml−1. Approximately 1×106 plaques were screened using a 32P-labeled PM2.1 cDNA (Chatthai and Misra 1998) as a probe. The probe was labeled with α-[32P]dCTP using the Random Primer DNA Labeling System (Gibco, Burlington, Ontario, Canada) according to the manufacturer’s instructions.

DNA sequencing

DNA sequences were determined from both strands by the dideoxynucleotide chain termination method using Sequenase version 2.0 as per the manufacturer’s instructions (United States Biochemical, Cleveland, OH) together with commercially available and custom-synthesized primers.

Primer extension analysis

Thirty micrograms of total RNA from megagametophytes at early- and mid-cotyledonary stages was combined with 1×105 cpm of 5′-end-labeled primer 5′PMMT (5′-ACAGCCACAGTCTTTGCCGTCAG-3′) complementary to the 5′-terminus of the PM2.1 coding region. The primer was 5′-end-labeled with γ-[32P]ATP. The primer–RNA mixture was precipitated with ethanol in 0.3 M sodium acetate (pH 5.2) for 30 min at −20°C. The nucleic acid was dissolved in 30 μl of aqueous hybridization buffer [3 M NaCl, 0.5 M Hepes (pH 7.5), 1 mM EDTA], heated for 10 min at 85°C, annealed for 12 h at 55°C and precipitated with ethanol. First, single-strand cDNAs were produced by dissolving the pellet in 20 μl of reverse transcriptase buffer [50 mM Tris–HCl (pH 7.6), 60 mM KCl, 10 mM MgCl2, 1 mM DTT, 1 mM dNTPs, 1 unit·ml−1 RNASIN, 50 units of AMV reverse transcriptase] and incubating for 2 h at 42°C. The reaction was stopped by addition of 25 mM EDTA and 2.5 units RNase H (Gibco). The reaction was incubated for 30 min at 37°C and heated for 5 min at 95°C. The primer extension products were extracted once with phenol:chloroform (1:1, v/v) and precipitated with ethanol in 0.3 M sodium acetate (pH 5.2) for at least 2 h at −20°C. The pellet was dissolved in 10 µl formamide loading buffer (80% formamide, 1 mM EDTA, 0.1% xylene cyanol, 0.1% bromophenol blue), heated for 3 min at 90°C, and electrophoresed on a 6% acrylamide sequencing gel alongside dideoxy chain termination sequencing reactions to determine the size of the products.

Construction of GUS expression vectors

For construction of pMTP0.9–GUS, the 0.9-kb fragment of the 5′-flanking sequence of the gPmMTa gene was PCR-amplified from the plasmid gPMMTa-Exo1.2 using a pair of primers creating PstI/SalI sites at the 5′-end (5′-primer; 5′-CTGCAGTCGACCCCCTAGAGAGTTCTGAAT-3′) and XbaI/BamHI sites at the 3′-end (3′-primer; 5′GGATTCTCTAGAGTTTGCAGATCGTTAACA-3′) of the promoter fragment. After partial digestion with XbaI/PstI, the 0.96-kb PCR product was cloned into pBI221 vector (Clontech) in place of the XbaI/PstI fragment of the CaMV 35S promoter region. For construction of pMTP0.2, the plasmid pMTP0.9 was partially digested with HindIII/XbaI and the isolated 0.28-kb fragment was cloned into pBI221 vector, replacing the HindIII/XbaI fragment of the CaMV 35S promoter region. PCR was utilized to create the promoter deletion constructs, pMTP0.7 using the 5′-mtp1 primer (5′-GATGATAATGTCGACGATAATAG-3′) and 3′-primer, and pMTP0.5 using the 5′-mtp2 primer (5′-CCCTATAAAATGTCGACTTCACG-3′) and 3′-primer. In both cases, the upstream primers (5′-mtp1 and 5′-mtp2) contained sequences that created a SalI site at the 5′-end of the promoter. After digestion with SalI/XbaI, each PCR product was cloned into the promoterless pBI101 vector. The resulting plasmids were digested with HindIII/XbaI and the released promoter fragments were cloned into pBI221 vector in place of the CaMV 35S promoter.

