Journal of Biotechnology 174 (2014) 49–56
Contents lists available at ScienceDirect
Journal of Biotechnology journal homepage: www.elsevier.com/locate/jbiotec
Strong seed-specific protein expression from the Vigna radiata storage protein 8SG˛ promoter in transgenic Arabidopsis seeds Mo-Xian Chen b,1 , Shu-Xiao Zheng b,1 , Yue-Ning Yang a,1 , Chao Xu a , Jie-Sheng Liu a , Wei-Dong Yang a , Mee-Len Chye b,∗∗ , Hong-Ye Li a,∗ a b
College of Life Science and Technology, Jinan University, Guangzhou 510632, China School of Biological Sciences, The University of Hong Kong, Pokfulam, Hong Kong, China
a r t i c l e
i n f o
Article history: Received 5 December 2013 Received in revised form 23 January 2014 Accepted 27 January 2014 Available online 3 February 2014 Keywords: Genome-walking -Glucuronidase Promoter Seed-specific Transgenic Arabidopsis Vigna radiata
a b s t r a c t Vigna radiata (mung bean) is an important crop plant and is a major protein source in developing countries. Mung bean 8S globulins constitute nearly 90% of total seed storage protein and consist of three subunits designated as 8SG␣, 8SG␣ and 8SG. The 5 -flanking sequences of 8SG˛ has been reported to confer high expression in transgenic Arabidopsis seeds. In this study, a 472-bp 5 -flanking sequence of 8SG˛ was identified by genome walking. Computational analysis subsequently revealed the presence of numerous putative seed-specific cis-elements within. The 8SG˛ promoter was then fused to the gene encoding glucuronidase (GUS) to create a reporter construct for Arabidopsis thaliana transformation. The spatial and temporal expression of 8SG˛::GUS, as investigated using GUS histochemical assays, showed GUS expression exclusively in transgenic Arabidopsis seeds. Quantitative GUS assays revealed that the 8SG˛ promoter showed 2- to 4-fold higher activity than the Cauliflower Mosaic Virus (CaMV) 35S promoter. This study has identified a seed-specific promoter of high promoter strength, which is potentially useful for directing foreign protein expression in seed bioreactors. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Plants have served as bioreactors in the production of recombinant proteins for over two decades. As a production platform in molecular farming, plants are relatively low cost and possess tremendous potential through delivery in seeds which have emerged to be more favorable than any other plant-based expression system (Lau and Sun, 2009; Ma et al., 2003). In spite of the high storage protein content in mature seeds, heterologous proteins can still accumulate in transgenic seeds (Stoger et al., 2005). Given the low metabolic activity in a dehydrated environment, these heterologous proteins remain bioactive for several years (Lau and Sun, 2009; Ma et al., 2003). Furthermore, subsequent downstream purification processes can be omitted if the recombinant protein is to be delivered orally, as would be the case for edible vaccines (Daniell et al., 2001). Several monocot and dicot plants have already been used as seed production platforms. For example, chicken avidin generated in maize was the first commercially released plant-derived protein (Hood et al., 1997). Rice, has also acted as a seed production platform for orally delivered vaccines
∗ Corresponding author. Tel.: +86 20 85228470. ∗∗ Corresponding author. Tel.: +852 22990319. E-mail addresses:
[email protected] (M.-L. Chye),
[email protected] (H.-Y. Li). 1 These authors contributed equally to the manuscript. 0168-1656/$ – see front matter © 2014 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jbiotec.2014.01.027
(Tiwari et al., 2009; Yang et al., 2008) and examples of vaccine antigens expressed in rice seeds include the Newcastle disease virus envelope fusion (F) glycoprotein (Yang et al., 2007), the Vibrio cholera toxin B subunit (Nochi et al., 2007; Oszvald et al., 2008) and the human insulin-like growth factor I (Xie et al., 2008). Amongst the factors affecting the expression of foreign proteins in seeds, the choice in promoter is deemed highly important. Currently, the categories of promoters available include the constitutive, inducible and spatial- or temporal-specific promoters. The use of constitutive promoters, such as the Cauliflower Mosaic Virus (CaMV) 35S promoter (Odell et al., 1985) and the maize ubiquitin promoter (Christensen and Quail, 1996), results in unnecessary expression in non-target organs, and hormonal- or chemical-inducible promoters require external triggers; therefore, a seed-specific promoter represents an alternative for efficient seed-targeted recombinant protein production. Although the 35S promoter is often used to drive the selectable marker besides the target gene but transcriptional silencing may occur if the same promoter is used in multitude (Peremarti et al., 2010). For example, different seed promoters have been utilized to optimize the expression of multiple proteins in Arabidopsis to engineer docosahexaenoic acid and eicosapentaenoic acid biosynthesis (Haslam et al., 2013; Lopez et al., 2013). Given this potential problem, other plant-derived promoters are preferred. Several endosperm-specific promoters from monocots, such as the rice glutelin GluA-2 promoter (Wu et al., 1998), maize zein
50
M.-X. Chen et al. / Journal of Biotechnology 174 (2014) 49–56
promoter (Russell and Fromm, 1997) and barley D hordein promoter (Horvath et al., 2000), were effective in driving strong expression in various transgenic cereals. In dicots, the most well studied seed-specific promoters originate from legumes including those of genes encoding the common bean ARCELIN5 (Goossens et al., 1999) and -phaseolin (Burow et al., 1992; Kawagoe and Murai, 1992; van der Geest and Hall, 1996), and soybean lectin (Philip et al., 1998), -conglycinin ␣ subunit (Nishizawa et al., 2003) and glycinin (Nielsen et al., 1989). Use of the common bean ARCELIN5-I promoter culminated in target protein accumulation to 36.5% in Arabidopsis seeds (De Jaeger et al., 2002), exceeding foreign protein expression achieved in monocot seeds (Lau and Sun, 2009; Ma et al., 2003). Legumes such as soybean, pea, common bean and mung bean are particularly rich in protein content, and the level of storage proteins attain about 20% of total protein in comparison to cereals seeds such as maize which has a lower range of 7 to 10% (Lau and Sun, 2009). Also, the oil-rich dicot seeds have additional advantages. For example, by using the oleosin-fusion system in safflower, recombinant proteins such as biosimilar insulin could be purified in a single step through the use of oleosin, an oil-body associated small protein (Xu et al., 2012). Mung bean (Vigna radiata L.) contains a high amount of nutrient sources including proteins, minerals and vitamins and constitutes the major protein source especially in developing countries. However, few promoters from mung bean have been reported and one such promoter of the gene encoding ACC synthase was observed to drive constitutive expression in transgenic tobacco and Arabidopsis (Cazzonelli et al., 2005). Similar to other legumes, mung bean seeds are rich in storage proteins and contain 17% to 26% protein content. Three types of mung bean seed storage proteins have been characterized and designated as the 7S, 8S and 11S globulins. (Itoh et al., 2006; Mendoza et al., 2001), The 8S globulins constitute nearly 90% of total storage protein (Itoh et al., 2006; Mendoza et al., 2001). The high abundance of 8S proteins prompted us to investigate the promoter activities of their corresponding genes in seeds. The three highly conserved subunits of the 8S globulins have been classified as 8SG␣, 8SG␣ and 8SG (Bernardo et al., 2004). Recently, the 5 -flanking region of mung bean 8SGˇ was shown to display moderate promoter strength in cotyledonary protoplasts (Yang et al., 2011), while the 8SG˛ promoter was confirmed to direct high expression in Arabidopsis cotyledonary embryos (Chen et al., 2013). Nonetheless, 8SG˛ expression was also detected in other organs albeit at lower levels, thus it was deemed not seed-specific (Chen et al., 2013). In this study, the spatial and temporal expression patterns of the mung bean 8SG˛ promoter was investigated in transgenic Arabidopsis. The 5 -flanking sequence of 8SG˛, isolated by genome walking, was fused to the gene encoding -glucuronidase (GUS) to create a reporter construct for Arabidopsis transformation. Based on our prior experience using this reporter gene (Du et al., 2013a,b; Rawat et al., 2005; Zheng et al., 2012), GUS histochemical assays revealed that the 5 -flanking sequence of 8SG˛ directed seedspecific expression in Arabidopsis and showed higher activity than the conventional 35S promoter, making it potentially useful as a molecular tool in targeting heterologous protein production in plant seed bioreactors. 