© 2013 Scandinavian Plant Physiology Society, ISSN 0031-9317
Physiologia Plantarum 150: 493–504. 2014
Cloning and functional analysis of the promoters that upregulate carotenogenic gene expression during flower development in Gentiana lutea Changfu Zhua,b,† , Qingjie Yangc,d,† , Xiuzhen Nia , Chao Baib , Yanmin Shenga , Lianxuan Shid , Teresa Capellb , Gerhard Sandmanne and Paul Christoub,f,∗ a
School of Life Sciences, Changchun Normal University, Changchun 130032, China ` Departament de Produccio´ Vegetal i Ciencia Forestal, Universitat de Lleida-Agrotecnio Center, Lleida 25198, Spain c College of Landscape Architecture, Northeast Forestry University, Harbin 150040, China d School of Life Sciences, Northeast Normal University, Changchun, 130024 China e Department of Molecular Biosciences, J.W Goethe Universitaet, Frankfurt, D-60054 Germany f ´ Companys, Barcelona 08010, Spain Institucio Catalana de Recerca i Estudis Avancats, Passeig Lluis b
Correspondence *Corresponding author, e-mail: [email protected] Received 17 June 2013; revised 24 October 2013 doi:10.1111/ppl.12129
Over the last two decades, many carotenogenic genes have been cloned and used to generate metabolically engineered plants producing higher levels of carotenoids. However, comparatively little is known about the regulation of endogenous carotenogenic genes in higher plants, and this restricts our ability to predict how engineered plants will perform in terms of carotenoid content and composition. During petal development in the Great Yellow Gentian (Gentiana lutea), carotenoid accumulation, the formation of chromoplasts and the upregulation of several carotenogenic genes are temporally coordinated. We investigated the regulatory mechanisms responsible for this coordinated expression by isolating five G. lutea carotenogenic gene (GlPDS , GlZDS , GlLYCB, GlBCH and GlLYCE ) promoters by inverse polymerase chain reaction (PCR). Each promoter was sufficient for developmentally regulated expression of the gusA reporter gene following transient expression in tomato (Solanum lycopersicum cv. Micro-Tom). Interestingly, the GlLYCB and GlBCH promoters drove high levels of gusA expression in chromoplastcontaining mature green fruits, but low levels in chloroplast-containing immature green fruits, indicating a strict correlation between promoter activity, tomato fruit development and chromoplast differentiation. As well as core promoter elements such as TATA and CAAT boxes, all five promoters together with previously characterized GlZEP promoter contained three common cis-regulatory motifs involved in the response to methyl jasmonate (CGTCA) and ethylene (ATCTA), and required for endosperm expression (Skn1_motif, GTCAT). These shared common cis-acting elements may represent binding sites for transcription factors responsible for co-regulation. Our data provide insight into the regulatory basis of the coordinated upregulation of carotenogenic gene expression during flower development in G. lutea.
Abbreviations – ABRE, abscisic acid responsive element; BCH, β-carotene hydroxylase; ERE, ethylene response element; GlBCH, Gentiana lutea β-carotene hydroxylase; GlLYCB, Gentiana lutea lycopene β-cyclase; GlPDS, Gentiana lutea phytoene desaturase; GlLYCE, Gentiana lutea lycopene -cyclase gene; GlZDS, Gentiana lutea ζ -carotene desaturase gene; GlZEP, Gentiana lutea zeaxanthin epoxidase; IMG, immature green; LA-PCR, long accurate polymerase chain reaction; LYCB, lycopene β-cyclase; MeJA, methyl jasmonate; MG, mature green; PCR, polymerase chain reaction; PDS, phytoene desaturase; PSY, phytoene synthase; UTR, untranslated region; ZDS, ζ -carotene desaturase; ZEP, zeaxanthin epoxidase. † These
authors contributed equally to this work.
