Planta

, Volume 232, Issue 4, pp 925–936

The oil palm metallothionein promoter contains a novel AGTTAGG motif conferring its fruit-specific expression and is inducible by abiotic factors

Authors

  • Vahid Omidvar
    • Department of Agriculture Technology, Faculty of AgricultureUniversity Putra Malaysia (UPM)
    • Laboratory of Plantation CropsInstitute of Tropical Agriculture, UPM
    • Department of Agriculture Technology, Faculty of AgricultureUniversity Putra Malaysia (UPM)
    • Laboratory of Plantation CropsInstitute of Tropical Agriculture, UPM
  • Amir Izadfard
    • Department of Agriculture Technology, Faculty of AgricultureUniversity Putra Malaysia (UPM)
  • Chai Ling Ho
    • Department of Cell and Molecular Biology, Faculty of Biotechnology and Biomolecular SciencesUniversity Putra Malaysia (UPM)
  • Maziah Mahmood
    • Department of Cell and Molecular Biology, Faculty of Biotechnology and Biomolecular SciencesUniversity Putra Malaysia (UPM)
Original Article

DOI: 10.1007/s00425-010-1220-z

Cite this article as:
Omidvar, V., Abdullah, S.N.A., Izadfard, A. et al. Planta (2010) 232: 925. doi:10.1007/s00425-010-1220-z

Abstract

The 1,053-bp promoter of the oil palm metallothionein gene (so-called MSP1) and its 5′ deletions were fused to the GUS reporter gene, and analysed in transiently transformed oil palm tissues. The full length promoter showed sevenfold higher activity in the mesocarp than in leaves and 1.5-fold more activity than the CaMV35S promoter in the mesocarp. The 1,053-bp region containing the 5′ untranslated region (UTR) gave the highest activity in the mesocarp, while the 148-bp region was required for minimal promoter activity. Two positive regulatory regions were identified at nucleotides (nt) −953 to −619 and −420 to −256 regions. Fine-tune deletion of the −619 to −420 nt region led to the identification of a 21-bp negative regulatory sequence in the −598 to −577 nt region, which is involved in mesocarp-specific expression. Gel mobility shift assay revealed a strong interaction of the leaf nuclear extract with the 21-bp region. An AGTTAGG core-sequence within this region was identified as a novel negative regulatory element controlling fruit-specificity of the MSP1 promoter. Abscisic acid (ABA) and copper (Cu2+) induced the activity of the promoter and its 5′ deletions more effectively than methyl jasmonate (MeJa) and ethylene. In the mesocarp, the full length promoter showed stronger inducibility in response to ABA and Cu2+ than its 5′ deletions, while in leaves, the −420 nt fragment was the most inducible by ABA and Cu2+. These results suggest that the MSP1 promoter and its regulatory regions are potentially useful for engineering fruit-specific and inducible gene expression in oil palm.

Keywords

Biolistic transformationFruit-specific promoterGel mobility shift assayInducible expressionOil palm metallothioneinTransient expression assay

Abbreviations

UTR

Untranslated region

GUS

β-Glucuronidase

GFP

Green fluorescent protein

EMSA

Electrophoretic mobility shift assay

MT

Metallothionein

ERE

Ethylene-responsive element

ABRE

Abscisic acid-responsive element

MeJa-RE

Methyl jasmonate-responsive element

WUN-RE

Wounding-responsive element

Cu2+-RE

Copper-responsive element

Introduction

Metallothioneins (MTs) are low-molecular weight cysteine-rich proteins that can bind and sequester ionic forms of certain metals, and have potential roles in heavy metal detoxification, homeostasis of essential metal ions, and protection against oxidative damage (Wong et al. 2004). Based on the arrangement of cysteine residues, MTs are categorized into three classes (Robinson et al. 1993; Rauser 1999; Chatthai et al. 2004). Class I includes MTs from mammalians and vertebrates, class II refers to MTs from plants, fungi, and invertebrate animals, and class III MTs are enzymatically synthesized polypeptides with a poly γ-glutamylcysteine structure. The EcMT protein from wheat was the first MT identified in higher plants (Lang et al. 1987). Since then, many MT genes have been isolated and characterized from plants. Plant MTs are further divided into four types based on the distribution and arrangement of their cysteine residues (Robinson et al. 1993; Cobbett and Goldsbrough 2002). Type 1 MTs are primarily expressed in plant roots and leaves (Foley et al. 1997; Hsieh and Huang 1998), while type 2 MTs are expressed abundantly in leaves and roots of mature plants (Hsieh et al. 1996). Type 3 MT transcripts are mainly found in developing fruits of plants such as banana, kiwi, apple, oil palm and mandarin (Ledger and Gardner 1994; Clendennen and May 1997; Reid and Ross 1997; Siti Nor Akmar et al. 2002; Endo et al. 2007), even though there are reports of their occurrence in leaves (Bratic et al. 2009), and embryos (Dong and Dunstan 1996). Type 4 MT genes such as the wheat EcMT and rice OsMT-11-1a are preferentially expressed in developing seeds (Kawashima et al. 1992; Zhou et al. 2005). Plant MT genes were also reported to be induced in response to a wide range of biotic and abiotic factors (for review see Rauser 1999). In all cases, the differential tissue-specific and inducible-expression pattern of MT genes suggest that they may have different transcriptional regulation mechanisms and specialized functions in different plant tissues.

Several studies have focused on the tissue-specificity of plant MT promoters (Fordham-Skelton et al. 1997; Hsieh and Huang 1998; Chatthai et al. 2004). For instance, 5′ deletion analysis of the pea PsMTA promoter in transgenic Arabidopsis showed that the −583/−285 nt region was responsible for GUS expression in roots (Fordham-Skelton et al. 1997). Analysis of the buckwheat FeMT3 promoter in transgenic tobacco revealed strong activity of the promoter in leaves, roots, and pollen (Bratic et al. 2009). Characterization of the rice OsMT2b promoter in transgenic Arabidopsis led to the identification of regulatory regions responsible for differential expression in vegetative and reproductive organs (Ren and Zhao 2009).

