Characterization of a Novel Rice Metallothionein Gene Promoter: Its Tissue Specificity and Heavy Metal Responsiveness

Abstract

The rice (Oryza sativa L.) metallothionein gene OsMT-I-4b has previously been identified as a type I MT gene. To elucidate the regulatory mechanism involved in its tissue specificity and abiotic induction, we isolated a 1 730 bp fragment of the OsMT-I-4b promoter region. Histochemical β-glucuronidase (GUS) staining indicated a precise spacial and temporal expression pattern in transgenic Arabidopsis. Higher GUS activity was detected in the roots and the buds of flower stigmas, and relatively lower GUS staining in the shoots was restricted to the trichomes and hydathodes of leaves. No activity was observed in the stems and seeds. Additionally, in the root of transgenic plants, the promoter activity was highly upregulated by various environmental signals, such as abscisic acid, drought, dark, and heavy metals including Cu²⁺, Zn²⁺, Pb²⁺ and Al³⁺. Slight induction was observed in transgenic seedlings under salinity stress, or when treated with Co²⁺ and Cd²⁺. Promoter analysis of 5′-deletions revealed that the region −583/−1 was sufficient to drive strong GUS expression in the roots but not in the shoots. Furthermore, deletion analysis indicated important promoter regions containing different metal-responsive cis-elements that were responsible for responding to different heavy metals. Collectively, these findings provided important insight into the transcriptional regulation mechanisms of the OsMT-I-4b promoter, and the results also gave us some implications for the potential application of this promoter in plant genetic engineering.

Introduction

Heavy metals such as Cu and Zn are essential for normal plant growth and development, functioning as the cofactors or structural components of many enzymes and other proteins (Hall 2002). However, elevated concentrations of both essential and non-essential heavy metals are toxic and can result in growth inhibition (Clemens 2001). Plants have evolved a range of potential cellular mechanisms that may be involved in the detoxification of heavy metals and thus tolerance to metal stress. Chelation of metals by high-affinity ligands is potentially one very important mechanism. Such ligands include amino acids, organic acids, and two classes of cysteine-rich peptides, phytochelatins (PCs) and metallothioneins (MTs). MTs bind metal ions in the metal-thiolate cluster, and therefore contribute to metal detoxification by buffering cytosolic metal concentration (Cobbett and Goldsborough 2002). Induction of the synthesis of PCs is based on the posttranscriptional activation of PC synthase but not under transcriptional control. In contrast, MTs are encoded by a family of genes (Cobbett and Goldsborough 2002). Therefore, it is important to characterize the promoter region of the MT gene to reveal its expression pattern, including the tissue specificity and heavy metal responsiveness.

At present, more than 100 plant MT gene sequences can be obtained from public databases. An increasing number of reports indicate that plant MTs are organ-specific, and that they may play important roles in developmental processes, such as fruit ripeness, root development, suberization, pollination, and seed development (Chatthai et al. 1997; Clendennen and May 1997; Mir et al. 2004; Moyle et al. 2005; Zhou et al. 2005; Yuan et al. 2008). MT isoforms with different expression patterns may be transcriptionally regulated in different manners and have specialized functions in the corresponding tissues. To study the transcriptional regulation of each MT gene further, the 5′ promoter regions have also been isolated and studied to reveal their tissue-specificity (Chatthai et al. 1997; Fordham-Skelton et al. 1997; Hsieh and Huang 1998; Guo et al. 2003; Fukuzawa et al. 2004). For example, in the pea PsMTA promoter, the region −583/−285 is responsible for the expression of the downstream GUS reporter in the root of transgenic Arabidopsis (Fordham-Skelton et al. 1997). Another example is from a rice MT gene (ricMT) promoter, which contains three regions that are primarily responsible for its tissue-specific expression. These include the 5′-distal region (−1382/−910), which is important for the aerial parts of transgenic seedlings, a 5′-proximal region (−194/−1) for the roots and an internal region (−909/−708) for the floral organs in transgenic Arabidopsis plants (Lü et al. 2007). Taken together, these data indicate that the regulatory mechanism for MT genes in plant development is complicated.