Physiol. Plant. 150, 2014
493
Introduction The carotenoids are a large class of yellow, orange and red pigments derived from isoprenoid precursors. In higher plants, they accumulate in the chloroplasts of leaves and in the chromoplasts of many flowers and fruits. Carotenoids function as accessory pigments during photosynthesis and help to prevent photooxidation (Frank and Cogdell 1996, Demmig-Adams and Adams 2002). They are also precursors of the plant hormones abscisic acid (Creelman and Zeevart 1984) and strigolactone (Gomez-Roldan et al. 2008, Umehara et al. 2008). In flowers and fruits, the bright colors imparted by carotenoids help to attract pollinating insects and seed-dispersing animals (Bartley and Scolnik 1995), and the colors also provide agronomic value in fruit and vegetable crops as well as ornamental plants (Bartley and Scolnik 1995, Zhu et al. 2010). Carotenoids play a fundamental role in human nutrition as antioxidants and precursors of vitamin A, and a high dietary intake of carotenoids reduces the risk of several diseases (Fraser and Bramley 2004, Zhu et al. 2009, Bai et al. 2011, Farre et al. 2011). Genes encoding carotenogenic enzymes have therefore been cloned and used for metabolic engineering to increase the nutritional value and health-promoting properties of plants by improving the carotenoid content and composition (Sandmann et al. 2006, Zhu et al. 2007, 2013, Giuliano et al. 2008, Farre et al. 2010, 2011, Bai et al. 2011). However, the regulation of endogenous carotenoid formation in higher plants is poorly understood, restricting the extent to which the impact of metabolic engineering in crops can be predicted (Sandmann et al. 2006, Giuliano et al. 2008, Farre et al. 2010). The flowers in the Great Yellow Gentian (Gentiana lutea) contain abundant amounts of β-carotene and xanthophylls, making it a useful model to investigate the regulation of carotenoid biosynthesis in flowers (Zhu et al. 2002, 2003). The petals possess chromoplasts that originate from fully developed chloroplasts, and there is a temporal correlation among the accumulation of carotenoids, the formation of chromoplasts and the induction of the carotenogenic genes PSY (phytoene synthase), PDS (phytoene desaturase), ZDS (ζ -carotene desaturase), LYCB (lycopene β-cyclase), BCH (β-carotene hydroxylase) and ZEP (zeaxanthin epoxidase) (Zhu et al. 2002, 2003) (Fig. 1). Transcription of carotenogenic genes in G. lutea controls the accumulation of carotenoids during flower development (Zhu et al. 2002, 2003). The G. lutea zeaxanthin epoxidase (GlZEP ) promoter was recently cloned and characterized in transgenic tomato plants (Yang et al. 494
2012), where it was shown to drive the developmentally regulated expression of gusA, encoding the reporter β-glucuronidase (GUS). High levels of gusA expression were observed in chromoplast-containing fruits and petals. In contrast, only low levels of expression was seen in the immature green (IMG) chloroplast-containing fruits and leaves. GlZEP promoter activity was strictly associated with fruit development and chromoplast differentiation (Yang et al. 2012). Here we describe the isolation and analysis of five G. lutea carotenogenic gene promoters (Gentiana lutea phytoene desaturase gene (GlPDS ), Gentiana lutea ζ -carotene desaturase gene (GlZDS ), Gentiana lutea lycopene β-cyclase gene (GlLYCB), Gentiana lutea β-carotene hydroxylase (GlBCH ) and Gentiana lutea lycopene cyclase gene (GlLYCE ) promoters, the latter controlling the expression of lycopene -cyclase gene). We investigated the regulatory basis of coordinated upregulation during G. lutea flower development and searched for common cis-regulatory elements that could provide insight into the hierarchal control of carotenoid biosynthesis in flower petals.
Materials and methods Plant material
Gentiana lutea leaves were obtained from the Hokkaido Experimental Institute of Health Science, Japan. The tissues were frozen in liquid nitrogen immediately after harvesting and stored at −80◦ C. Tomato (Solanum lycopersicum cv. Micro-Tom) plants were grown in the greenhouse at 25◦ C with a 16-h photoperiod. Cloning the promoter sequences
Gentiana lutea Genomic DNA was extracted from 5 g of leaf tissue according to Edwards et al. (1991). Genomic DNA (20 μg) was completely digested with a restriction enzyme appropriate to the corresponding promoter sequence (Table 1) and self-ligated using 10 Weiss units of T4 DNA Ligase (Invitrogen, Carlsbad, CA) to generate circular molecules. These were used as templates for amplification by the long accurate polymerase chain reaction (LA-PCR), following the recommendations provided with the Takara LA-PCR Kit (Takara, Shuzo, Japan). The products were cloned into vector PCR® II TOPO® (TA Cloning Kit, Invitrogen, Carlsbad, CA) for sequencing using the Big Dye Terminator v3.1 Cycle Sequencing Kit on a 3130x1 Genetic Analyzer (Applied Biosystems, Foster City, CA). The restriction enzymes and primer sequences required for each of the five promoters are listed in Table 1.
Physiol. Plant. 150, 2014
Fig. 1. Carotenoid biosynthesis pathway in plants. Abbreviations: CRTISO, carotenoid isomerase; CYP97C, heme-containing cytochrome P450 carotene -ring hydroxylase; GGPP, geranylgeranyl diphosphate; HYDB, β-carotene hydroxylase [non-heme di-iron β-carotene hydroxylase (BCH) and heme-containing cytochrome P450 β-ring hydroxylases (CYP97A and CYP97B)]; LYCB, lycopene β-cyclase; LYCE, lycopene -cyclase; PDS, phytoene desaturase; PSY, phytoene synthase; VDE, violaxanthin de-epoxidase; ZDS, ζ -carotene desaturase; ZEP, zeaxanthin epoxidase; Z-ISO, ζ -carotene isomerase. This figure was modified based on Zhu et al. (2010, 2013).