Transient expression

Immature seeds corresponding to early- and mid-cotyledonary stages were collected on July 11 1998 and July 31 1998, respectively. The seeds were surface-sterilized in 1% sodium hypochlorite for 5 min and rinsed 3 times in sterilized water before megagametophytes and zygotic embryos were separated. All samples were placed on BM-3 medium (Gupta and Pullmann 1991) in a 60-mm-diameter Petri dish and used for particle bombardment.

Particle bombardment was carried out with the PDS-1000/He Particle Delivery System (Bio-Rad Laboratories, Richmond, CA, USA) according to the manufacturer’s instructions. DNA was precipitated onto gold particles (1.5–3.0 μm diameter; Aldrich Chemicals) as described in Klein et al. (1988). The following parameters were used for each bombardment: the sample chamber was evacuated to 6.77×104 Pa, the gap distance between the rupture disk and macrocarrier was 0.6 cm, the macrocarrier travel distance was 1.6 cm, and target tissues were placed 6.0 cm from the stopping screen and bombarded once at 8.96×106 or 1.07×107 Pa. Each experiment was repeated three to four times, on different days, and with freshly prepared new batches of DNA-coated gold particles. Bombarded explants remained on the same plates for 24–48 h before they were subjected to histochemical localization of GUS activity (Jefferson et al. 1987).

Tobacco transformation

Transgenic tobacco (Nicotiana tabacum L. cv. Xanthi) plants were generated by leaf-disc transformation as described in Horsch et al. (1985). Regenerated plants were tested for the presence of chimeric gene constructs using PCR-amplification of genomic DNA. The transgenic tobacco plants were then transferred to soil and developing seeds were collected. GUS activity in seeds and tissues was detected using histochemical localization (Jefferson et al. 1987).

Results

Characterization of the Douglas-fir gPmMTa gene

A Douglas-fir genomic library constructed in the λEMBL3 bacteriophage was screened with the PM2.1 cDNA probe. From approximately 1×106 genomic clones, three clones strongly hybridized to the PM2.1 cDNA. Restriction analysis of these clones revealed different restriction patterns (data not shown), suggesting different MT-like genes in Douglas-fir. A 4.5-kb XbaI fragment of λPMMTa genomic clone was subcloned into plasmid pUC19, giving rise to the gPmMTa genomic clone.

The nucleotide sequence of 2.3 kb of gPmMTa and the deduced primary structure of a Douglas-fir MT-like polypeptide are shown in Fig. 1. Comparison of the nucleotide sequences of the gPmMTa gene and PM2.1 cDNA indicated that the gPmMTa transcription unit consisted of three exons interrupted by two introns that located at amino acid positions 17 and 39. The introns are 125 and 78 bp in length. Both regions exhibit characteristics of the putative plant introns such as the A/T abundance, the :GU...AG: boundary and the consensus YUNAN (where Y is C or U, and N is any nucleotide) branchpoint sequence (Brown et al. 1996). The coding region of the gPmMTa gene was not identical to that of the PM2.1 cDNA; however, they shared high degrees of similarity at both nucleotide (94%) and amino acid (98.5%) levels.

Fig. 1
figure 1

Structure of the Douglas-fir (Pseudotsuga menziesii) gPmMTa gene. The nucleotide and deduced amino acid sequences of the gPmMTa gene. The transcription start site (cytosine) is referred to as +1. Numbers shown on the left indicate the nucleotide position with respect to the transcriptional initiation site. Capitalized nucleotides indicate translated sequences. Introns, 5′- and 3′-untranslated sequences are shown in italics. The putative TATA box in the promoter region is underlined. Highlighted in bold type are G-box like motifs (ACGT), an E-box (E-box), an A/T-rich sequence (A/T rich), a motif similar to the metal regulatory element (MRE) of animal MT genes and three repeats of a sequence similar to the ethylene-responsive element (ERE). Arrow lines above the sequence indicate two almost identical sequences. The predicted amino acid sequence (single-letter abbreviations) is shown below the nucleotide sequence within the open reading frame

Full size image

A primer extension assay using total RNA from the stage-5 megagametophyte and the 23-mer oligonucleotide 5′PMMT, complimentary to the 5′-terminus of the PM2.1 coding sequence, was performed to locate the transcription start site (Fig. 2). The longest transcript, as deduced from the sequence ladder produced with the same primer, was initiated from the cytosine, 92 nucleotides upstream from the ATG initiation codon. The transcription start site at the cytosine was assigned as the start site and is designated hereafter as +1.