To understand the metal responsiveness of plant MT promoters, more detailed research is required. In particular, the cis-elements conferring heavy metal responsiveness remain to be identified. One cis-element has been identified in the green algae Chlamydomonas reinhardtii: a copper response element (CuRE) with the conserved core sequence 5′-GTAC-3′ (Quinn and Merchant 1995; Quinn et al. 2000). A copper response regulator, CRR1, binds to CuRE sites and mediates the expression of downstream genes under copper-deficient conditions (Quinn et al. 2000). In addition, another report showed that both CuRE and CRR1 are required for nickel response (Quinn et al. 2003).

In contrast, heavy metal-responsive elements and the corresponding transcription factors have been identified in yeast and animals in detail. In yeast, it has been reported that copper MT genes are induced by a metal-responsive element (MRE, with the consensus core sequence 5′-HTHNNGCTGD-3′) and its corresponding metal-responsive transcription factor, ACE1. ACE1 can bind to MRE sequences and regulate the downstream MT genes (Furst et al. 1988; Dixon et al. 1996). In animals, metal regulation of MT genes is mediated by metal-responsive elements (MREs), which contain a highly conserved core sequence (5′-TGCRCNC-3′) and can be recognized by a MRE-binding transcription factor-1 (MTF-1) (Stuart et al. 1985; Heuchel et al. 1994). MREs are always present in multiple copies upstream of MT genes (Stuart et al. 1985). In plants, Lü et al. reported that a special region (−331/−194) of a rice MT promoter, which contained an element similar to animal MRE, was indeed required for copper-activated expression of the reporter GUS gene in transgenic Arabidopsis seedlings (Lü et al. 2007). Similar cases have also been reported to show the presence of animal MRE-like motifs in promoters of plant MT genes, including PsMTA (Fordham-Skelton et al. 1997), LeMTB (Whitelaw et al. 1997), PmMT (Chatthai et al. 2004), and PvSR2, a novel heavy metal-specific responsive gene isolated from bean (Qi et al. 2007). However, the exact role of these MRE-like motifs in the metal-responsiveness of plant MT genes and whether there are other metal-responsive elements need to be identified further.

We previously isolated a novel MT gene, OsMT-I-4b, from rice (Oryza sativa L.) that was expressed at a high level only in the root and not in the leaves and sheaths (Zhou et al. 2006). Among the type I MT genes in rice, OsMT-I-4b was the gene whose expression was most highly induced by Cu²⁺ in the roots of rice seedlings. Expression of OsMT-I-4b in lead-sensitive yeast mutant cells resulted in enhanced lead tolerance (Xu et al. 2007). All of the above promoted us to investigate the tissue-specific and metal-regulatory expression pattern of OsMT-I-4b in detail. In the present study, we characterized the promoter region of OsMT-I-4b, aided by GUS reporter in transgenic Arabidopsis plants. Deletions from the 5′-end revealed that different combinations of CuRE and MRE-like motifs were responsible for the specific responsiveness to different metal ions.

Results

Cis-acting element analysis of the OsMT-I-4b promoter region

cDNA of the rice type I MT gene, OsMT-I-4b (accession no. NM_001073598) was isolated previously in our laboratory (Zhou et al. 2006). To investigate the transcriptional activity of the OsMT-I-4b promoter, we cloned a 1 730-bp DNA fragment and designated it as POsMT-I-4b. The A of the translation initiation codon (ATG) of the OsMT-I-4b gene was defined as +1 (Figure 1). A motif search was carried out using the PLACE database (http://www.dna.affrc.go.jp/htdocs/PLACE/) to reveal the putative cis-elements. As shown in Figure 1, a number of regulatory motifs that potentially relate to environmental signals were found in the OsMT-I-4b promoter upstream of two putative basic cis-elements, the TATA (−109/−103) and CAAT (−214/−211) boxes. Four MRE-like sequences, which were similar to the conserved core MRE sequence involved in the heavy metal-induced expression of MT genes in animals (Stuart et al. 1985), were located at positions −1542/−1536, −1370/−1364, −371/−365 and −275/−269. Another heavy metal response element, CuRE (Quinn and Merchant 1995; Quinn et al. 2003), was also found in six copies (−1498/−1495, −1041/−1038, −1014/−1011, −670/−667, −600/−597 and −119/−116). These putative MREs and CuREs suggested that the promoter region of OsMT-I-4b might respond to heavy metal stress via a complex mechanism.