In silico analysis of several plant MT promoters revealed the presence of cis-acting regulatory elements that confer responsiveness to abiotic and biotic stresses, such as the ethylene-responsive elements (ERE) (Whitelaw et al. 1995), the abscisic acid-responsive elements (ABRE) (Zhou and Goldsbrough 1995), the W-box involved in wounding response (Endo et al. 2007), the E-box involved in defense signaling, and the MYB-binding sites involved in drought inducibility (Lü et al. 2007). Tissue-specific elements such as the root-specific element (RSE) have been identified in MT promoters from pea (Fordham-Skelton et al. 1997) and oil palm (Siti Nor Akmar et al. 2002). A few reports described the occurrence of a metal-responsive element (MRE) in plant MT promoters including the pea PsMTA promoter (Fordham-Skelton et al. 1997), the tomato LeMTB promoter (Whitelaw et al. 1995), the Douglas Fir PmMT promoter (Chatthai et al. 2004), and the rice ricMT promoter (Lü et al. 2007). The fact that plant MT promoters are enriched with tissue-specific, stress-inducible and metal-responsive regulatory elements suggests the presence of a complex regulatory mechanism controlling the expression of MT genes. Although several plant MT promoters have been characterized in detail, there are very limited reports on their applications to drive the expression of transgenes.

Promoters of type 3 MT genes are important in particular due to their fruit-specific activity. A fruit-specific type 3 MT gene (MT3-A) was previously isolated from oil palm and shown to be developmentally regulated during fruit ripening (Siti Nor Akmar et al. 2002). Analysis of the promoter region of the MT3-A gene (designated MSP1; GenBank accession no. EU499363) revealed a strong activity in the mesocarp tissue of oil palm (Siti Nor Akmar and Zubaidah 2007). In a recent study, the fruit-specific activity of the MSP1 promoter was demonstrated using GUS histochemical assay and the promoter was employed to drive the transient fruit-specific expression of the entire polyhydroxybutyrate biosynthetic gene pathway in oil palm (Omidvar et al. 2008). Fruit-specific promoters are crucial for improving the nutritional value and quality of fruits by genetic engineering. The MSP1 promoter would be specifically useful for functional gene studies and biotechnology applications through engineering genes of interest for improved palm oil metabolism, as well as for production of high-value novel derivatives and biodiesel to sustain the oil palm industry. However, a detailed analysis of the MSP1 promoter regulatory regions is required to provide a better understanding of the mechanism controlling the fruit-specific and developmentally regulated expression of the MT3-A gene.

The present study aimed specifically to (1) evaluate the potential of using a transient expression assay for the functional analysis of the MSP1 promoter in oil palm; (2) study the tissue-expression pattern of the MSP1 promoter and identification of the cis-elements involved in tissue-specific regulation of the MT3-A gene; (3) study the effect of abiotic factors and heavy metal toxicity on the MSP1 promoter activity. In order to achieve these objectives, a series of oil palm MSP1 promoter deletion fragments were fused to GUS reporter gene and transiently expressed in different oil palm tissues and under stress conditions. Electrophoretic mobility shift assay (EMSA) was employed to confirm the interaction of a newly identified 21-bp fruit-regulatory region with leaf nuclear protein extract.

Materials and methods

Construction of 5′ deletions of the MSP1 promoter

The 5 deletions of the MSP1 promoter containing the 5-UTR were generated by PCR amplification. The primers used for PCR are provided in Suppl. Material S1. The amplified fragments were digested with HindIII and BglII, and cloned as translational fusions with the GUS reporter gene into the HindIII/BglII digested pCAMBIA1305.1 vector (CAMBIA, Brisbane, Australia), replacing the CaMV35S promoter, and resulting in the following recombinant constructs: D1 (−983/+72), D2 (−619/+72), D3 (−420/+72), D4 (−256/+72), D5 (−116/+72), and D6 (−76/+72) (Fig. 1). In addition, the D1 construct carrying the full length promoter region just upstream to the transcription start site (−983/+1) was generated to examine the effect of the 5-UTR on promoter activity. The presence and orientation of the 5′ deletion fragments were confirmed by sequencing. The pCAMBIA1305.1 carrying the 35S:GUS fusion and the 35SpEGFP1 with 35S:GFP (green fluorescent protein) fusion were used as controls (Suppl. Material S2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-010-1220-z/MediaObjects/425_2010_1220_Fig1_HTML.gif
Fig. 1

Schematic representation of constructs containing the full length promoter with the 5′-UTR (D1), without the 5′-UTR (D1′), and 5′ deletions of D1 (D2, D3, D4, D5, and D6). The numbers indicate the nucleotide position of each deletion. The symbols on each promoter region represent deduced cis-acting elements that are annotated below the schematic diagram. TSS transcription start site

Fine-tune 5′ deletion analysis of the −619 to −420 nt region

Successive deletions were made at the region located between nt −619 and −420 by generating fragments differing about 20 bp in length using PCR amplification. The primers used for PCR are provided in Suppl. Material S3. The amplified fragments were cloned as translational fusions with the GUS gene by replacing the existing CaMV35S promoter in the pCAMBIA 1305.1 vector, resulting in the following chimeric constructs: D2 (−619/+72), D2.1 (−598/+72), D2.2 (−577/+72), D2.3 (−558/+72), D2.4 (−539/+72), D2.5 (−520/+72), D2.6 (−501/+72), D2.7 (−481/+72), D2.8 (−462/+72), and D2.9 (−443/+72). The presence and orientation of the 5′ deletions were confirmed by sequencing.