Fig. 2
figure 2

Transcriptional initiation site of the gPmMTa gene. A 19-mer oligonucleotide complementary to the sequence from +118 to +96 of the gPmMTa gene was labeled at the 5′-end and used for primer extension analysis using total RNA prepared from megagametophytes at 31 days after fertilization. Products of the primer extension reaction (lane P) were electrophoresed on an 8% polyacrylamide gel containing 7 M urea. The gPmMTa sequencing reactions (lanes G, A, T, C) were co-electrophoresed using the same primer. The corresponding nucleotide position of the primer extension product is shown by an asterisk

Full size image

When the gPmMTa promoter sequence was compared to the analogous regions of other MT-like genes, no significant similarity between gPmMTa and any other sequences could be detected. However, a number of cis-regulatory motifs common to these genes were evident in the Douglas-fir gPmMTa promoter upstream of the putative TATA box (Fig. 1). These include an E-box (CANNTG; Stålberg et al. 1996) at position −493, three copies of palindromic G-box-like motifs (AACGTT, CACGTG; Foster et al. 1994) between positions −104 and −165, and three copies of putative EREs (AWTTCAAA; Montgomery et al. 1993) around positions −580, −590 and −745. The gPmMTa gene also contains a motif similar to the core metal-responsive element (MRE) (TGCRCNC; Thiele 1992). The MRE motif has only been found in pea PsMTA (Fordham-Skelton et al. 1997) and oil palm MT3-B (Siti Nor Amar et al. 2002). The unique and interesting feature found in the proximal 5′-upstream sequence of the Douglas-fir gPmMTa gene is the presence of two 150-bp direct repeat units (−134/+20 and −34/−175 regions). Both sequences contain the putative TATA-box and identical sequences flanking the predicted transcriptional start site. It is unknown whether transcription is initiated from the distal TATA-box.

Analysis of the gPmMTa-uidA chimeric gene activity in Douglas-fir

Functional analyses of the Douglas-fir MT-like gene promoter were performed using intact and deleted gPmMTa promoter sequences fused to the β-glucuronidase (uidA or GUS) coding region.

Using the particle-bombardment method, a 0.9-kb gPmMTa promoter/uidA chimeric gene construct (pMTP0.9), the promoter-less reporter gene (pBI101), and the CaMV 35S promoter/uidA gene (pBI221) were delivered into developing Douglas-fir megagametophytes, immature zygotic embryos and mature somatic embryos. The GUS activity was determined as GUS-expression units (GEUs) shown as blue spots after the histochemical GUS assay 2 days after bombardment (Fig. 3). Bombardment with pBI101 produced no visible GUS transient expression. By contrast, GUS activity was evident when Douglas-fir tissues were bombarded with pBI221 and pMTP0.9. The efficiency of the 0.9-kb gPmMTa promoter in comparison to that of the constitutive CaMV 35S promoter was estimated by counting the number of GEUs per tissue. As shown in Fig. 5, bombardment with the pMTP0.9 construct produced 2- to 3-fold higher GEUs in both megagametophytes and zygotic embryos than the pBI221 plasmid. The results indicate that the gPmMTa promoter is capable of driving expression of GUS in megagametophytes, zygotic embryos and somatic embryos of Douglas-fir.

Fig. 3
figure 3

Transient-expression analysis of the gPmMTa promoter. Douglas-fir megagametophytes (STAGE 5 MG and STAGE 6 MG), zygotic embryos (STAGE 6 ZE), and mature somatic embryos (MATURE SE) were bombarded with the plasmids pBI101, pBI221, or the gPmMTa/uidA construct (pMTP0.9). Blue spots represent cells expressing GUS activity. Bars = 1 mm