Figure 1. The sequence of the putative promoter region of the OsMT-I-4b gene. The nucleotide A in the translation initiation codon (ATG) was designated as +1. The putative CAAT and TATA boxes are shaded and underlined. The cis-elements that are potentially involved in heavy metal responsiveness are shown red, shaded and boxed. The remaining deduced cis-acting elements are boxed and annotated beneath the figure.

Also, in the OsMT-I-4b promoter there were some conserved regulatory motifs that are present in most plant gene promoters (Figure 1). These included ABRE (the abscisic acid response element, −996/−989, Simpson et al. 2003), two copies of MYC (−1082/−1077, −415/−410) and three copies of MYB (a recognition site in the dehydration responsive gene rd22, −1467/−1462, −1297/−1292 and −856/−851), all of which are required in abscisic acid (ABA) signaling and in response to dehydration stress (Abe et al. 2003; Lee et al. 2005). Two copies of the I-box (−1008/−1003 and −311/−306, which are thought to be involved in the light response) were also found (Lam and Chua 1989). Finally, two copies of the W-box (a binding site for the WRKY transcription factor, at −465/−461 and −336/−332, which plays a role in wound response) were present in the promoter region (Ulker and Somssich 2004). The presence of these stress-related cis-elements suggested that the promoter region of OsMT-I-4b might respond to a variety of environmental stresses, such as stresses from dehydration, light, wounding, and heavy metals.

Tissue-specific expression pattern of OsMT-I-4b promoter in Arabidopsis

To define the tissue-specific expression pattern of OsMT-I-4b promoter precisely, the promoter::GUS chimeric construct (POsMT-I-4b::GUS) was introduced into Arabidopsis, and histochemical GUS staining was carried out in various organs throughout plant development. As shown in Figure 2A, we detected GUS activity in 1-d-old transgenic seedlings. The GUS staining was stronger in the cotyledons and radicles but relatively weaker in the hypocotyls. However, in the 3-d-old and 5-d-old seedlings, GUS activity was present exclusively in the radicles and hypocotyls, but not in the cotyledons (Figure 2B and C). In the 10-d-old seedlings, strong GUS expression was maintained in the roots, and the expression was detectable in the first true leaves (Figure 2D). These results were in good agreement with the previous report, indicating that OsMT-I-4b was expressed at a high level only in the roots, not in the leaves and sheaths (Zhou et al. 2006). Interestingly, no GUS activity was detected in the root tips in 1–10-d-old seedlings (Figure 2A–D).

Figure 2. The histochemical localization of β-glucuronidase (GUS) aactivity in representative transgenic Arabidopsis plants harboring PMT-I-4b::GUS. (A–D) Seedlings grown on Murashige and Skoog (MS) plates at day 1 (A), day 3 (B), day 5 (C) and day 10 (D). (E–H) Different leaf types of the transgenic plants (E, a rosette leaf at day 20; F, a rosette leaf at day 35; G, a cauline leaf at day 35; H–I, a senescent leaf). (J) The stem. (K–N) Reproductive organs of the transgenic plants (K, mature flower; L, young flower; M, young silique and N, old silique).

In 20-d-old and 35-d-old transgenic plants grown in soil, strong GUS activities could still be detected in the roots (data not shown). However, in the rosette leaves of the soil-grown plants, GUS staining could be detected only in the trichomes and hydathodes (Figure 2E,F). In the cauline leaves of 35-d-old transgenic plants, an increased number of trichomes could be stained, as were the hydathodes (Figure 2G). GUS staining was also detectable at cutting sites in cauline leaves (Figure 2G), but not at cutting sites in rosette leaves (Figure 2F). In the senescent leaves, no GUS activity was detected, except at the hydathodes (Figure 2H–I). However, no GUS activity was detected in the stems of the soil-grown transgenic plants (Figure 2J).

To examine GUS expression in reproductive organs of transgenic Arabidopsis, we also stained whole flowers (Figure 2K–N). As a result, intense GUS activity appeared in the bud stigmas. Relatively weaker GUS staining occurred in the filaments and sepals (Figure 2K–L). There was also slight GUS staining in the vascular tissues of the petals (Figure 2K). In young siliques, GUS expression was restricted to the upper zone, and there was a dramatic decrease in intensity as the siliques matured (Figure 2M). When the siliques matured and split, the GUS staining disappeared (Figure 2N). Moreover, the seeds were not stained (Figure 2N). All of the above results indicated that the OsMT-I-4b promoter modulates precise developmental regulation in transgenic Arabidopsis, in both the vegetative tissues and reproductive organs.