Preparation of target tissue for bombardment

Preparation of target tissue was carried out based on the procedure described by Agius et al. (2005). Oil palm (Elaeis guineensis, D × P) fruits (12 weeks postanthesis) and young leaves were obtained from MPOB (Malaysian Palm Oil Board, Malaysia) and surface sterilized in 20% (v/v) Clorox, followed by thorough rinsing with sterile distilled water. Fruit mesocarp and leaves were excised into 1 × 1 cm sections, and placed at the centre (approximately 3 cm in diameter) of petri dishes containing Murashige and Skoog medium (MS; Murashige and Skoog 1962). The tissue sections were kept at 28°C in a growth chamber for 1 day prior to bombardment. Abscisic acid, methyl jasmonate, and metal treatments were performed by incubating the tissue sections for 4 h in MS media containing 100 μM ABA, 100 μM MeJa or 150 μM CuSO4, respectively. Ethylene treatment was performed for 6 h in 2 mL of 40% (w/v) ethephon, and 59.52 mM sodium bicarbonate (ethylene gas is liberated by ethephon under such conditions) (Zhang et al. 2004).

Biolistic transformation

Gold microcarriers were prepared following instructions of the manufacturer (Bio-Rad Laboratories, Richmond, CA, USA). Sixty milligrams of 1-μm diameter gold particles were suspended in 1 mL of 100% ethanol, vigorously vortexed for 2 min, and then centrifuged at 12,000g for 1 min. The recovered particle pellet was washed twice with sterile distilled water before resuspension in 1 mL of sterile distilled water. For each co-bombardment, 7 μg of each deletion construct, 3 μg of 35SpEGFP1 construct (internal standard), 80 μL of CaCl2 (2.5 M) and 50 μL of spermidine (0.1 M) were added one by one into an aliquot of 80 μL of gold particle suspension with continuous vortexing. The mixture was vortexed for an additional 3 min, and then the DNA-particle pellet was recovered by centrifugation at 12,000g for 10 s. The pellet was washed twice with 250 μL of 100% ethanol, and was resuspended in 60 μL of 100% ethanol.

Biolistic transformation was carried out using the PDS-1000/He Biolistic® Particle Delivery System (Bio-Rad Laboratories) by co-bombardment of each 5′ deletion construct and 35SpEGFP plasmid into mesocarp and leaf tissues. In a parallel experiment, co-bombardment of pCAMBIA1305.1 and 35SpEGFP1 plasmids was carried out in order to provide a quantitative reference for the comparison of the MSP1 and CaMV35S promoters. For each co-bombardment, 10 μL of DNA–gold complex was used and the system was set up based on the parameters optimized for transient transformation of oil palm tissues as described by Zubaidah and Siti Nor Akmar (2003). The bombardment was carried out using rupture disks at 1,550 and 1,100 psi for mesocarp and leaf tissues, respectively, while the vacuum was maintained at 27 mmHg pressure. The bombarded tissues were kept at 28°C for 48 h prior to reporter gene assay.

GUS and GFP fluorometric assays

Quantitative analysis of the GUS activity in transiently transformed mesocarp and leaf tissues was carried out according to Jefferson (1987). The bombarded tissues were ground into a fine powder in liquid nitrogen and homogenized in 2 volumes of GUS extraction buffer (50 mM NaPO4 buffer, pH 7.0, 10 mM EDTA, pH 8, 0.1% SDS, 0.1% Triton X-100, 10 mM β-mercaptoethanol, and 25 μg mL−1 phenylmethylsulfonyl fluoride). The crude protein extract was recovered by centrifugation at 14,000g for 15 min at 4°C. Protein extract (20 μL) was added to pre-warmed (37°C) 4-methylumbelliferyl-β-d-glucuronide (4-MUG) assay buffer (1 mM 4-MUG in GUS extraction buffer) (1 mL), mixed thoroughly and incubated at 37°C. Each of the reaction was stopped at 30, 60, and 90 min by transferring 100 μL of the reaction mixture into a tube containing 900 μL of stop buffer (1 M sodium carbonate). The fluorescence generated from each sample was measured based on 4-methylumbelliferone (4-MU) standard curve, using the TECAN Microplate Reader (SiberHegner Sdn. Bhd., Malaysia). Protein concentration in each sample was determined using the Quick Start™ Bradford Protein Assay Kit (Bio-Rad Laboratories). The GUS activity was calculated in pmol MU min−1 mg protein−1. GFP quantification was performed according to the protocol described by Clontech (Clontech, Mountain View, CA, USA). Protein samples were diluted to 10 μg mL−1, and 100 μL of each diluted sample were used for GFP analysis. A series of dilutions of the standard rAcGFP1 (Clontech) (0.1–1 μg mL−1) was made in 100 μL of protein extract from non-transformed tissues containing 10 μg mL−1 of protein. The spectral properties of rAcGFP1 resemble those of EGFP. The GFP fluorescence intensity of each sample was measured at 485 nm excitation and 550 nm emission using the TECAN Microplate Reader.

Electrophoretic mobility shift assays

EMSAs were carried out using the LightShift Chemiluminescent EMSA Kit (Pierce, Rockford, IL, USA). The 21-bp oligonucleotide identified at nt −598 to −577 (CATTTTGATTAGTTAGGCTAA) was synthesized in both forward and reverse directions. Labelling was carried out using Biotin 3′ End DNA Labelling Kit (Pierce). The terminal deoxynucleotidyl transferase (TdT) incorporates 1–3 biotinylated ribonucleotides onto the 3′ end of the DNA strand. Each single-stranded oligonucleotide was labelled in a reaction containing 5 μL of oligo-templates (1 μM), 5 μL of TDT enzyme (2 U μL−1), 5 μL of biotin-11-dUTP (5 μg), 10 μL of 5× TdT reaction buffer, and 25 μL of ultrapure water. The end-labelled complementary oligos were annealed to form the biotin-labelled double-stranded probe.