Full size image

To localize regulatory regions responsible for expression of MT-like genes in Douglas-fir, a series of deletion mutants of the gPmMTa gene promoter were constructed. Four 5′-deletion gPmMTa promoter fragments were linked to the uidA gene with nos terminator in the vector pBI221, and designated pMTP0.9, pMTP0.7, pMTP0.5 and pMTP0.2 (Fig. 4). All constructs included the 5′-UTR of the transcript and the initiation codon of gPmMTa. Each construct was assayed by transient expression after particle bombardment into stage-5 and stage-6 megagametophytes, stage-6 zygotic embryos and mature somatic embryos of Douglas-fir (Figs. 4, 5). pMTP0.9 (−856/+88) generated approximately 80 and 33 GEUs per tissue in stage-5 and stage-6 megagametophytes, respectively. There was no significant difference in GUS activity when the tissues were bombarded with pMTP0.7 (−677/+88). Deletions to position −453 (pMTP0.5) and position −190 (pMTP0.2) consistently produced only 50% of the GUS expression generated by pMTP0.9 and pMTP0.7. Results revealed that the −677/−453 region of the gPmMTa gene contains a positive regulatory element(s) for the expression in megagametophytes.

Fig. 4
figure 4

Transient expression of the gPmMTa/uidA chimeric gene in Douglas-fir. Douglas-fir megagametophytes (STAGE 5 MG and STAGE 6 MG) and zygotic embryos (STAGE 6 ZE) were bombarded with the construct pMTP0.9, pMTP0.7, pMTP0.5 or pMTP0.2. Blue spots represent cells expressing GUS activity. Top panel Four deletion constructs showing a uidA reporter gene linked to each of four 5′-deletion regions of the gPmMTa promoter. Colored boxes indicate the positions of ethylene-responsive elements (ERE; yellow), metal-responsive elements (MRE; red) and G-box motifs (ACGT; blue). Bars = 1 mm

Full size image
Fig. 5
figure 5

Transient-expression analysis of deletion constructs of the gPmMTa promoter. Douglas-fir megagametophytes (stage 5 MG and stage 6 MG) and zygotic embryos (stage 6 ZE) were bombarded with the construct pBI221, pMTP0.9, pMTP0.7, pMTP0.5 or pMTP0.2. Ten megagametophytes or embryos were used in each of three bombardments. For each replicate, the number of GUS-transient expression units (GEUs) were scored and averaged out as GEUs per tissue. The average numbers of GEUs per tissue were calculated from three independent bombardments. Means + SE

Full size image

In stage-6 zygotic embryos, pMTP0.7 produced the highest GUS activity (19 GEUs per embryo) which accounted for 2-fold higher activity than the construct pMTP0.9. The pMTP0.5 construct generated only 3 GEUs per embryo; however, further deletion to −190 caused an increase in GUS activity by 3-fold as compared to the construct pMTP0.5. This is in contrast to what was found in megagametophytes where the efficiencies of pMTP0.5 and pMTP0.2 were similar. These observations indicate that the −677/−453 region of the gPmMTa gene promoter is essential for the high level of gene expression in zygotic embryos whereas the −853/−677 and −453/−190 regions may contain negative regulatory elements. GUS activity was evident in somatic embryos of all constructs studied; however, with the exception of pMTP0.5, GEUs generated were too densely arranged to count.

Analysis of the gPmMTa-uidA chimeric gene activity in transgenic tobacco

In tobacco plants, four 5′-deletion fragments of the gPmMTa gene promoter fused to the GUS-coding region (replacing the CaMV 35S promoter in pBI121) were introduced via Agrobacterium-mediated transformation. The chimeric gene integration and the number of uidA gene copies in the transformed tobacco were verified using PCR-amplification and Southern blot analysis (data not shown). Figure 6 shows histochemical staining for the GUS activity in developing seeds of the transgenic tobacco containing the 0.9-kb gPmMTa gene promoter–uidA gene fusion (full-length promoter). Very low GUS activity was detected in whole seeds, embryos and endosperm of transgenic tobacco. Similar results were observed in transgenic tobacco carrying other chimeric gene constructs.