Expression of the OsMT-I-4b promoter in response to environmental stresses

The presence of many environmental signal-responsive elements in the OsMT-I-4b promoter (Figure 1) gave us some hints that the activity of the promoter may respond to various abiotic and heavy metal stresses. To test this hypothesis, we carried out quantitative GUS activity assays on 10-d-old-seedlings. Results from treating the transgenic seedlings with abiotic stresses are shown in Figure 3A. GUS activity was significantly induced in the seedling roots at 6 h after treatment with 100 μM ABA and drought, and showed marked increases of 7.65-fold and 5.38-fold over the control, respectively. Furthermore, a 3.57-fold increase was detected in GUS activity after 6 h of dark treatment. Additionally, GUS activity was slightly induced by treatment with high salinity (200 mM NaCl), at 364.51 pmol MU/min/mg protein (approximately 2.43-fold greater than that in the mock control samples). These data clearly indicate that the OsMT-I-4b promoter can respond to various abiotic stresses at different levels of intensity.

Figure 3. β-glucuronidase (GUS) activity assay of the OsMT-I-4b promoter::GUS construct in response to various stresses. (A) 10-d-old transgenic seedlings were treated with abscisic acid (ABA) (100 μM) and other abiotic stresses for 6 h. Then, the roots were harvested and the GUS activities were measured. (B) The response of the OsMT-I-4b promoter to metal stress. The seedlings were treated with 100 μM each of CuSO₄, ZnSO₄ or CoCl₂, 50 μM each of CdCl₂ or AlCl₃, or 20 μM of PbSO₄. (C) The dose response of GUS activity in the seedling roots that were treated with different concentrations of metals. The GUS activities are recorded as the fold-induction caused by the stress in comparison with untreated control seedlings. The values shown are the means ±SE (n= 3). *P < 0.05, **P < 0.01 compared with the corresponding value obtained for the respective mock controls. NT, untreated control (the value was set to 1); Mock, transgenic seedlings treated with distilled water only.

To assay the OsMT-I-4b promoter in response to various heavy metals, we also measured GUS activities in transgenic seedlings treated with 100 μM CuSO₄, 100 μM ZnSO₄, 100 μM CoCl₂, 50 μM CdCl₂, 20 μM PbSO₄, and 50 μM AlCl₃ for 24 h. As shown in Figure 3B, the highest inducible GUS activity in the roots was observed in the seedlings that were treated with Pb²⁺ (10.38-fold induction). The next highest activities were found after Al³⁺ treatment (8.24-fold induction) and Zn²⁺ treatment (8.06-fold induction). Cu²⁺ was another potential inducer of GUS activity (about 6.64-fold induction). However, the transgenic seedlings that were treated with 50 μM Cd²⁺ showed very slight induction in GUS activity (2.07-fold), and no significant induction (P > 0.05) was detected in the seedlings treated with 100 μM Co²⁺, compared with the mock control (Figure 3B).

We also studied the effects of increasing heavy metal concentrations (Cu²⁺, Zn²⁺, Pb²⁺ and Al³⁺) on POsMT-I-4b-driven GUS activity. Cu²⁺ was readily taken up by plants, and the GUS activity increased as the Cu²⁺ concentration increased from 0 (mock control, untreated with Cu²⁺) to 100 μM. Thereafter, GUS activity declined (Figure 3C). This decrease presumably reflected the toxic effects of Cu²⁺ at high concentrations. Similar GUS activity patterns were detected in seedlings treated with Zn²⁺ and Al³⁺ (Figure 3C). However, when the seedlings were treated with Pb²⁺, GUS activity increased as the concentration of metal ions increased from 0 to 50 μM (Figure 3C). Taken together, it can be concluded that the OsMT-I-4b promoter (−1730/−1) responds to various heavy metals, and its activity varies according to the type and concentration of the metal ions. GUS activity was also detected in the shoots of stress-treated seedlings, but the results were not significantly different between the stress-treated and control samples (data not shown).