Physiol. Plant. 150, 2014
495
Table 1. PCR primers used to clone gene promoters, to construct promoter-gusA fusion plasmids and to clone genomic DNA to confirm the isolated promoter sequences.
Gene PDS
Promoter cloning (digested with EcoRI) Promoter-GUS construct
Promoter cloning (digested with BamHI) Promoter-GUS construct Genomic DNA cloning
LYCB
Promoter cloning (digested with EcoRI) Promoter-GUS construct Genomic DNA cloning
BCH
Promoter cloning (digested with PstI) Promoter-GUS construct Genomic DNA cloning
LYCE
Accession numbers for cDNA or genomic DNA (gDNA)
F1: 5 -GGACACATATCTGCTGTTAACATAGGTAGGCAAGG-3 R1: 5 -GGTAGGATAAAATTCACTAAGTTGAAGGTGAAAGGG-3 F2: 5 -GTCGACAATTCATGAGTTCAAACCCGTGATTCGTTC-3 R2: 5 -GGATCCATATCAAAGCTGGTACCAAACAGAGCAAAC-3 F3: 5 -AATTCATGAGTTCAAACCCGTGATTCGTTC-3 R3: 5 -GGTAGGATAAAATTCACTAAGTTGAAGGTG-3 F1: 5 -AGGCTTGTTTCCACCGGAACCTGAACATTATCGG-3
+145 to +179 +1 to +36 −1077 to −1048 −1 to −30 −1077 to −1048 +7 to +36 +415 to +448
EF203257 (cDNA)
Purposes
Genomic DNA cloning ZDS
Primers
Positions (+1 is the first nucleotide of cDNA)
Promoter cloning (digested with BglII) Promoter-GUS construct Genomic DNA cloning
R1: 5 -TCAATGCAATCCTGAAGAACTTAGACCATTGCTGT-3 F2: 5 -GTCGACGATCCGTTTAACGCTTAGATCGTCGTCATC-3 R2: 5 -GGATCCTTAGATTATGTTAAAACAAGCATCAAAGCT-3 F3: 5 -GATCCGTTTAACGCTTAGATCGTCGTCATC-3 R3: 5 -TCAACGCAATCCTGAAGAACTTAGACCATT-3 F1: 5 -CACCCTTTATGTGGGTTTGTTGATAAAGTCTGCTCC-3
+123 to +157 −766 to −737 −1 to −30 −766 to −737 +128 to +157 +385 to +420
R1: 5 -CCAATTGTTGCCCAGAAAGCAAACTGAATTTCGAG-3 F2: 5 -GTCGACAATTCAACATCAAATGGCTAGTTGGACTTT-3 R2: 5 -GGATCCTCTGCGACACGTCTGACGGTGGAGTTAATT-3 F3: 5 -AATTCAACATCAAATGGCTAGTTGGACTTT-3 R3: 5 -CTTTTCTCTATTGGGTGTACTTTATTTCTT-3 F1: 5 -TTGGTCTCCGGTAGAAACAGCAACATTCATTGCCGT-3
+259 to +293 −1506 to −1477 −1 to −30 −1506 to −1477 +313 to +342 +135 to +170
R1: 5 -GTGACCGGAAACCGGAGAAACAGCTGAGAGGAGAG-3 F2: 5 -GTCGACGCCTAGGCCTACAATGCACAACCTAATACC-3 R2: 5 -GGATCCACCACATGTTTTCTTTGGATTGTGAAACTC-3 F3: 5 -GCCTAGGCCTACAATGCACAACCTAATACC-3 R3: 5 -TACCGGAAAATGGTGACCGGAAACCGGAGA-3 F1: 5 -TATTCGAGACGATCAAGAAGAAGGAGGATTCTCAGTG-3
+73 to +107 −1744 to −1715 −1 to −30 −1744 to −1715 +90 to +119 +310 to +346
R1: 5 -GTTGTTTGAAGTGTGAGCGCAGGGCAGCCGCTAAAG-3 F2: 5 -GTCGACGATCTTAAGTTATGTTCTAGTAAGTAAAGC-3 R2: 5 -GGATCCTCGTTTCAAGAGTTGCCACGTGGCTTGAAA-3 F3: 5 -GATCTTAAGTTATGTTCTAGTAAGTAAAGC-3 R3: 5 -GTTTGGCCCTTTATTTGTGTAGTTTCCGTG-3
+31 to +66 −938 to −909 −1 to −30 −938 to −909 +226 to +255
Promoter-gusA constructs The promoter sequences were fused to the gusA gene in vector pBI101 (Clontech Laboratories, Mountain View, CA) (Jefferson et al. 1987) by amplifying each promoter using primers containing additional sequences to provide appropriate restriction sites. For example, the full-length GlPDS promoter region was amplified from G. lutea genomic DNA using forward primer F2 (5 -GTC GAC AAT TCA TGA GTT CAA ACC CGT GAT TCG TTC-3 , introducing a Sal I site in italics) and reverse primer R2 (5 -GGA TCC ATA TCA AAG CTG GTA CCA AAC AGA GCA AAC-3 , introducing a BamHI site in italics). The 1077-bp amplified promoter fragment was then transferred to the PCR® II TOPO® vector using the Invitrogen TA Cloning® Kit, to yield intermediate vector pCR-GlPDSPro. The pCR-GlPDSPro and pBI101 vectors were digested with Sal I and BamHI, allowing 496