Nuclear protein extracts were prepared from intact mesocarp, leaf, root and kernel tissues using the NE-PER Nuclear and Cytoplasmic Extraction Kit (Pierce), in order to study their binding pattern with the 21-bp regulatory region of the MSP1 promoter. Binding of the DNA probe with nuclear extract was carried out in a 20 μL binding reaction containing 1× binding buffer, 2.5% (v/v) glycerol, 5 mM MgCl2, 50 ng μL−1 of poly(dI-dC), 0.05% (v/v) NP-40, 3 μL of nuclear protein extract, and 20 fmol of biotin-labelled probe. Four picomoles of the unlabelled probe was used for competition with the labelled probe in EMSAs. The binding reactions were separated on a 6% native polyacrylamide gel at 100 V for 1 h, transferred onto a nylon membrane and UV cross-linked. DNA–protein interactions were detected using the Chemiluminescent Nucleic Acid Detection Module (Pierce). The chemiluminescent signals were captured using the gel documentation system equipped with a CCD camera and a chemiluminescent filter.

Experimental design and data analysis

For particle bombardment experiments, samples were arranged in a completely randomized design with three replications (five samples per replications). The data obtained from quantification of GUS and GFP were subjected to analysis of the variance (ANOVA) and Duncan New Multiple Range Test (DNMRT) using SAS software version 6.12. In order to normalize the differences in each bombardment efficiency on GUS activity, expression data for each 5′ deletion was reported as the ratio of GUS to GFP values.

Results

Sequence characterization of the oil palm MSP1 promoter

The sequence of the MSP1 promoter (Suppl. Material S4) was analysed for the presence of putative fruit-specific and other cis-acting elements involved in regulation of the MT3-A gene expression using PLACE and PlantCARE databases (Higo et al. 1999; Lescot et al. 2002). A number of potential regulatory elements were identified corresponding to known cis-acting elements of eukaryotic genes (Table 1). A putative TATA box at position −34 and a CAAT (CCAATT) box at position −348, which are common to eukaryotic promoters, were identified. Several important motifs were detected, including two copies of I-box (Terzaghi and Cashmore 1995) and MeJa-responsive element (RE) (Rouster et al. 1997), one copy of ABRE (Hobo et al. 1999), an ABA-/sugar-RE element (Acevedo-Hernandez et al. 2005), a G-box (Nash et al. 1990), a WUN-RE element (Pastuglia et al. 1997), and an ERE (Siti Nor Akmar et al. 2002). In addition, two sequences that are similar to the copper (Cu2+)-RE element of the green alga Chlamydomonas Cyc6 gene (Quinn and Merchant 1995) were identified at positions −434 and −252 of the MSP1 promoter.
Table 1

Putative cis-regulatory elements enriched in the MSP1 promoter

Name and sequence of element

Function

Position relative to transcription start site

TATA box

(+)TATAAAT

Common core promoter element

−34

CAAT-box

(+)CCAATT

Common cis-acting enhancer

−348

I-box

(+)GATAA

Light-responsive element

−942, −603

G-box

(+)CACGAC

Light-responsive element

−286

ABRE

(+)ACGTAGC

Abscisic acid-responsive element

−237

ABA- and Sugar-RE

(+)CACCTCCA

Abscisic acid and sugar responsive element

−931

MeJa-RE

(+)CGTCA

Methyl jasmonate-responsive element

−340, −289

ERE

(−)TTAGAATT

Ethylene-responsive element

−317

WUN-RE

(+)AAATTTCCT

Wounding-responsive element

−358

Cu2+-RE

(+)GTAC

Copper-responsive element

−434, −252

MYB motif

(+)WAACCA

Dehydration-responsive element

−809, −729, −521,−491, −367, −156

Dof cis-acting element

(+)AAAG

Involved in expression of a variety of signal response and/or tissue-specific genes in

−837, −537, −513, −504, −469, −451, −392, −373, −276, −189, −169, −90

POLLEN1LETAL52 motif

(+)AGAAA

Involved in pollen-specific expression

−539, −516, 502, −454, −193

Skn-1 motif

(−)GTCAT

Involved in endosperm-specific expression

−177

XYLAT motif

(+)ACAAAGAA

Involved in xylem-specific expression

−839, −278

The MSP1 promoter sequence is provided in Supplementary Material S4. Elements were identified using PLACE (http://www.dna.affrc.go.jp/PLACE/) (Higo et al. 1999) and PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al. 2002) databases

W A+T, (+): sense strand, (−): antisense strand

Deletion analysis of the MSP1 promoter in mesocarp tissue

To identify the functional significant domains within the MSP1 promoter, a series of 5′ deletions were prepared as a translational fusion with the GUS reporter gene in the pCAMBIA1305.1 (Fig. 1) For data normalization, all the test constructs were co-bombarded with the 35SpEGFP1 construct carrying the GFP reporter gene under the control of the CaMV35S promoter. The pCAMBIA1305.1 vector carrying the CaMV35S promoter was used as a reference for evaluating the strength of the MSP promoter. The promoter activity of each 5′ deletion construct was measured in mesocarp and leaf tissues as a ratio of GUS to GFP values (Fig. 2).
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-010-1220-z/MediaObjects/425_2010_1220_Fig2_HTML.gif
Fig. 2

Transient expression of the MSP1 promoter and its 5′ deletion constructs in oil palm mesocarp (a) and leaf (b) tissues. The activity of each deletion is reported as the ratio of GUS (pmol MU min−1 mg protein −1) to GFP (μg mL−1). Data are expressed as means of three separate bombardments (each replicated five times) with the standard error. Bars with the same letter are not significantly different at P = 0.05 according to DNMRT

The D1 construct carrying the full length sequence of the MSP1 promoter produced the highest level of expression (GUS/GFP ratio) in mesocarp tissue (Fig. 2a), indicating that the 1,053-bp region upstream of the translation start site is functionally required for the maximal MSP1 promoter activity. The D2 deletion construct (−619 to +72) showed lower expression (twofold less) compared to that of the D1 construct (−983 to +72), indicating the presence of a strong enhancing region in the −983 to −619 nt deleted fragment, which is required for high activity of the promoter in the mesocarp. Further deletion to nt −420 (D3 construct: −420 to +72) resulted in a slight increase in the promoter activity, indicating the presence of a negative regulatory region in the −619 to −420 nt deleted fragment. The D4 deletion construct (−256 to +72) showed a lower expression level (1.3-fold less) than the D3 construct, indicating the presence of positive regulatory element(s) in the deleted −420 to −256 nt region. No further drop in activity of D5 (−116 to +72) and D6 (−76 to +72) deletion constructs was observed, indicating that the 148-bp region (−76 to +72) that is located upstream of the translation start site and containing the TATA box is considered as the MSP1 minimal promoter. It is noted that the 1,053-bp full length promoter displayed a 1.5-fold higher activity than the constitutive CaMV35S promoter in mesocarp tissue.