Deletion analysis of the OsMT-I-4b promoter

In an attempt to localize the regions responsible for the heavy metal responsiveness of the OsMT-I-4b promoter and to reveal its complex regulatory mechanisms, we constructed a series of 5′ deletions (R2 to R5) fused to the GUS reporter gene (Figure 4). Another 5′ deleted version (R5m) was also constructed in which the CuRE proximal to the coding region (−119/−116) was mutated (from GTAC to TTCC, Figure 4). Each construct was introduced into Arabidopsis and the GUS activity in the shoots (gray bars) and roots (black bars) of 10-d-old seedlings was tested. The negative control, which contained a promoterless GUS gene, had undetectable GUS activity in this assay (Figure 4).

Figure 4. Deletion analysis of the OsMT-I-4b promoter. Schematic diagrams of the constructs that were used for the β-glucuronidase (GUS) activity assays are shown on the left. The black boxes indicate the putative TATA boxes (−109/−103). The gray boxes indicate the animal metal-regulatory element (MRE)-like motifs. The white ellipses indicate the putative copper response elements (CuREs), and the crosshatched ellipse indicates the mutated CuRE sequence. The numbers on the far left indicate the 5′-end points of each deletion relative to the translation start site. In each experiment, Arabidopsis plants were transformed with each of the constructs shown, and the GUS activities of the shoots and roots were detected. A plasmid containing a promoter-less GUS (pCAMBIA1381) was used as a negative control. The values recorded are the means ±SE (n= 3). n.d., not detectable.

In the shoots of the transgenic seedlings, the highest level of GUS activity (24.68 pmol MU/min/mg protein) was detected using the R1 construct, which contained the full-length OsMT-I-4b promoter (−1730/−1, Figure 4). However, GUS activity decreased in order from the R2 to R5 construct. Mutation of the CuRE sequence (R5m) decreased the expression of GUS by approximately 26.4% compared with its wild-type version (Figure 4). In the seedlings, GUS activity was significantly higher in the roots than in the shoots (Figure 4). The construct containing the deletion to −583 (R4) showed the highest GUS activity (225.96 pmol MU/min per mg protein), and the GUS activity of the R1 construct was slightly lower than that of R4. A deletion up to −142 (R5) resulted in almost complete loss of GUS activity; just 23.47 pmol MU/min per mg protein (approximately 10.4% of the value obtained for the R4 construct) could be detected. In the mutated version of the R5 construct (R5m), GUS activity was reduced by just 8.6%, with no significant difference from the wild-type version (R5). These data revealed that the DNA fragment −583/−1 might be sufficient for promoter activity in the root.

We also measured the GUS activities of the successive deletions in the presence or absence of various heavy metals, and the results are shown in Figure 5. When treated with 100 μM Cu²⁺ (Figure 5A), the highest increase in GUS activity was observed in the seedlings that harbored the R1 construct (6.93-fold). The next-highest increase was observed in the R2 construct (6.16-fold). In the R3 construct, Cu²⁺ was still able to induce 3.08-fold greater GUS activity, despite the fact that the distal second and third CuREs had been deleted. However, in the R4 construct (a deletion up to −583, with the deletion of fourth and fifth CuREs) a complete loss of Cu²⁺-induced GUS expression was observed (0.77-fold). These results indicate that the −1052/−583 region of the OsMT-I-4b promoter, containing four CuREs, can lead to efficient Cu²⁺ induction.

Figure 5. β-glucuronidase (GUS) activities in the roots of transgenic plants harboring different deletion constructs. The figure shows the response to treatment with Cu²⁺ (100 μM, A), Pb²⁺ (50 μM, B) and Al³⁺ (50 μM, C). The GUS activities were calculated as the ratio of the promoter activity in the presence of metal stress compared with that in the absence of stress (the induction ratio). The values shown are the means ±SE (n= 3). R1, R2, R3, R4, R5 and R5m indicate the different constructs shown in Figure 4.