DQ226992 (gDNA) DQ226992 (gDNA) EF203258 (cDNA)
JQ417647 (gDNA) JQ417647 (gDNA) EF203253 (cDNA)
JQ417648 (gDNA) JQ417648 (gDNA) EF203255 (cDNA)
EF203261 (gDNA) EF203261 (gDNA) EF203256 (cDNA)
EU592045 (gDNA) EU592045 (gDNA)
the GlPDSPro fragment to be inserted upstream of gusA, yielding the final construct pBI-GlPDSPro-GUS (PDSgusA). The constructs pBI-GlZDSPro-GUS (ZDS-gusA), pBI-GlLYCBPro-GUS (LYCB-gusA), pBI-GlBCHPro-GUS (BCH-gusA) and pBI-GlLYCEPro-GUS (LYCE-gusA) were constructed in an analogous manner using the specific primers listed in Table 1. The integrity of all intermediate and final constructs was confirmed by sequencing.
Transient expression of promoter-gusA constructs in tomato Plasmids pBI101, pBI121 (35S-gusA), pBI-GlPDSProGUS (PDS-gusA), pBI-GlZDSPro-GUS (ZDS-gusA), pBI-GlLYCBPro-GUS (LYCB-gusA), pBI-GlBCHPro-GUS (BCH-gusA) and pBI-GlLYCEPro-GUS (LYCE-gusA) were introduced into Agrobacterium tumefaciens
Physiol. Plant. 150, 2014
strain LBA 4404 by electroporation (Mattanovich et al. 1989). Individual colonies were seeded into 5-ml aliquots of YEM medium (0.5% beef extract, 0.1% yeast extract, 0.5% peptone, 0.5% sucrose, 2 mM MgSO4 , pH 7.2) containing 50 μg ml−1 kanamycin and 25 μg ml−1 rifampicin, and were shaken at 300 rpm, 28◦ C overnight. Each culture was then used to inoculate 50 ml of induction medium (YEM medium supplemented with 20 μM acetosyringone, 10 mM MES, pH 5.6) containing 50 μg ml−1 kanamycin and 25 μg ml−1 rifampicin, followed by incubation as above. Bacteria were recovered by centrifugation (2700 g ), resuspended in infiltration medium (10 mM MgCl2 , 10 mM MES, 200 μM acetosyringone, pH 5.6) to an OD600 of approximately 1.0, and then incubated at room temperature with gentle agitation (20 rpm) for 3 h. Approximately 450 μl of the infiltration medium was then injected into fruits at the immature green stage (fruit diameter 0.8–1.0 cm, about half full size, 15–20 days after anthesis. Fruits were not fully expanded and still green, defined as ‘immature green’) (Akihiro et al. 2008), and 600 μl was injected into fruits at the mature green (MG) stage (fruit diameter 1.5–2.0 cm, 30–35 days after anthesis. Fruits were fully expanded and green, defined as ‘mature green’) (Akihiro et al. 2008), in both cases through the stylar apex using a 1-ml syringe with needle (Orzaez et al. 2006). Injected fruits were left on the vine for 3 days before harvesting and sectioning for histochemical staining. All the experiments were repeated six times in six independent Micro-Tom plants. Histochemical GUS assay Histochemical GUS assays were carried out according to Jefferson et al. (1987) with minor modifications. Tissues were incubated at 37◦ C for 12 h in the dark in 1 mM XGluc (5-bromo-4-chloro-3-indolyl-β-D-glucuronide) in 100 mM sodium phosphate (pH 7.0), 10 mM EDTA (ethylenediaminetetraacetic acid), 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide, 0.3% (v/v) Triton X-100 and 20% (v/v) methanol to eliminate endogenous GUS expression (Kosugi et al. 1990). After staining, the tissues were destained in an ethanol series (50, 70, 80 and 95%) to remove chlorophyll, and then stored in 70% (v/v) ethanol, and photographed with a digital camera.