Deletion analysis of the MSP1 promoter in leaves and assessment of tissue-specificity of the different 5′ deletions

As illustrated in Fig. 2b, the activity of the 1,053-bp full length MSP1 promoter and its 5′ deletions was considerably low in leaves, while in mesocarp, the activity of the full length promoter was sevenfold higher. This reflects the strong mesocarp-specificity of the MSP1 promoter. Interestingly, the D2 deletion construct (−619 to +72) was found to direct the least expression of the GUS reporter gene. Deletion of the −619 to −420 nt region resulted in a dramatic 2.6-fold increase in the activity of the D3 deletion construct (−420 to +72) in leaves, which was surprisingly higher (1.3-fold) than that of the full length promoter. This observation suggests the presence of negative cis-acting regulatory element(s) in the deleted −619 to −420 nt region, which probably control the fruit-specific activity of the promoter by suppressing its activity in leaves. However, sequence analysis of this region did not reveal any known cis-acting regulatory elements involved in fruit-specific expression. Further deletions to positions −256 (D4 construct) and −116 (D5 construct) resulted in a gradual decrease of the promoter activity while the activity level did not drop by deleting the promoter to position −76 (D6 construct).

The −598 to −577 nt region is responsible for fruit-specific activity of the MSP1 promoter

In order to identify the critical region conferring mesocarp-specific activity of the MSP1 promoter, the region between nt −619 and −420 was subjected to fine-deletion analysis. The D2.1 construct directed the lowest expression of the reporter gene in leaves, while deletion of the 21-bp at position −598 to −577 (D2.2 construct) induced the highest promoter activity in leaves by 3.5-fold (Table 2). Interestingly, expression of the rest of the constructs (D2.3–D2.9) in leaves remained significantly higher than the D2.1 construct. This finding indicates that the region between nt −598 and −577 acts as a negative cis-acting regulatory element conferring mesocarp specificity by inhibiting the gene expression in leaves, which could be due to the interaction of the leaf-specific transcription factors with the identified regulatory region. No previously identified motif was found in the 21-bp sequence, suggesting the presence of novel motif(s) in this region.
Table 2

Transient expression of the MSP1 promoter successive deletions at nt −598 to −577 (D2.1–D2.9) in oil palm mesocarp and leaf tissues

5′ deletion

GUS/GFP

Mesocarp

Leaf

D2 (−619/+72)

1,110 ± 127 (a)

170 ± 34 (c,d)

D2.1 (−598/+72)

1,290 ± 151 (a)

140 ± 29 (d)

D2.2 (−577/+72)

1,400 ± 169 (a)

490 ± 56 (a)

D2.3 (−558/+72)

1,370 ± 160 (a)

340 ± 41 (b)

D2.4 (−539/+72)

1,350 ± 112 (a)

290 ± 43 (b,c,d)

D2.5 (−520/+72)

1,290 ± 171 (a)

380 ± 45 (a,b)

D2.6 (−501/+72)

1,300 ± 142 (a)

300 ± 50 (b,c)

D2.7 (−481/+72)

1,360 ± 163 (a)

320 ± 31 (b)

D2.8 (−462/+72)

1,310 ± 138 (a)

350 ± 53 (a,b)

D2.9 (−443/+72)

1,230 ± 165 (a)

360 ± 41 (a,b)

The activity of each deletion is reported as the ratio of GUS (pmol MU min−1 mg protein−1) to GFP (μg mL−1). Data are expressed as means of three separate bombardments (each replicated five times) with the standard error. Means followed by the same letter are not significantly different at P = 0.05 according to DNMRT

Sequence-specific interaction of the 21-bp negative regulatory region with the leaf nuclear proteins

To examine the DNA sequence-specific protein binding of the identified 21-bp negative regulatory region (designated NR), EMSA was carried out using nuclear protein extracts from different oil palm tissues including mesocarp, leaf, root, and kernel tissues (Fig. 3). Nuclear extracts from mesocarp, root, and kernel tissues did not cause any retardation of the free NR probe (Fig. 3, lanes 1, 3 and 4), indicating the absence of any DNA–protein interaction in these tissues. However, challenging the NR probe with the leaf nuclear extract resulted in a major retarded band, showing the presence of specific DNA–protein interaction in leaves (Fig. 3, lane 2). In order to examine the specificity of the interaction in leaves, the labelled NR probe was challenged with 200-fold molar excess of the unlabeled probe in a competition EMSA (Fig. 3, lane 5), and it is shown that the band shift observed with the NR probe and leaf extract (Fig. 3, lane 2) can be inhibited by competition from excess amount of the unlabeled specific probe.
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-010-1220-z/MediaObjects/425_2010_1220_Fig3_HTML.gif
Fig. 3

EMSA showing the interaction of the 21-bp regulatory region (NR) with nuclear protein extract from mesocarp, leaf, root, and kernel tissues. Lane 0 position of the free NR probe, lanes 14 interaction of the NR probe with the nuclear extracts from mesocarp, leaf, root, and kernel tissues, respectively. Lane 5 competition of the NR probe with 200-fold molar excess of the competitor (unlabeled probe). F free probe, B bound complex

Identification of the core-sequence responsible for the specific DNA–protein binding in leaves