Regarding the induction of GUS activity by Pb²⁺, the greatest increase (10.38-fold) was also found using the R1 construct. This level of induction decreased gradually with increasing deletion size from the 5′-end, until 2.01-fold induction remained when using the R5 construct (Figure 5B). In the R5m construct, 1.51-fold induction was detected. In the seedlings treated with 50 μM Al³⁺, deletion of the region −1052/−914 strongly reduced the induction of POsMT-I-4b::GUS expression (Figure 5C). In the R2 construct (−1052/−1), Al³⁺ induced the activity of GUS by 7.97-fold, but the level of induction was only 1.41-fold in the R3 construct (−914/−1). These results indicated that all of the elements included in the 1 730-bp fragment were required in the induction of the OsMT-I-4b promoter by Pb²⁺, and that the region −1052/−914, which contains two CuREs (localized at −1041/−1038 and −1014/−1011), is responsible for the induction of OsMT-I-4b promoter activity by Al³⁺. Combining the results of the deletion analyses, we conclude that different combinations of metal-responsive elements (CuREs and MREs) are involved in the response to different metals.

Discussion

In the present study, we characterized the promoter of a novel rice MT gene, OsMT-I-4b, to reveal its tissue specificity and heavy metal responsiveness. Our analysis indicated that transgenic Arabidopsis plants carrying the OsMT-I-4b promoter demonstrate significant tissue specificity with more intense GUS staining in the roots of seedlings (Figure 2A–D). This is consistent with previous northern blotting results in rice seedlings (Zhou et al. 2006). However, this expression pattern is different from that of another rice MT gene (ricMT, encoding a type II MT protein) promoter, which could mediate GUS accumulating more in aerial parts than in roots (Lü et al. 2007). Of particular interest to us, the relatively low GUS staining in the leaves of transgenic plants was localized mainly in the trichomes and hydathodes (Figure 2E–I), which had been speculated as a pathway for secreting excess heavy metals outside the mesophyll. The detoxification role of trichomes and hydathodes has been demonstrated by previous reports showing that cadmium accumulates preferentially in trichomes in Indian mustard and that nickel accumulates in trichomes in Alyssum lesbiacum (Salt 1995; Kramer et al. 1997). During reproductive growth, histochemical GUS staining is observed in the stigma, filaments and sepals of flowers (Figure 2K–L). It is suggested that OsMT-I-4b may play an important role in flower development by facilitating the transfer and exchange of metal ions required for flower formation and maturation. Similar trichome- and flower-specific expression patterns were detected in other MT genes, including ricMT (Lü et al. 2007), PsMTA (Fordham-Skelton et al. 1997), and AtMT1a (Garcia-Hernandez et al. 1998), indicating distinct but overlapping expression patterns of plant MT genes. Additionally, one point that should be notable is that our results derived from Arabidopsis seem to be not as original as that from rice. However, the present data are consistent with the northern results described previously (Zhou et al. 2006), indicating that there are similar regulatory mechanisms on MT genes expression in monocot rice and dicot Arabidopsis plants. Furthermore, these similarities have also been found in other rice MT genes, including ricMT (Lü et al. 2003) and OsMT-I-1a (our unpublished data). Due to their similarities, as well as the powerful tools available for work with Arabidopsis and the practical advantages of its small size and rapid life cycle, Arabidopsis has been often chosen as a model system for many promoter studies.

We also detected the heavy metal responsiveness of the OsMT-I-4b promoter (Figure 3B,C) and identified the different regions that are relevant for different types of metals (Figure 5). In the promoter region, we found many cis-elements that had been identified previously to be involved in heavy metal-induced gene expression. These included four copies of animal MRE-like motifs and six copies of putative CuRE sequences (Figure 4). In the 5′ deletion analysis, we found that deletion of promoter regions containing different combinations of MREs and CuREs attenuated the metal-inducible function of the OsMT-I-4b promoter to varying degrees (Figure 5). Based on this result, we conclude that different combinations of heavy metal-responsive elements are probably involved in the regulatory variations of downstream MT gene expression in response to different metals. An additional question is whether there are some other metal-regulatory cis-elements, distinct from the core sequences of the known heavy metal-responsive elements. Recently, a novel promoter region (Region II, −187/−147) that confers heavy metal responsiveness has been identified in the promoter of the PvSR2 gene. Sequence analysis found no similarity to any cis-acting elements that had been previously identified to be involved in heavy metal induction, suggesting the presence of a novel heavy metal-responsive element (Qi et al. 2007). To resolve this question and to demonstrate the contribution of each of the known heavy metal-responsive elements in the promoter to the metal-inducible gene expression, further studies using loss of function analyses are required. These include deletion analysis, linker scanning and point mutations.