Results Cloning the promoter sequences The promoters for GlPDS , GlZDS , GlLYCB, GlBCH and GlLYCE were cloned by inverse PCR using cleaved and
Physiol. Plant. 150, 2014
circularized G. lutea genomic DNA as the template and outward-facing primers based on the corresponding cDNA sequences (Zhu et al. 2002, 2003). The primers, GenBank accession numbers and restriction enzymes used to prepare the genomic templates are listed in Table 1. After sequencing the products, DNA fragments of 1113 bp (GlPDS ), 923 bp (GlZDS ), 1848 bp (GlLYCB), 1863 bp (GlBCH ) and 1193 bp (GlLYCE ) were isolated directly from genomic DNA using gene-specific primers (Primers and GenBank accession numbers are listed in Table 1). All the fragments comprised the upstream promoters and the 5 -untranslated region (UTR). The full-length promoter fragments were defined as 1077 bp (GlPDS ), 766 bp (GlZDS ), 1506 bp (GlLYCB), 1744 bp (GlBCH ) and 938 bp (GlLYCE ) with position +1 assigned to the first nucleotide of the cDNAs (Zhu et al. 2002, 2003). Construction of promoter-gusA fusion genes and transient expression in tomato fruits The full-length promoter regions were amplified from G. lutea genomic DNA using forward primers containing a Sal I restriction site (5 -GTCGAC-3 ) and reverse primers containing a BamHI restriction site (5 -GGATCC-3 ). The promoter fragments were inserted upstream of the gusA gene in vector pBI101, which had been digested with Sal I and BamHI. Tomato fruits at the IMG and MG stages were injected with bacterial cultures carrying the vectors pBI101 (promoterless gusA), pBI121 (35S-gusA, constitutive Cauliflower mosaic virus 35S promoter), pBI-GlPDSPro-GUS (PDS-gusA), pBI-GlZDSPro-GUS (ZDS-gusA), pBI-GlLYCBPro-GUS (LYCB-gusA), pBIGlBCHPro-GUS (BCH-gusA) and pBI-GlLYCEPro-GUS (LYCE-gusA). Fruits were harvested 3 days later and transverse sections were stained for GUS expression. As expected, both IMG fruits (Fig. 2) and MG fruits (Fig. 3) showed significant GUS expression when transformed with the 35S-gusA construct but no GUS expression when transformed with the promoterless control vector pBI101. We evaluated the five carotenoid constructs (PDS-gusA, ZDS-gusA, LYCB-gusA, BCH-gusA and LYCE-gusA) by transient expression in tomato fruits as above, using the 35S-gusA vector as a control. Histochemical staining revealed no GUS expression in IMG fruits transformed with PDS-gusA, ZDS-gusA and LYCEgusA, and with only low GUS expression levels in IMG fruits transformed with the LYCB-gusA and BCH-gusA constructs (Fig. 2) but higher GUS expression for these constructs, as well as PDS-gusA and LYCE-gusA, in MG fruits (Fig. 3). The ZDS-gusA construct produced low level GUS expression in the mesocarp of MG fruits 497
Fig. 2. Histochemical GUS staining of typical transgenic MicroTom IMG fruit expressing pBI101, 35S-gusA, PDS-gusA, ZDS-gusA, LYCB-gusA, BCH-gusA, LYCE-gusA, respectively. Abbreviations: pBI101, pBI101; 35S-gusA, pBI121; PDS-gusA, pBI-GlPDSPro-GUS; ZDS-gusA, pBI-GlZDSPro-GUS; LYCB-gusA, pBI-GlLYCBPro-GUS; BCH-gusA, pBIGlBCHPro-GUS (BCH-gusA); LYCE-gusA, pBI-GlLYCEPro-GUS.
(Fig. 3). Overall, the GUS expression of the five constructs during fruit development was similar, indicating that all cis-acting elements necessary to confer GUS expression in tomato fruits are contained within the isolated promoter sequences. Promoter analysis A search of the PlantCARE database of plant cis-regulatory elements (Lescot et al. 2002, http://bioinformatics. psb.ugent.be/webtools/plantcare/html) revealed potential TATA and CAAT boxes in all five promoters (Table 2; Figs S1–S5, Supporting Information). In addition to these core promoter elements, we found that all five promoters contained three common cis-regulatory motifs: CGTCAmotif which is linked to methyl jasmonate (MeJA) signaling (Rouster et al. 1997), and ATCTA-motif which was recently found in the AtPSY promoter and is known to interact with the ethylene response transcription factor RAP2.2 (Welsch et al. 2003, 2007). Another motif present was Skn-1_motif (GTCAT) that is required for endosperm expression (Takaiwa et al. 1991). Carotenoid biosynthesis is known to be regulated by light (Von Lintig et al. 1997, Simkin et al. 2003, Li et al. 498
Fig. 3. Histochemical GUS staining of typical transgenic MicroTom MG fruit expressing pBI101, 35S-gusA, PDS-gusA, ZDS-gusA, LYCB-gusA, BCH-gusA, LYCE-gusA, respectively. Abbreviations: pBI101, pBI101; 35S-gusA, pBI121; PDS-gusA, pBI-GlPDSPro-GUS; ZDS-gusA, pBI-GlZDSPro-GUS; LYCB-gusA, pBI-GlLYCBPro-GUS; BCH-gusA, pBIGlBCHPro-GUS (BCH-gusA); LYCE-gusA, pBI-GlLYCEPro-GUS.