Fig. 6
figure 6

Histochemical localization of GUS activity in transgenic tobacco (Nicotiana tabacum cv. Xanthi) seeds. Seeds of transgenic tobacco transformed with the uidA chimeric gene construct under the control of the gPmMTa promoter were assayed for GUS activity. To assay GUS activity, seeds (30 days after flowering) of transgenic tobacco were incubated in reaction buffer containing 5-bromo-4-chloro-3-indolyl-β-d-glucuronide (X-Gluc) overnight, dissected to separate embryos and endosperm and photographed. Bars = 1 mm (two left panels), 0.25 mm (right panel)

Full size image

However, when 2-week-old T1 transgenic tobacco seedlings were used for histochemical staining, the GUS activity was detected in the cotyledons of plants containing pMTP121–0.9 (Fig. 7c), pMTP121–0.7 (Fig. 7d), and pMTP121–0.5 (Fig. 7e) constructs, while no GUS activity was detected in plants containing the construct pMTP121–0.2 (Fig. 7f). The GUS expression appears to be confined to the meristematic and mitotic tissues of the vascular system, auxiliary buds and stem elongation system. Plantlets germinated from wild-type seeds were used as a negative control (Fig. 7a); plants expressing GUS under the control of CaMV 35S promoter were used as a positive control (Fig. 7b). Deletion of 179 bp from the 5′-end of the gPmMTa promoter fragment caused an increase in the level of GUS activity in transgenic tobacco, whereas further deletions caused significant reductions. In older plantlets, however, the low GUS activity was restricted to the auxiliary buds only (data not shown). Also, it is important to point out that none of the MT-like promoters tested (full length or truncated) showed any GUS activity in the roots of plantlets.

Fig. 7a–f
figure 7

Histochemical localization of GUS activity in transgenic tobacco seedlings. Tobacco Xanthi was transformed with plasmids pBI121 (b), pMTP121-0.9 (c), pMTP121-0.7 (d), pMTP121-0.5 (e), or pMTP121-0.2 (f). Two-week-old seedlings were assayed to localize GUS activity. Non-transformed tobacco Xanthi was used as a negative control (a)

Full size image

Discussion

Very little is known about the control of conifer gene expression during germination and early seedling development. The identification of genes expressed during plant embryogenesis is a fundamental step towards the understanding of the molecular biology of seed development. Up to date, relatively few promoters from forest trees have been isolated and characterized. Gene-transfer technology has enabled a number of plant genes to be introduced into heterologous hosts and has facilitated the analysis of qualitative and quantitative aspects of their expression. The demonstration that genes retain regulated patterns of expression in the foreign host has permitted the delineation of cis-acting sequence elements involved in tissue-specific and developmental control.

In an attempt to identify promoter regions responsible for the regulated expression of the gPmMTa gene, chimeric genes containing a series of promoter deletions fused to the uidA reporter gene were constructed and introduced into cells of immature seeds and somatic embryos of Douglas-fir by particle bombardment. The results showed that a sequence of 190 bp upstream of the gPmMTa transcription start site is sufficient to direct GUS expression in the megagametophyte, and in zygotic and somatic embryos of Douglas-fir. The proximal 190-bp sequence, in addition to a putative TATA-box sequence, contains an inverted repeat of the ACGT-core motif. The core sequence is part of the G-box (CACGTG) which appears to be conserved in the promoters of active MT-like genes identified so far, including Arabidopsis MT1a (Zhou and Goldsbrough 1995), pea PsMTa (Fordham-Skelton et al. 1997), tomato LeMTB (Whitelaw et al. 1997), rice rgMT (Hsieh et al. 1995) and cgMT1 from the tropical tree Causarina glauca (Laplaze et al. 2002). The G-box and related sequences are required for the induction of the differential expression of genes by stress, light, abscisic acid (Busk and Pagè 1998; Guilfoyle 1997; Shen and Ho 1997) and ethylene (Sessa et al. 1995).