2. Materials and methods 2.1. Plant materials Mung bean (Vigna radiata L.) and Arabidopsis thaliana (ecotype Columbia-0) were cultivated under 16-h light (23 ◦ C)/8-h dark (21 ◦ C) cycles in a growth chamber. Seeds of A. thaliana were surface sterilized using 20% bleach supplemented with 0.05% Tween-20 followed by three washes in distilled water (Du et al., 2013a). They were then germinated on Murashige and Skoog (MS) medium
(Murashige and Skoog, 1962) containing 50 g/mL kanamycin and chilled for 4 days at 4 ◦ C in the dark for stratification. The plates were incubated under 16-h light (23 ◦ C)/8-h dark (21 ◦ C) cycles for 2 weeks. Seedlings were then grown in pots containing soil in a growth chamber under 16-h light (23 ◦ C)/8-h dark (21 ◦ C) cycles. Arabidopsis tissues were harvested and subsequently used in histochemical stains or quantitatively assayed for fluorescent GUS activity following Chen et al. (2013). 2.2. DNA extraction and genome walking Mung bean DNA was extracted from leaves using conventional DNA extraction methods. Tissues were ground using buffer provided by the DNA Extraction Kit (Takara, Japan) according to the manufacturer’s instructions. The purity of mung bean DNA was subsequently determined using spectrophotometry and agarose gel electrophoresis. Polymerase Chain Reaction (PCR)-based genome walking was used to amplify the 472-bp 5 -flanking sequence of 8SG˛ from data on the 8SG˛ mRNA (GenBank accession EF990627). Reverse primers P1 (5 -TGGAACCACCTGTCAGAGTTGA-3 ), P2 (5 -GTGTACGATGCCGAAGGAGACA-3 ), P3 (5 -GTATTAGTGAATATTGGCGGAGTTG-3 ) and degenerate adaptor primers were supplied by the Genome Walking Kit (Takara). PCR amplification was carried out as follows: first PCR at 94 ◦ C for 1 min, 98 ◦ C for 40 s, 94 ◦ C for 30 s, 65 ◦ C for 1 min, and 72 ◦ C for 2 min for 5 cycles; 94 ◦ C for 30 s, 25 ◦ C for 3 min, and 72 ◦ C for 2 min; 94 ◦ C for 30 s, 65 ◦ C for 1 min, and 72 ◦ C for 2 min; 94 ◦ C for 30 s, 65 ◦ C for 1 min, and 72 ◦ C for 2 min for 15 cycles; 94 ◦ C for 30 s, 44 ◦ C for 1 min, 72 ◦ C for 2 min, and 72 ◦ C for 10 min. The second and third PCR reactions each consisted of 94 ◦ C for 30 s, 65 ◦ C for 1 min, and 72 ◦ C for 2 min; 94 ◦ C for 30 s, 65 ◦ C for 1 min, and 72 ◦ C for 2 min; 94 ◦ C for 30 s, 44 ◦ C for 1 min, and 72 ◦ C for 2 min for 15 cycles; and an extension at 72 ◦ C for 10 min. The resultant PCR product was cloned into vector pMD18-T (Takara) and was confirmed by DNA sequencing. The forward primer for 8SG˛-omega was P10 (5 -CCCAAGCTTTGAAGGCCCTAAGAAGGATTAAAT-3 ) and reverse primer P11 containing the omega leader sequence which is underlined (5 -CGGGATCCATGGTAATTGTAAATGTAATTGTAATGTTGTT TGTTGTTTGTTGTTGTTGGTAATTGTTGTAAAAATTGTATTAGTGAATATTGGCGGAG-3 ). 2.3. Analysis of the 8SG˛ 5 -flanking region BLAST (http://www.ncbi.nlm.nih.gov) was used to analyze the 5 -flanking sequence of 8SG˛. It was compared to the 784-bp 5 -flanking sequence of 8SG˛ (GenBank accession No. HQ214071, Chen et al., 2013) and the 661-bp 5 -flanking sequence of 8SGˇ (GenBank accession No. GU176353, Yang et al., 2011). Software PlantProm of Softberry (http://www.softberry.com) was utilized to predict the transcription start site (TSS) and the TATA box. Software PlantCARE (http://bioinformatics.psb.ugent.be/ webtools/plantcare/html/) and PLACE (http://www.dna.affrc.go.jp/ database/) were used to predict cis-elements. 2.4. Transformation of Arabidopsis with the p8SG˛::GUS construct A DNA region (472-bp) consisting of the 5 -flanking sequence of 8SG˛ was cloned in the binary vector, pBI101.3 (Clontech), to generate a p8SG˛::GUS fusion (Fig. S1) for Agrobacterium tumefaciens transformation. Construct p8SG˛ ::GUS was generated by cloning the 784-bp 5 -flanking sequence of 8SG˛ in vector pBI101.3 (Clontech) as described by Chen et al. (2013). Plasmid pBI121 (Clontech) expressing 35S::GUS was used as a positive control and plasmid pBI101.3, as the negative control.