2008, Welsch et al. 2008), and we identified at least one light-response motif in each promoter, e.g. Box I, Box II, ATC, GT-1, TCT, Sp1, G-box, ACE and AEbox (Table 3; Figs S1–S5) providing the basis for the regulation of carotenogenic gene expression according to day length and other cues involved in the control of flower development. At least one MYB binding site, involved in drought tolerance (Abe et al. 2003), was also found in each promoter (Table 3). We also identified a number of additional cis-acting elements shared among some of the promoters, including an ethylene response element (ERE) in the GlPDS and GlLYCE promoters, abscisic acid responsive element (ABRE) and auxin response element (TGA-element) in the GlBCH and GlLYCE promoters, and a W1-box that responds to fungal elicitors in the GlLYCB and GlLYCE promoters. The GlZDS , GlLYCB and GlLYCE promoters shared a GCN4 motif common to endosperm-specific promoters, whereas the GlLYCE and GlBCH promoters contained four TA-rich regions and an ATGCAAAT motif that binds GCN4 in rice, respectively (Table 2).
Discussion The carotenoid biosynthesis pathway (Fig. 1) has been completely elucidated through a combination of biochemical, genetic and transgenic approaches
Physiol. Plant. 150, 2014
Physiol. Plant. 150, 2014
499
TATA-box
CAAT-box
ERE
PDS −76 −624, −552 (CAAT); −287, −1036, −1003 (1077 bp) (TATATATA) −39 (CAAAT); −433, (ATTTCAAA) −168, −87 (CAATT) ZDS −47 (TAATA); −654, −626, −285 (CAAT); (766 bp) −16 (TTTA) −706, −580, −560, −501; −380, −60 (CAAAT); −415, −353, −157 (CAATT) LYCB −44 −926, −735, −685, −579, (1506 bp) (TATAAA) −517, −403, −145 (CAAT); −1496, −1361, −725, −308 (CAAAT); −871, −831, −776, −533, −381, −158, −151, −49 (CAATT) BCH −142 −1733, −1550, −1147, (1744 bp) (TAATA) −1136, −353, −22 (CAAT); −1390, −1310, −1163, −647, −483 (CAAAT); −1540, −1208, −1154, −935, −928, −264, −195, −188, −155, −137 (CAATT) LYCE −73 −834 (CAAT); −683, −599, −157 (938 bp) (TAATA) −93 (CAAAT) (ATTTCAAA)
Promoter
−1399 (AACGAC)
−908 (AACGAC)
−440, −22 (CACGTG); −21 (ACGTGGC)
TGA-element (auxin response)
−1670, −738 −1144 (CACGTG); (TGACG−70 TAA) (TACGTG)
ABRE
TGA-box (auxin response)
−696 (TTGACC)
−1123 (TTGACC)
Box-W1
−27 (CAAGCCA)
−626 (TGAGTCA)
−302 (TGT GTCA)
GCN4
−1166 (ATACAAAT)
ATGCAAAT motif
−391, −389, −387, −385 (TATATATA TATATATA TATATA)
TA-rich region
Table 2. Specific cis-acting regulatory elements found in different carotenogenic gene promoters from Gentiana lutea. The positions of the cis-acting regulatory elements are denoted relative to the positions of cDNA (+1 is the first nucleotide of cDNA). GenBank accession numbers for the promoter sequences analyzed are as follows: PDS, DQ226992; ZDS, JQ417647; LYCB, JQ417648; BCH, EF203261; ZEP, EF203262; LYCE, EU592045.
Table 3. The consensus cis-acting regulatory elements in different promoters of carotenogenic genes from Gentiana lutea. The positions of the cis-acting regulatory elements are denoted relative to the positions of cDNA (+1 is the first nucleotide of cDNA). GenBank accession numbers for the promoter sequences analyzed are as follows: PDS, DQ226992; ZDS, JQ417647; LYCB, JQ417648; BCH, EF203261; ZEP, EF203262; LYCE, EU592045.