Furthermore, GUS activity assays indicate that the activity of the OsMT-I-4b promoter might be upregulated by ABA and abiotic stresses (Figure 3A). The involvement of ABA in plant environmental stress responses has long been recognized. Abiotic stresses can cause an increased biosynthesis and accumulation of ABA, and many stress-responsive genes are upregulated by ABA (Rock 2000; Xiong et al. 2002). The induced activity of OsMT-I-4b promoter by both ABA and abiotic stresses revealed the participation of OsMT-I-4b protein in adaptation against various environmental stresses. Furthermore, oxidative stress, resulting from a disturbance in the balance between the production and scavenging of ROS, is an inevitable response when plants are challenged with abiotic stresses. Recently, a number of investigations have demonstrated that MTs can be efficient scavengers of ROS in animals (Li et al. 2006; Peng et al. 2007). This may protect against ROS-induced DNA degradation with an even higher molar efficiency than glutathione (Jourdan et al. 2004). In plants, there is also evidence that MT proteins have a ROS-scavenging function. For example, the expression of LSC54 (a rape MT1 gene) was induced by ROS production and is related to the misbalance of ROS during leaf senescence (Navabpour et al. 2003). In another example, when cgMT1 from beefwood (Casuarina glauca) is overexpressed in transgenic Arabidopsis, a reduced accumulation of H₂O₂ is observed (Obertello et al. 2007). OsMT2b may also function as a ROS scavenger when involved in the response to bacterial blight and blast fungus infections in rice (Wong et al. 2004). The stress-induced OsMT-I-4b promoter activity suggests a potential role for OsMT-I-4b as a ROS scavenger during abiotic stresses.

Taken together, the present work provides a comprehensive description of the regulation of the OsMT-I-4b promoter. In plant transgenic engineering, the choice of promoter is very important. Overexpression of target genes mediated by a constitutive promoter may be detrimental to the plant, and it may be impossible to obtain stable transformants with high levels of widespread overexpression. In this case, the heavy metal-inducible promoter may be advisable. A good example is the yeast copper-inducible promoter system, which has been well characterized in transgenic tobacco plants fused with the GUS reporter gene (Mett et al. 1993) and cytokinin synthase (McKenzie et al. 1998). However, several promoters that are usually induced by Zn²⁺ and Cd²⁺ were found to be inactive in transgenic plants (Qi et al. 2007). Here, we have shown that the OsMT-I-4b promoter can respond to a variety of metal ions in transgenic Arabidopsis, in particular, Cu²⁺, Zn²⁺, Pb²⁺, and Al³⁺ (Figure 3B). Furthermore, the heavy metal-responsive activity of the OsMT-I-4b promoter depended on the type and concentration of the metal ions (Figure 3B,C) and was restricted mainly in the root (Figure 4). The root is the main ion-exchange interface between plants and their environments. Therefore, the OsMT-I-4b promoter could be used as a novel, heavy metal-responsive promoter to control target genes for increasing the heavy metal tolerance of plants or for the accumulation of heavy metals in transgenic plants. In addition, the OsMT-I-4b promoter contains different regions that are relevant to the metal-specific responsiveness. These regions, when fused with reporter genes, could be developed as bioluminescent bioreporters to monitor environmental heavy metal contamination.

Materials and Methods

Plant growth and stress treatment

Arabidopsis thaliana L. (ecotype Columbia) was used in the present study. The plants were grown on agar plates or in soil in a growth chamber (16:8 h light: dark at 22 °C) after vernalizing for 2 d at 4 °C. For growth under sterile conditions, seeds were surface-sterilized and sown on half-strength Murashige and Skoog (MS) salts supplemented with 1.5% sucrose and 0.75% (w/v) agar in Petri dishes (pH 5.8).

To evaluate the ABA and abiotic stress responses of transgenic seedlings harboring the OsMT-I-4b promoter (PMT-I-4b::GUS), we harvested 10-d-old seedlings from the MS agar plates. These were then transferred to Whatman 3 mm paper soaked with distilled water (mock) or 100 μM ABA. To initiate osmotic stress, the seedlings were placed on Whatman paper soaked with 200 mM NaCl (salt stress). For drought stress, the seedlings were dehydrated on Whatman 3 mm paper at 60% humidity. In each case, the plants were subjected to the stress treatments for 6 h, after which all of the shoots and roots of the seedlings were harvested and frozen in liquid nitrogen for quantitative GUS activity assay.