The second functional domain identified in the gPmMTa promoter extends from positions −677 to −453. This region conferred a high level of expression of the uidA gene in both megagametophyte and embryos. Sequence analysis of this region revealed the presence of several putative regulatory elements, including two copies of the ERE, an A/T-rich region and E-box motif. In a number of plants, ethylene is implicated in regulating tissue-specific and developmentally regulated expression of MT-like genes. For example, MT-like cDNAs showed induction during leaf senescence in Brassica napus (Buchanan-Wollaston 1994) and Arabidopsis (Zhou and Goldsbrough 1995), during ethylene-stimulated leaf abscission in Sambucus nigra (Coupe et al. 1995), and during fruit-ripening in kiwi (Ledger and Gardner 1994), apple (Reid and Ross 1997), papaya (Lam and Abu Baker 1996), and cherry (Wiersma et al. 1998). The transient-expression assays showed that the deletion constructs lacking two copies of the ERE exhibited a significant decrease in GUS activity in developing seeds. Similar behavior was observed in the case of the PsMTA promoter, when the deletion of three copies of the ERE caused deficiency in expression of a reporter gene within roots and senescent aerial tissues of transgenic Arabidopsis (Fordham-Skelton et al. 1997). The second sequence, CATTTG, at position −493, resembles the putative E-box (CANNTG) that was shown to be a recognition site for DNA-binding proteins in the promoter of the bean β-phaseolin gene (Kawagoe and Murai 1992), and was responsible for quantitative and correct seed-specific expression of napin genes (Ellerström et al. 1996; Stålberg et al. 1996). Finally, the identified A/T-rich 39-bp sequence extending from −637 to −604 is believed to act as a general enhancer of expression (Stålberg et al. 1996).

The 5′-flanking region of gPmMTa contains a putative metal regulatory element (MRE), suggesting the possibility of metal-regulated transcription. This is in agreement with the results from northern analyses showing metal-induced accumulation of PM2.1 transcripts in seeds and young seedlings of Douglas-fir (Chatthai et al. 1997). To date, there have been only two reports of the existence of MREs in plant MT-like genes. The pea PsMTA (Fordham-Skelton et al. 1997) and tomato LeMTB (Whitelaw et al. 1997) each contain one putative MRE in the 5′-flanking region of the genes; however, there is no evidence that the sequence is functional. The activity of the gPmMTa promoter is similar to the activity of other plant promoters unrelated to MTs but related to development and embryogenesis, i.e. the Hrgp (hydroxyproline-rich glycoprotein) promoter from maize (Menossi et al. 2003).

Studies on plant MT-like gene promoters using stable transformation have been limited. The GUS expression analysis in transgenic Arabidopsis showed that the promoter of the Brassica napus LSC54 gene was highly induced during leaf senescence and in response to wounding and pathogen infection (Butt et al. 1998). The barley B22EL8 gene promoter sequence directed the expression of a reporter gene in barley embryos but was not functional in transgenic tobacco (Klemsdal et al. 1991). When the gPmMTa promoter was used to drive gus gene expression in a heterologous system, the GUS expression pattern was very similar to that observed with MT-like gene promoters from rice (Yu et al. 2000), tomato (Whitelaw et al. 1997), and sunflower (Chang et al. 2002), all of which showed the highest expression in stems and leaves, but not roots. When 2-week-old T1 transgenic tobacco seedlings were used for the histochemical analysis, the GUS activity was detected in the cotyledon of transgenic tobacco plants. The deletion of 179 bp from the 5′-end of the gPmMTa promoter fragment caused an increase in the level of GUS activity in transgenic tobacco, whereas further deletions caused significant reduction. In summary, these results show that the Douglas-fir MT-like gene promoter functions in tobacco and that some components of the regulatory pathway are conserved between these two evolutionarily distinct plant species.

The activity of the gPmMTa promoter appears to be developmentally regulated and showed a decrease in activity with the age of transgenic tobacco seedlings. This finding may indicate that the role of MT-like proteins in developing young seedlings is to give them an advantage by sequestering essential ions while protecting them from toxic metal ions.

The roles and functions of MT-like proteins in plants are not clearly understood. They (or at least some of them) are involved in heavy-metal metabolism and tolerance. Their multiple cysteine residues could bind heavy metals and then release them when the cysteine residues are oxidized (Fabisiak et al. 1999). However the diversity of the sites of their expression and different stimuli used to activate them suggest that the plant MT-like proteins could be involved in many biological processes, including development, apoptosis and the defense against plant pathogens. We are currently investigating the role of MT-like protein in Douglas-fir embryogenesis, and its regulation in response to various signal molecules that may interact with the motifs identified in the gPmMTa promoter.

The characteristics of new promoters may also prove useful in the engineering of novel traits into conifer and angiosperm plant species, as well as expression/accumulation of foreign proteins in heterologous plant systems.