Promoter
Cis-acting regulatory element involved in light responsiveness
ATCTA-motif (RAP2.2 motif in AtPSY)
PDS (1077 bp)
−1035, −1002 (Box I, TTTCAAA); −795, −607 (ATC-motif, AGTAATCT)
−603 (ATCTA)
ZDS (766 bp)
−420 (TCT-motif, TCTTAC); −397 (GT-1 motif, GGTTAAT); −49 (Box 4, ATTAAT) −1407 (TCT-motif, TCTTAC)
−6 (ATCTA)
LYCB (1506 bp) BCH (1744 bp)
ZEP (2225 bp)
LYCE (938 bp)
−1670, −738 (G-box, CACGTG); −1427 (G-box, CACGTT); −114 (G-box, TGACGTGG); −70 (G-box, TACGTG); −1516, −1254 (Box 4, ATTAAT); −1453 (AE-box, AGAAACTT); −1410 (AE-box, AGAAACAT); −112 (ACE, ACGTGGA); −1725, −1420 (MRE, AACCTAA); −33 (GAG-motif, AGAGAGT) −2217 (GT-1-motif, GCGGTAATT); −2160, −275 (Box I, TTTCAAA); −1613 (G-box, CACATGG); −1585, −1355 (G-box, CACGTC); −1567 (chs-CMA2a, TCACTTGA); −1068 (GAG-motif, AGAGAGT); −599, −262, −237, −215 (Box 4, ATTAAT) −614 (Sp1, GGGCGG); −531 (ATC-motif, AGCTATCCA); −464 (G-box, CACATGG); −440, −22 (G-Box, CACGTG); −156 (Box I, TTTCAAA); −23 (Box II, CCACGTGGC)
−597 (CGTCA-motif, CGTCA); −266 (TGACG-motif, TGACG) −743 (CGTCA-motif, CGTCA)
Skn-1_motif required for endosperm expression
MYB binding site involved in droughtinducibility
−596 (GTCAT)
−46 (MBS, TAACTG)
−742, −299 (GTCAT)
−546 (MBS, TAACTG)
−1352, −954 (ATCTA) −1637, −1617, −1529, −415 (ATCTA)
−18 (CGTCA-motif, CGTCA) −1144, −114 (TGACG-motif, TGACG); −817 (CGTCA-motif, CGTCA)
−623, −190 (GTCAT) −1287, −967, −816, −227, −207 (GTCAT)
−1372 (MBS, CAACTG) −1602 (MBS, TAACTG)
−2191, −1512 (ATCTA)
−2064, −1527 (CGTCA-motif, CGTCA); −950 (TGACG-motif, TGACG)
−658, −557 (GTCAT)
−1191, −997 (MBS, CAACTG)
−765, −83 (ATCTA)
−588 (TGACG-motif, TGACG); −485 (CGTCA-motif, CGTCA)
−539, −435 (GTCAT)
−524 (MBS, CAACTG)
(Zhu et al. 2007, 2010, 2013, Giuliano et al. 2008, Farre et al. 2010). During flower development in G. lutea, carotenoids accumulate in the petals, accompanied by a shift in profile from lutein derived from α-carotene to xanthophylls derived from β-carotene (Zhu et al. 2003). PSY and ZDS mRNA levels increase six- to sevenfold during this process, whereas PDS and other genes required for β-carotene formation and metabolism (LYCB, BCH and ZEP ) are induced approximately twofold (Zhu et al. 2002, 2003). In contrast, the abundance of the LYCE transcript is reduced to less than half of its normal value (Zhu et al. 2003). These 500
Cis-acting regulatory element involved in MeJAresponsiveness
antagonistic changes shift the relative abundance of carotenoids from the α to β branches. These data, combined with the developmental regulation of the GlZEP promoter during tomato fruit development and chromoplast differentiation (Yang et al. 2012), suggest that carotenogenesis in the flower petal is regulated by multiple developmental and environmental cues. It is clear that carotenogenesis in G. lutea petals during flower development is under transcriptional control (Zhu et al. 2002, 2003), but the mechanisms are largely unknown. The coordinated expression of several genes may reflect similarities between their promoters,
Physiol. Plant. 150, 2014
or potentially could reflect clustering within particular chromosomal domain (Caron et al. 2001, Winzer et al. 2012). Similar expression profiles may also be caused by the coordinated action of more than one set of transcription factors, in which case the promoter regions of co-regulated genes would be heterogeneous, e.g. the well-known C1 and R transcription factors which control the flavonoid biosynthesis pathway (Grotewold et al. 2000, and other references therein). To investigate the basis of coordinated carotenogenic gene expression during G. lutea flower development, we cloned promoter fragments upstream of GlPDS , GlZDS , GlLYCB, GlBCH and GlLYCE directly from G. lutea genomic DNA by inverse PCR. Sequencing identified putative TATA and CAAT boxes in all the promoters (Table 2). We investigated GUS expression driven by these promoters in ripening tomato fruits which provide as a useful model for studying the regulation of carotenogenic gene expression in chromoplastcontaining petals (Yang et al. 2012). We evaluated the constructs PDS-gusA, ZDS-gusA, LYCB-gusA, BCHgusA and LYCE-gusA by transient expression in tomato and found that all the promoters were sufficient for developmentally regulated reporter gene expression in the tomato fruits. The GlLYCB and GlBCH promoters sustained strong GUS expression in chromoplast-containing MG fruits (Fig. 3) but only low levels in chloroplast-containing IMG fruits (Fig. 2). This is similar to the endogenous gene expression profiles of the genes in G. lutea (Zhu et al. 2002, 2003) and the expression of the GlZEP promoter in transgenic tomatoes (Yang et al. 2012), confirming that carotenogenic gene promoter activity is strictly associated