M.-X. Chen et al. / Journal of Biotechnology 174 (2014) 49–56
Agrobacterium LBA4404 was transformed with the above constructs by electroporation (McCormac et al., 1998). These GUS fusions were introduced into A. thaliana using the “floral dip” approach (Clough and Bent, 1998). The primary transformants (T0 ) were selected on MS-medium (Murashige and Skoog, 1962) containing kanamycin (50 g/mL) and were subsequently grown to set seed. The resultant seeds were harvested and germinated on MS plates containing kanamycin (50 g/mL). Putative T1 transformants with true leaves emerging on kanamycin plates were confirmed by genotyping using PCR primers specific to the 8SG˛ promoter. PCR-confirmed transformants were transferred to soil to yield T2 plants. About three to ten independent T2 lines per construct, each with single-copy inserts that had shown a simple Mendelian 3:1 segregation ratio to kanamycin, were selected. Seeds from independent lines were harvested and T3 seedlings that were 100% kanamycin-resistant were considered as homozygous for followup experiments. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jbiotec. 2014.01.027. 2.5. Histochemical GUS assays on plant materials Histochemical GUS stains of Arabidopsis were conducted following Jefferson et al. (1987). Plant materials were harvested and incubated in GUS histochemical X-Gluc (5-bromo-4chloro-3-indolyl--d-glucuronide) standard buffer (50 mM sodium phosphate, pH 7.5, 2 mM K3 Fe(CN)6 , 2 mM K4 Fe(CN)6 , 0.1% (v/v) Triton X-100, 1 mg/mL X-Gluc). Samples were vacuuminfiltrated for 1 h in GUS staining solution followed by 2-h incubation at 37 ◦ C. Chlorophylls were subsequently removed using 70% ethanol and 20-min incubation at 65 ◦ C. Light microscopy was used to view the GUS-stained samples. Comparison of promoter activities were conducted amongst the 472-bp 5 -flanking
51
sequence of 8SG˛, the 784-bp 5 -flanking sequence of 8SG˛ and the conventional 35S promoter using transgenic Arabidopsis seeds transformed with their respective GUS constructs. 2.6. Fluorometric assays to determine GUS activity Fluorometric assays of GUS activity were carried out according to Jefferson et al. (1987). In brief, total protein was extracted using GUS extraction buffer (50 mM sodium phosphate buffer, pH 7.0, 10 mM EDTA-disodium salt, 0.1% SDS, 0.1% Triton X-100, 10 mM -mercaptoethanol). The protein supernatant (50 l) and GUS extraction buffer (150 l) containing 1 mM 4-methylumbelliferyl -d-galactopyranoside (4-MUG; Invitrogen) substrate were used in each reaction. A Beckman Coulter DTX 880 multimode detector (Fullerton, USA) with excitation at 360 nm and emission at 450 nm was used to measure fluorescence. GUS activity (nmole 4-MU min−1 mg−1 protein) was normalized to total protein concentration. The protein quantity was determined using Bradford reagent (Sigma–Aldrich, SA) and bovine serum albumin (Sigma–Aldrich) was used to construct a standard curve. Six biological replicates were performed in each experiment and at least two independent experiments were completed for each batch of results. 2.7. Detection of GUS protein by western blot analysis In western blot analysis (Jasik et al., 2011), total proteins were extracted from transgenic Arabidopsis seeds (10 mg) using extraction buffer (75 mM sodium phosphate buffer, pH 7.0, 1 mM EDTA-disodium salt, 1 mM dithiothreitol). Subsequently, total protein (15 g) was separated in a 10% SDS-PAGE gel and transferred to Hybond-C membrane (Amersham) using a Trans-Blot® cell (Bio-Rad, USA). Rabbit polyclonal primary antibodies against GUS (Abcam, UK) and the SuperSignal West Pico Chemiluminescent Substrate (Thermo, USA) were used to detect cross-reacting bands.