In order to identify the core-sequence involved in leaf-specific protein–DNA binding, a series of truncated versions of the NR probe were synthesized, with every four successive nucleotides removed in each version. These unlabeled truncated probes were designated as tNR1, tNR2, tNR3, tNR4, and tNR5, and used as competitors with the NR probe in EMSAs (Fig. 4). The tNR1, tNR2, tNR3, and tNR5 probes showed strong competition with the NR probe by preventing the band shift, while the tNR4 probe completely lost its competition with the NR probe due to removal of TTAG nucleotides (Fig. 4). In order to find the actual length of the motif and further examine its sequence-specific binding, mutant versions of the NR probe were prepared by introducing single point mutations in the TTAG sequence as well as in its flanking three nucleotides at either side. The mutated derivatives were referred to as mNR1, mNR2, mNR3, mNR4, mNR5, mNR6, mNR7, mNR8, mNR9 and mNR10, and used as competitors with the non-mutant NR probe in EMSAs (Fig. 5). The mNR2 to mNR8 probes lost their competition with the NR probe due to a single-base mutation in their sequence, while mutations in the mNR1, mNR9, and mNR10 probes did not suppress their protein-binding activity (Fig. 5). Therefore, the AGTTAGG sequence was considered as the actual motif involved in fruit-specific activity of the MSP1 promoter.
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-010-1220-z/MediaObjects/425_2010_1220_Fig4_HTML.gif
Fig. 4

Competition EMSAs showing interaction of the 21-bp regulatory region (NR) and its truncated derivatives with nuclear protein extract from leaves. Lane 0 position of the free NR probe. Lane 1 interaction of the NR probe with leaf nuclear extract. Lanes 26 interaction of the truncated tNR1–tNR5 probes with leaf nuclear extract, respectively in competition with the NR probe. F free probe, B bound complex

https://static-content.springer.com/image/art%3A10.1007%2Fs00425-010-1220-z/MediaObjects/425_2010_1220_Fig5_HTML.gif
Fig. 5

Competition EMSAs showing interaction of the 21-bp regulatory region (NR) and its mutated derivatives with nuclear protein extract from leaves. Lane 0 interaction of the NR probe with nuclear extract. Lanes 110 interaction of the mutated mNR1–mNR10 probes with leaf nuclear extract, respectively, in competition with NR probe. The mutated nucleotide is shown in green colour. F free probe, B bound complex

Responsiveness of the MSP1 promoter and its 5′ deletions to abiotic stress and heavy metal toxicity in mesocarp and leaf tissues

As illustrated in Table 3, the activity of the full length MSP1 promoter was highly induced by ABA (1.6-fold) and Cu2+ (1.4-fold) treatments in mesocarp tissue. In comparison, the MSP1 promoter activity was shown to be less responsive to MeJa and ethylene treatments. Notably, a deletion from nt −983 to −619 resulted in a dramatic decrease in the promoter activity in response to ABA (almost twofold) in mesocarp tissue. In addition, deletion from nt −256 to −116 resulted in 1.43-fold decrease in activity of the −116 deletion construct in response to ABA, while deletion of this region did not cause much difference in the promoter activity in untreated mesocarp tissue. This suggests that the −983 to −619 and −256 to −116 nt regions contain potential cis-regulatory elements that are responsive to ABA. Interestingly, in silico analysis of the MSP1 promoter revealed the presence of two ABA-/sugar-responsive and ABRE elements at positions −931 and −237, respectively (Table 1).
Table 3

Transient expression of the MSP1 promoter and its 5′ deletions in response to various stresses in oil palm mesocarp and leaf tissues

5′ deletion

Fold increase in GUS/GFP

ABA

Cu2+

MeJa

Ethylene

Mesocarp

Leaf

Mesocarp

Leaf

Mesocarp

Leaf

Mesocarp

Leaf

D1 (−983 to +72)

1.6 (a)

1.7 (b)

1.4 (a)

1.5 (a)

1.2 (a)

1.3 (b)

1.1 (a)

1.25 (b)

D2 (−619 to +72)

1.57 (c)

1.6 (c)

1.35 (b)

1.4 (b)

1.3 (b)

1.14 (c)

1.2 (b)

1.3 (c)

D3 (−420 to +72)

1.4 (b)

1.9 (a)

1.1 (b)

1.5 (a)

1.03 (b)

1.4 (b)

1.05 (b)

1.3 (a)

D4 (−256 to +72)

1.4 (c)

1.3 (c)

1.4 (b)

1.1 (b)

1.1 (c)

1.05 (c)

1.04 (c)

1.05 (c)

D5 (−116 to +72)

1.1 (d)

1.1 (c)

1.01 (c)

1.1 (b)

1.02 (d)

1.0 (c)

1.05 (c)

0.9 (c)

D6 (−76 to +72)

1.04 (d)

1.1 (c)

1.01 (c)

1.1 (b)

1.01 (d)

1.1 (c)

1.03 (c)

1.0 (c)

The activity of each deletion is reported as the ratio of GUS (pmol MU min−1 mg protein−1) to GFP (μg mL−1), and is represented as the fold change in response to ABA, MeJa, Cu2+ and ethylene relative to no treatment. Data are expressed as means of three separate bombardments (each replicated five times) with the standard error. Means followed by the same letter are not significantly different at P = 0.05 according to DNMRT

An interesting pattern was observed in activity of the 5′ deletion series in response to Cu2+ in mesocarp tissue. Deletion of the region at nt −256 to −116 resulted in 1.5-fold decrease in activity of the −116 deletion construct in treated mesocarp tissue (Table 3), but produced no effect in untreated mesocarp tissue. Sequence analysis of the −256 to −116 nt region revealed the presence of a Cu2+-RE element (Table 1). It was shown that the 5′ deletion series were relatively less responsive to MeJa and ethylene treatments than ABA and Cu2+ metal ions (Table 3). Although MeJa and ethylene treatments did induce the activity of the MSP1 promoter and its 5′ deletions to some extent, not much significant changes were observed in comparison to the promoter activities in untreated mesocarp tissue. Furthermore, activity of the minimal 148-bp promoter remained constant in untreated and treated mesocarp tissue, indicating the absence of cis-acting regulatory elements that are responsive to ABA, Cu2+, MeJa and ethylene in this promoter fragment.