For the heavy metal responsiveness assay of MT-I-4b promoter, we analyzed 10-d-old seedlings from each transgenic line harboring the different promoter regions. The seedlings were transferred to filter paper soaked with CuSO₄ (100 μM), ZnSO₄ (100 μM), CoCl₂ (100 μM), CdCl₂ (50 μM), PbSO₄ (20 μM), and AlCl₃ (50 μM) for 24 h. A fresh water-only control was conducted in parallel. For the dose-dependent response assay, we applied different concentrations of CuSO₄ (50, 100, 200 μM), ZnSO₄ (50, 100, 200 μM), PbSO₄ (10, 20, 50 μM) and AlCl₃ (25, 50, 100 μM). Finally, the roots were collected and assayed for GUS activity. All experiments were carried out with three technical replications, each containing 30–40 seedlings per line.

Generation of the transgenic plants

A series of nested 5′-deletions of OsMT-I-4b promoter fragments (1 730, 1 052, 914, 583, and 142 bp) were amplified by PCR with genomic rice DNA as template using a common antisense primer R-R1 (5′-GCACTAGTCTTGATCTTCTGGGTC-3′) and following different sense primers, F-R1 (5′-CAGGATCCCCCCTCAAAAACTG-3′), F-R2 (5′-TTGGATCCAAATTGAATTC-GTACAGC-3′), F-R3 (5′-CTGGATCCATGATAAACTTAAGTTC-3′), F-R4 (5′-CAGGATCCAAACAGCTAAGAACTTT-3′), and F-R5 (5′-TTGGATCCAATTCCGCAGCTTCTT-3′), respectively. Similarly, we mutated the CuRE sequence of the promoter fragment (−142/−1, R5m) using the primers R-R1 and F-mR5 (5′-CAGGATCCTCTCAAGGTTCCTA-3′). After sequence verification, these derivatives were cloned into the BamHI-SpeI site of pCAMBIA-1381 (Clontech, Palo Alto, CA, USA) upstream of GUS reporter gene taking the place of the CaMV 35S promoter region.

The recombinant plasmids generated were introduced into Agrobacterium tumefaciens GV3101 using the liquid nitrogen freeze-thaw method (Lü et al. 2007) and transferred into Arabidopsis plants using the floral-dip method (Clough and Bent 1998). Homozygous plants were selected from the T₂ progeny and confirmed in the T₃ generations on the basis of hygromycin (25 μg/mL) resistance. Homozygous plants were then used in the experiments.

Histochemical localization and quantitative analysis of GUS activity

GUS activity was assayed histochemically using a modification of the method of Jefferson et al. (1987). Intact young seedlings and various parts of mature PMT::GUS transgenic plants were placed directly in 100 mM phosphate buffer (pH 7.0) containing 10 mM ethylenediaminetetraacetic acid (EDTA), 0.5 mM K₃[Fe(CN)₆], 0.5 mM K₄[Fe(CN)₆], 0.1% Triton X-100 and 1 mM 5-bromo-4-chloro-3-indoyl-β-D-glucuronide (X-Gluc, Sigma, St. Louis, MO, USA) at 37 °C for 16–24 h. After the chlorophyll had been removed using 70% ethanol, we photographed the samples under an SZX12 anatomy microscope (Olympus, Tokyo, Japan) equipped with a DP70 digital camera system.

GUS activity was analyzed quantitatively in 10-d-old transgenic PMT::GUS seedlings using fluorometric measurement with 4-methyl umbelliferyl glucuronide (4-MUG) as the substrate according to the method previously described by Lü et al. (2007).

Statistical analyses

All data are expressed as the means ±SD. For the statistical analysis, we used a two-tailed, unpaired Student's t-test. All differences were considered significant at P < 0.05.

(Co-Editor: Peter Doerner)

Acknowledgements

We thank members of the Laboratory of Molecular Biology at Tsinghua University for comments and participation in discussions. This work was supported by grants from the National Transgenic Animals & Plants Research Project (2009ZX08009-069B), the State Key Basic Research and Development Plan of China (2006CB101706), and the National Natural Science Foundation of China (30870197).