Fig. 1. Alignment of the 5 -flanking sequences from the genes encoding mung bean 8S globulin subunit 8SG␣, 8SG␣ and 8SG. The start codon “ATG” (underlined) is marked for each sequence. Asterisks represent conserved nucleotides. The start site of transcription in each sequence is boxed. The predicted cis-elements are denoted. ABRE: ABA-responsive element; CAAT: CAAT consensus sequence; DPBF: Dc3 promoter-binding factor binding core sequence; GARE: gibberellic acid-responsive element; MYB: core sequence for MYB transcription factor; MYC: binding site for MYC transcription factor; RY: RY repeat, cis-element in response to ABA-induction; SEF3: SEF3 motif homologous to the binding site for soybean embryo factor SEF3; and Skn-1: cis-element required for endosperm-specific expression; TATA: TATA-box.
52
M.-X. Chen et al. / Journal of Biotechnology 174 (2014) 49–56
3. Results 3.1. Cloning the 5 -flanking sequence of 8SG˛ The putative 8SG˛ 5 -flanking sequence was identified as a 0.7-kb band after three rounds of genome walking. The cloned 5 -flanking region of 8SG˛ of 472 bp was deposited into GenBank (GenBank accession No. FJ792642). No significant similarity to the 5 -flanking region of the genes encoding Phaseolus vulgaris ␣- and -phaseolin (GenBank accession Nos. X52626 and J01263, respectively) were observed in a BLAST search (Fig. S2). However, the sequence demonstrated 50% similarity to the 5 -flanking sequence of mung bean 8SG˛ (GenBank accession No. HQ214071), and 68% similarity to mung bean 8SGˇ (GenBank accession No. GU176353). When these 5 -flanking DNA sequences up to the start codon were aligned (Fig. 1), the sequences nearest the start codon showed higher conservation. Supplementary material related to this article can be found, in the online version, at http://dx.doi.org/10.1016/j.jbiotec. 2014.01.027. 3.2. Identification of putative cis-elements The 8SG˛ 5 -flanking sequence, comprised of >60% AT, in agreement that transcription factors bind AT-rich regions (Joazeiro et al., 1996). Using promoter analysis software (PLACE, PlantCARE and PlantProm), dozens of putative cis-elements were predicted
within the 8SG˛ 5 -flanking sequence (Table 1), including the TATA box (-23) and CAAT enhancer motif (Joshi, 1987; Mantovani, 1998). In addition, putative hormone-responsive elements (e.g. auxin-responsive factor binding sites and the MeJA-responsive CGTCA motif), stress-related motifs (e.g. anaerobic responsive elements, cold-inducible RAV1a binding sites, copper-responsive elements, defense-responsive TC-rich repeats and the W box), and circadian/light-regulated cis-elements (e.g. Box 4, G-Box, and TCCCmotif) were identified (Table 1). Similar to the P. vulgaris ˛- and ˇ-phaseolin promoters, the prediction software identified several cis-elements related to seed expression clustered in the 400-bp proximal region relative to the TSS (Table 1). These include hormone regulatory motifs such as the ABA-responsive elements (ABREs, −422/−416 and −190/−186), RY repeat (−168/−162), Dc3 promoter binding factor (DPBF) binding site (−105/−98), MYB binding sites (−286/−281, −85/−80 and −61/−56), MYC binding sites (−320/−315, −269/−264, −259/−254 and −210/−205) and gibberellic acid (GA)-responsive elements (GARE, −248/−241) (Fig. 1), indicating that the 8SG˛ 5 -flanking sequence may be subject to regulation during development. Besides hormone-responsive motifs, embryo- and endospermspecific cis-elements were predicted. Two consensus sequences (−163/−157, −142/−136) that can potentially bind soybean embryo factor (SEF)3 were identified adjacent to each other (Fig. 1). SEF binding sites have been identified in the 5 -flanking sequence of the gene encoding ␣ subunit -conglycinin and SEF is known to positively regulate the gene encoding -conglycinin during