Data analysis of the effect of the hormonal and heavy metal treatments on promoter activity and its 5′ deletions in leaves (Table 3) showed that ABA and Cu2+ induced activity of the full length 1,053-bp promoter up to 1.7- and 1.5-fold, respectively, while MeJa and ethylene increased its activity just up to 1.32- and 1.25-fold, respectively. Among the 5′ promoter deletions, the −420 nt deletion displayed the highest activity in leaves in response to ABA in comparison to other treatments. Although ABA, Cu2+, MeJa and ethylene treatments resulted in an overall increase in activities of the 5′ deletions in leaves, they had the most inducing effect on the −420 deletion of the promoter, which arguably could be due to the removal of the putative mesocarp-specific regulatory region at position −619 to −420. The −256, −116, and −76 nt deletion constructs did not show any significant differences in activity in treated and untreated leaves.

Effect of the 5′-UTR on the MSP1 promoter activity in mesocarp and leaf tissues

In order to investigate the effect of the 5′-UTR region on the MSP1 promoter activity, the D1′ construct was generated carrying the full length promoter up to the transcription start site (without the 5′-UTR). It is shown that the D1 construct with the 5′-UTR was able to direct the expression of the reporter gene 1.5-fold more than the D1′ construct in mesocarp tissues (Fig. 6a), indicating the importance of the 5′-UTR as a positive regulatory region for the promoter activity. Interestingly, almost the same pattern was observed in the activities of the D1 and D1′ constructs in the mesocarp tissue in response to hormonal and heavy metal treatments (Fig. 6a). The 5′-UTR did not exert any effect on the promoter activity in leaves, since the D1 and D1′ constructs showed no significant difference in the promoter activity in either treated or untreated leaves (Fig. 6b). This indicates that although the 5′-UTR served as a strong positive regulatory region for mesocarp-specific expression, it does not influence the level of expression in response to abiotic stresses in both tissues.
https://static-content.springer.com/image/art%3A10.1007%2Fs00425-010-1220-z/MediaObjects/425_2010_1220_Fig6_HTML.gif
Fig. 6

Transient expression of the MSP1 promoter with the 5′-UTR (D1) or without the 5′-UTR (D1′) in oil palm mesocarp (a) and leaf (b) tissues. The activity of each deletion is reported as the ratio of GUS (pmol MU min−1 mg protein−1) to GFP (μg mL−1). Data are expressed as mean of three separate bombardments (each replicated five times) with the standard error. Bars with the same letter are not significantly different at P = 0.05 according to DNMRT

Discussion

The expression pattern of the oil palm MT3-A gene has been previously studied at different developmental stages in mesocarp tissue (Siti Nor Akmar et al. 2002). In the present study, functional analysis of the MSP1 promoter using a transient expression assay, revealed sevenfold higher activity in the mesocarp than in leaves, which indicates the strong fruit-specificity of the promoter. Furthermore, the full length promoter containing the 5′-UTR displayed 1.5-fold higher activity in mesocarp when compared to the constitutive CaMV35S promoter. Using 5′ deletion analysis in this study, two positive regulatory regions at position −983 to −619 and −420 to −256, and one negative regulatory region at position −619 to −420 were identified in the MSP1 promoter sequence. The −983 to −619 distal region of the promoter was shown to be potentially essential for the maximal promoter activity. By down-regulating the activity of the MSP1 promoter in leaves into a basal level, the negative regulatory region at position −619 to −420 enabled the promoter to maintain its mesocarp-specificity.

Transient expression of a series of successive 5′ deletions in the −619 to −420 nt region was performed to study the tissue-specificity of the MSP1 promoter. Deletion of the 21-bp sequence located at position −598 to −577, was shown to induce the expression of the reporter gene in leaves up to 3.5-fold. Presumably, this negative regulatory region is involved in switching the activity of the promoter from mesocarp-specific to constitutive-like expression. Tissue-specific expression is the result of a combinatorial synergic effect of positive and negative regulatory regions with tissue-specific transcription factors. Apparently, the mesocarp-specificity of the MSP1 promoter is not due to binding of positive cis-regulatory elements to transcription factors in fruits. EMSA demonstrated that the 21-bp fragment binds tightly with the nuclear proteins from leaves, but not mesocarp, root and kernel tissues. This finding confirms that leaf-specific transcription factor(s) interact with the 21-bp negative regulatory region and suppress the activity of the promoter, specifically in leaves, while the 21-bp region is not involved in controlling the activity of the MSP1 promoter in mesocarp, root, and kernel tissues. Previous studies showed that expression of the MT3-A gene is confined to the oil palm fruits (Siti Nor Akmar et al. 2002), and therefore, it is likely that other cis-regulatory region(s) might interact with nuclear factors in other non-fruit tissues and control the activity of the MSP1 by suppressing its activity through a complex regulatory mechanism. In order to pinpoint the core-sequence of the 21-bp region, a series of truncated and mutated versions of the 21-bp sequence were used in EMSAs in different tissues. An AGTTAGG sequence was identified as the critical protein-binding site in leaf tissues, and proved to be a novel motif involved in fruit-specific expression of the oil palm MSP1 promoter.

Studies on fruit-specific promoters have mainly focused on the identification of regulatory sequences related to fruit ripening, while only few reports have investigated regulatory sequences involved in fruit-specific expression. Yamagata et al. (2002) discovered a novel cis-regulatory enhancer element (TGTCACA) necessary for fruit-specific activity of the cucumisin gene promoter in melon (Cucumis melo L.). In a recent study, Yin et al. (2009) identified a novel negative cis-regulatory element (TCCAAAA) involved in fruit-specific expression of the ADP-glucose pyrophosphorylase promoter in watermelon, which interacts with leaf-specific nuclear factors and suppresses the expression in leaves; however, the authors did not study the protein-binding pattern of the identified motif in other tissues. These two motifs are not found in the oil palm MSP1 promoter sequence. Interestingly, the discovered novel motif in this study confers fruit-specificity of the promoter in the same manner as the TCCAAAA motif in melon, which is by suppressing the activity of the promoter in leaves. To the best of our knowledge, our finding is the first report on the identification of any cis-acting regulatory element responsible for fruit-specific expression in tree crops.