Table 1 Predicted cis-elements in 8SG˛ 5 -flanking region. Cis-element
Source organism of cis-element
Position on 8SG˛
Consensus sequence
Biological function in some organisms
ABRE ANAERO1 consensus
−422, −190 −323, −135
CACGTG AAACAAA
Cis-acting element involved in ABA responsiveness One of 16 motifs found in promoters of 13 anaerobic genes involved in the fermentative pathway
−265 −175
TGGTTT TGTCTC
Box 4
Arabidopsis Maize; Arabidopsis; pea; barley; rice; petunia; tomato Maize Arabidopsis; soybean; rice Parsley
−416
ATTAAT
CAAT-box CGTCA-motif
Brassica rapa; barley Barley
−74 −100
CAAAT/CAAT CGTCA
CURE core
−252, −188
GTAC
DPBF core
Chlamydomonas reinhardtii Carrot; Arabidopsis
−105
ACACNNG
GARE1
Rice
−248
TAACAGA
G-box MYB core
Snapdragon Arabidopsis; petunia
−191 −286, −85, −61
CACGTA WAACCA/CNGTTR
MYC consensus
Arabidopsis
−320, −269, −259, −210
CANNTG
RAV1A
Arabidopsis
−197, −194, −70, −54
CAACA
RY repeat
Brassica napus
−168
CATGCA
SEF3 motif
Soybean
−163, −142
AACCCA
Skn-1 motif
Rice
+10
GTCAT
TATA-box TCCC-motif TC-rich repeats
Pea; tobacco; bean Spinach Tobacco
−30, −31, −32 −15 −120
TATAAAT TCTCCCT ATTTTCTTCA
W-box
Tobacco
−316
TGACY
Cis-acting regulatory element essential for anaerobic induction ARF (auxin response factor) binding site, enriched in the 5 -flanking region of genes up-regulated by both IAA and BL Part of a conserved DNA module involved in light responsiveness Common cis-acting element in promoter and enhancer regions Cis-acting regulatory element involved in MeJA-responsiveness Core of a CuRE (copper-response element); involved in oxygen-response The bZIP transcription factors DPBF-1 and -2 (Dc3 promoter-binding factor) binding core sequence; Dc3 expression is normally embryo-specific and induced by ABA Gibberellin-responsive element (GARE) found in the promoter region of the gene encoding cystein proteinase (REP-1) Cis-acting regulatory element involved in light responsiveness Binding site for all animal MYB and at least two plant MYB proteins AtMYB1 and AtMYB2 involved in regulation of genes responsive to water stress Binding site of ICE1 (inducer of CBF expression 1) that regulates the transcription of CBF/DREB1 genes in the cold Binding consensus sequence of Arabidopsis transcription factor, RAV1 Required for seed-specific expression; ABRE-mediated transactivation by ABI3 and ABI3-dependent response to ABA Soybean embryo factor consensus sequence in the 5 -upstream region of the gene encoding -conglycinin (7S globulin) Cis-acting regulatory element required for endosperm expression Core promoter element around -30 of transcription start site Part of a light responsive element Cis-acting element involved in defense and stress responsiveness May be involved in activation of ERF3 gene by wounding
ARE AuxRE
Note: Biological functions were annotated according to software PlantCARE (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (http://www.dna.affrc.go.jp/database/). Abbreviations: ABA, abscisic acid; IAA, indole-3-acetic acid; BL, brassinolide; MeJA, methyl jasmonate.
and
PLACE
M.-X. Chen et al. / Journal of Biotechnology 174 (2014) 49–56
53
Fig. 2. Expression of 8SG˛::GUS in transgenic Arabidopsis in GUS histochemical stains. (A) Embryo in mid-torpedo stage, (B) embryo in bent-cotyledonary stage, (C) mature embryo, (D) seed coat of mature seed, (E) root tip, (F) 7-day-old seedling, (G) 4-week-old rosette leaf, (H) stems, (I) flower, and (J) siliques. Scale bar = 20 m (A–E), 1 cm (F and G), and 1 mm (H–J).