Expression of plant MT genes has been found to be induced by an array of biotic, abiotic and heavy metal ions (Lü et al. 2007). In order to understand the inducible-expression pattern of the MT3-A gene, its promoter sequence was examined for presence and function of the regulatory elements that are responsive to abiotic and heavy metal stresses. One striking feature of the MSP1 promoter is its high inducibility in response to ABA (1.6- and 1.7-fold increase in activity in mesocarp and leaves, respectively), which makes this promoter a good candidate for engineering of genes that are involved in ABA-responsive signaling pathways. Furthermore, the MSP1 promoter was shown to be induced by Cu2+ metal ion (1.4- and 1.5-fold increase in mesocarp and leaves, respectively). The MSP1 promoter contains a copper-responsive element, which resembles the GTAC sequence identified from the Chlamydomonas green alga Cyc6 gene (Quinn and Merchant 1995). A metal-responsive element similar to the core MRE sequence in animals (TGCRCNC) has been reported in the MT promoters of pea PsMTA (Fordham-Skelton et al. 1997), tomato LeMTB (Whitelaw et al. 1995), and oil palm MT3-B genes (Siti Nor Akmar et al. 2002). Further studies are needed to fully understand the metal-responsive mechanism of plant metallothioneins.

Functional analysis of the MSP1 promoter in stable transformants is a long-term process, since it takes several years to produce transgenic palm. The adopted biolistic-based transient expression assay in this study enabled rapid analysis of the promoter activity in different tissues and in response to treatments by inducing agents. Biolistic transfer of DNA into intact plant tissues has been proven to be a rapid approach to test the activity of the promoters in their homologous environments (Rasmussen and Donaldson 2006; Hwang et al. 2008), although there are reports indicating that promoters could act differently when they are expressed in heterologous model systems (Lü et al. 2007). There is a growing number of evidence showing that transient expression systems ultimately provide the same information as approaches using transgenic plants (Hwang et al. 2008). However, a major setback in biolistic-based reporter assays is the variability in the individual bombardment efficiencies of independent experiments, which makes it difficult to generate reproducible results. In order to overcome this difficulty, many researchers have adopted the strategy of simultaneous transformation of tissues with two different reporter constructs (Rasmussen and Donaldson 2006). In the same way, in this study, target tissues were co-bombarded with the MSP1 promoter, or its 5′ deletions, to drive the GUS reporter gene expression, and with the constitutive CaMV35S promoter to direct the GFP reporter gene expression. This provided a reliable way of taking into consideration the effect of differences in bombardment efficiencies on the promoter activity. The promoter activity was thus quantified as a ratio of the activity of the GUS reporter test construct to that of the GFP reporter reference.

Tissue-regulated and inducible promoters like the MSP1 promoter are important to restrict gene expression in a given tissue or induced for a specific time, thereby reducing the impact on non-target tissues to avoid a metabolic drain. Several fruit-specific promoters such as E4, E8, E11, PG and 2A11 from tomato and ACC-oxidase gene promoter from apple have been studied in detail (Atkinson et al. 1998; Hu et al. 2006). Yi et al. (2006) have characterized a fruit-specific/inducible promoter from tomato Lehsp23.8 gene, which was interestingly shown to be highly responsive to exogenous ABA, Cd2+, Cu2+, and Zn2+ heavy metal stresses. The MSP1 promoter showed a high mesocarp-specificity and a strong up-regulation in response to ABA and Cu2+. These findings indicate that the MSP1 promoter is regulated through a complex transcriptional regulatory mechanism with distinct functional positive and negative regulatory regions. Plant MT genes are involved in many biological processes, including growth, development, senescence and defense. The MSP1 promoter of the oil palm MT3-A gene can be exploited as a potential tool for engineering of an array of genes that are involved in the mentioned biological pathways.

This work provided important insights into the regions that control mesocarp-specificity and stress inducibility of the MSP1 promoter, supporting the future application of this promoter in genetic engineering programmes. The specificity of the MSP1 promoter in regulating gene expression in the fruit, especially in the mesocarp would be of great value in metabolic engineering to improve the fruit fatty acid content, as well as to produce high-value novel products from existing intermediates and metabolites (Omidvar et al. 2008). Efforts on oil palm genetic engineering thus far have provided nearly all the genes encoding the key enzymes for modification of the fatty acid biosynthetic pathway. Transformation cassettes carrying these genes under the regulation of the MSP1 promoter have been developed and used for production of transgenic palms. There are also efforts to channel the metabolites to high-value food and non-food products, such as lycopene and biodegradable plastics (Masani et al. 2008). In recent years, there is a growing interest to produce biofuel from renewable sources due to environmental issues and depletion of fossil fuels. Palm oil can be used for production of biodiesel and serve as environmental friendly and renewable source of energy (Naik et al. 2010). Thus, there is a potential use of the oil palm MSP1 promoter for targeting genes to increase the oil or fibre content of the oil palm fruit for high-value production of the bio-oil and biofuel/biodiesel. Furthermore, the enhanced response of the MSP1 promoter to abiotic factors and metal toxicity makes it a useful genomic tool for the inducible regulation of the engineered genes.

Acknowledgments

We thank the Ministry of Science, Technology and Innovation of Malaysia for providing the research grant through Agriculture Biotechnology Institute.

Supplementary material

425_2010_1220_MOESM1_ESM.doc (33 kb)
Supplementary material 1 (DOC 33 kb)
425_2010_1220_MOESM2_ESM.doc (136 kb)
Supplementary material 2 (DOC 135 kb)
425_2010_1220_MOESM3_ESM.doc (34 kb)
Supplementary material 3 (DOC 34 kb)
425_2010_1220_MOESM4_ESM.doc (38 kb)
Supplementary material 4 (DOC 37 kb)

Copyright information

© Springer-Verlag 2010