embryogenesis (Allen et al., 1989). The Skn-1 motif (GTCAT) has been reported to confer endosperm-specific expression of the gene encoding rice GluB-1 glutelin (Washida et al., 1999) and one such motif (+10/+15) was noted at the 5 -flanking region of 8SG˛ (Fig. 1). The putative CAAT box (−74/−70) was also identified (Fig. 1). 3.3. Expression of 8SG˛::GUS in transgenic Arabidopsis The expression pattern of three independent T3 lines showed that 8SG˛::GUS was expressed exclusively in seeds (Fig. 2A–D). There was no expression in root tips (Fig. 2E), young seedlings (Fig. 2F), mature rosette leaves (Fig. 2G), stems (Fig. 2H), flower parts (Fig. 2I) or siliques at various developmental stages (Fig. 2J). In developing seeds at mid-torpedo stage, 8SG˛::GUS stains initially appeared in the central zone of the developing embryo, including the cells destined to form the shoot meristem during seedling development (Laux et al., 2004). In the bent-cotyledonary stage, the expression of 8SG˛::GUS was most intense at the central zone and its expression spread bidirectionally, upwards to the cotyledons and downwards to the radicle (Fig. 2B). When the embryo matured, both the endosperm layer in the testa and the whole embryo tissue showed high expression of 8SG˛::GUS (Fig. 2C and D). 3.4. Comparison in promoter strength amongst 35S::GUS, 8SG˛::GUS and 8SG˛ ::GUS in Arabidopsis seeds When mature seeds from Arabidopsis transformants of 8SG˛::GUS were separated into the embryo and seed coat followed by GUS staining, embryos from three independent lines (8SG˛::GUS-1, 8SG˛::GUS-2, 8SG˛::GUS-3) appeared to show a more
intense blue color than 8SG˛ ::GUS and 35S::GUS embryos (Fig. 3A). Expression of 8SG˛::GUS was also observed at the endosperm layer whereas no such activity was detected for 8SG˛ ::GUS and 35S::GUS. In quantitative GUS enzyme activity measurements of whole seeds to compare the relative activity per unit time, expression of 8SG˛::GUS in all three independent lines was significantly higher (2- to 4-fold) than 35S::GUS and 8SG˛ ::GUS (Fig. 3B). Furthermore, on western blot analysis using antibodies against GUS, more GUS protein accumulated in the seeds of all three 8SG˛::GUS lines than those from 35S::GUS and 8SG˛ ::GUS (Fig. 3C), indicating that the 5 -flanking sequence of 8SG˛ displayed the strongest promoter activity. The amount of protein was observed to be substantially higher than the conventional 35S promoter (Fig. 3C). 4. Discussion Although only limited studies have reported the use of dicot plant seeds, they are promising hosts for efficient and high expression as demonstrated herein. Two of three 8S globulin (8SG˛ and 8SGˇ) 5 -flanking sequences, confer expression comparable to the conventional 35S promoter (Chen et al., 2013; Yang et al., 2011) but were not stronger than 8SG˛. According to nucleotide sequence alignment, the 8SG˛ 5 -flanking sequence did not show high similarity to either 8SG˛ and 8SGˇ or to those from seed storage proteins (Figs. 1 and S2). The divergence in upstream sequences amongst the three mung bean 8S subunits may result in their differential expression in planta as proven in promoter-GUS fusions. Transgenic Arabidopsis harboring the 8SG˛::GUS construct was exclusively expressed in seeds (Fig. 2), and showed higher GUS activity than either 35S::GUS or 8SG˛ ::GUS (Fig. 3), indicating that
54
M.-X. Chen et al. / Journal of Biotechnology 174 (2014) 49–56
Fig. 3. GUS histochemical stains and activity measurements of mature Arabidopsis seeds. (A) Freshly-harvested seeds of Arabidopsis transformants (35S::GUS, 8SG˛::GUS-1, 8SG˛::GUS-2, 8SG˛::GUS-3 and 8SG˛ ::GUS) were dissected to yield the embryo (upper panel) and seed coat (lower panel) followed by GUS staining for 2 h. Scale bar = 20 m. (B) Mature seeds (20 mg) harboring the vector control (VC), 35S::GUS, 8SG˛::GUS-1, 8SG˛::GUS-2, 8SG˛::GUS-3 and 8SG˛ ::GUS were used in fluorometric GUS activity assays. The bars in the histogram are mean (±SD) of three replicates. The student-T test was performed and the asterisks (**) represent values significantly higher than 35S::GUS (P < 0.01). (C) Western blot using GUS-specific antibodies. The 68-kD cross-reacting band is arrowed. Total protein (15 g/lane) was extracted from Arabidopsis seeds (10 mg) and was stained with Coomassie Blue as loading control (bottom).
the 8SG˛ 5 -flanking sequence is an ideal candidate for seed-specific expression of heterologous proteins. Several factors that influence promoter strength include the accessibility of the entire locus in chromatins (Li et al., 2001), the location of the promoter sequences in chromatin and the existence of matrix attachment regions (van der Geest and Hall, 1997). Cis-elements located in the regulatory promoter also control gene expression and some ABA- and GA-related elements were identified in the 5 -flanking sequence of 8SG˛ (Table 1). The plant hormones, ABA and GA, are key regulatory factors in seed desiccation and grain filling (Finkelstein et al., 2002; Yamaguchi, 2008). Several members of the Group A bZIP transcription factors in Arabidopsis designated as AREBs/ABFs have been shown to bind to ABREs (CACGTG) through yeast one-hybrid assays (Choi et al., 2000; Uno et al., 2000) and function redundantly in ABA signal transduction. Another member of the Group A bZIP protein, ABI5, has been demonstrated to play a positive role in seed maturation (Finkelstein et al., 2002; Yamaguchi, 2008). Furthermore, two bZIP transcription factors (DPBF1 and DPBF2) from sunflower bind to an
ABA-responsive, embryo-specific element (ACACNNG) in the carrot Dc3 promoter (Kim et al., 1997). In addition to ABREs, the RY repeat serves as a binding site for B3 domain-containing transcription factors during seed development (Ezcurra et al., 2000). Members of the AFL (ABI3/FUS3/LEC2) B3 proteins such as ABI3, FUS3 and LEC2, have been reported to promote embryogenesis (Suzuki and McCarty, 2008). Also, ABI3 is able to attenuate the expression of ABA-targeted genes by interacting with AREBs/ABFs and 14-3-3 proteins (Himmelbach et al., 2003). Furthermore, MYB and MYC proteins have been observed to act as transcription activators in ABA-signaling (Abe et al., 2003). It remains to be determined whether the seven putative MYB or MYC binding sites (Table 1) clustered within the 5 -flanking sequence of 8SG˛ attribute to the high expression of 8SG˛::GUS in transgenic Arabidopsis seeds. GA has been found to facilitate the repression of the AFL B3 regulatory network by the VAL (VP1/ABI3-LIKE) network (Suzuki and McCarty, 2008). A putative RY-repeat and a putative GARE motif were identified in the 8SG˛ 5 -flanking sequence, indicating the possibility in convergence of both ABA and GA signaling networks during seed development. The CAAT box is usually located in the −80/−100 region of TATAcontaining promoters (Joshi, 1987; Mantovani, 1998). The CAAT box binding protein, designated as NF-Y, is a heterotrimer consisting of three subunits (NF-YA, NF-YB and NF-YC) (Yang et al., 2005). Two NF-YB proteins, LEAFY COTYLEDON1 (LEC1) and LEC1-LIKE (L1L) have been observed to be important in seed development and depletion of any results in embryo defects in Arabidopsis (Kwong et al., 2003; Lee et al., 2003). Comparison in promoter strength of monocot seed-specific promoters have been reported (Furtado et al., 2008; Qu and Takaiwa, 2004), but no such studies have been conducted in dicots. Although the modified 5 -flanking region of the gene encoding common bean ARCELIN5 directs recombinant protein accumulation to 36.5% of total seed protein, the other dicot promoters show lower yield,