with tomato fruit development and chromoplast differentiation. The integration of expression profiles and promoter sequences can help to identify common and putative functionally relevant cis-acting elements (Werner 2001, Kim and Kim 2006, Yamamoto et al. 2011). Comparative bioinformatic studies on promoter regions of carotenoid genes may elucidate common binding motifs involved in carotenoid formation (Fraser and Bramley 2004). We therefore compared the isolated promoter sequences with the GlZEP promoter (Yang et al. 2012) using the Plant-CARE and PLACE databases (Higo et al. 1999, Lescot et al. 2002). All the promoters contained at least one CGTCA-motif involved in the response to MeJA, one ATCTA-motif that interacts with the ethylene response factor RAP2.2 in Arabidopsis thaliana (Welsch et al. 2003, 2007), one Skn-1_motif (GTCAT) that is required for endosperm expression, and consensus motifs involved in responses to light and drought (Table 3). These shared cis-acting elements could represent
Physiol. Plant. 150, 2014
binding sites for transcription factors responsible for co-regulation (Kreiman 2004, Haberer et al. 2006, Obayashi et al. 2007, Lenka et al. 2009). The existence of CGTCA and ATCTA motifs representing two hormone response pathways may give an example of functional cooperation or independent activities. The ATCTA motif in the AtPSY promoter binds RAP2.2, a member of the APETALA2/ERE-binding protein family. Overexpression of RAP2.2 induced only minor changes in the carotenoid profile of non-green Arabidopsis tissues, suggesting that additional factors may contribute to the regulation of PSY (Welsch et al. 2007). Single copy of the ATCTA motif is found in other carotenogenic gene promoters such as Arabidopsis DXS (encoding deoxyxylulose phosphate synthase) and PDS (Welsch et al. 2007), tomato and maize PDS (Welsch et al. 2007), and tomato CYC-B encoding lycopene βcyclase (Dalal et al. 2010). This motif is also present in several promoters involved in tocopherol biosynthesis (Welsch et al. 2007). MeJA response elements are often found in genes involved in plant pathogen interactions (Rouster et al. 1997). Lycopene synthesis in fruits treated with MeJA showed an inverted U-shaped dose response which significantly enhanced the lycopene content of the fruits and restored lycopene accumulation in mutants deficient in jasmonic acid (spr2 and def1) at a low concentration of 0.5 μM (Liu et al. 2012). The tomato DXS , GGPS (geranylgeranyl diphosphate synthase), PSY1 and PDS genes were up-regulated by MeJA treatment (Liu et al. 2012). Several other hormone response elements were identified in the promoters, including those responsible for the transduction of auxin and ethylene responses (Tables 2 and 3; Figs S1–S5). The GlLYCE promoter contained all the cis-regulatory elements found in other carotenogenic gene promoters except the ATGCAAAT motif present in the GlBCH promoter (Tables 2 and 3). The TA-rich region was unique to the GlLYCE promoter (Table 2). The diverse cis-regulatory elements of the GlLYCE gene promoter imply that diverse cis-acting elements may play crucial roles in determining the complex endogenous expression pattern in G. lutea when GlLYCE expression is first suppressed in petals from stage 1 (S1,
Physiologia Plantarum 150: 493–504. 2014
Cloning and functional analysis of the promoters that upregulate carotenogenic gene expression during flower development in Gentiana lutea Changfu Zhua,b,† , Qingjie Yangc,d,† , Xiuzhen Nia , Chao Baib , Yanmin Shenga , Lianxuan Shid , Teresa Capellb , Gerhard Sandmanne and Paul Christoub,f,∗ a
School of Life Sciences, Changchun Normal University, Changchun 130032, China ` Departament de Produccio´ Vegetal i Ciencia Forestal, Universitat de Lleida-Agrotecnio Center, Lleida 25198, Spain c College of Landscape Architecture, Northeast Forestry University, Harbin 150040, China d School of Life Sciences, Northeast Normal University, Changchun, 130024 China e Department of Molecular Biosciences, J.W Goethe Universitaet, Frankfurt, D-60054 Germany f ´ Companys, Barcelona 08010, Spain Institucio Catalana de Recerca i Estudis Avancats, Passeig Lluis b
Correspondence *Corresponding author, e-mail: [email protected] Received 17 June 2013; revised 24 October 2013 doi:10.1111/ppl.12129
Over the last two decades, many carotenogenic genes have been cloned and used to generate metabolically engineered plants producing higher levels of carotenoids. However, comparatively little is known about the regulation of endogenous carotenogenic genes in higher plants, and this restricts our ability to predict how engineered plants will perform in terms of carotenoid content and composition. During petal development in the Great Yellow Gentian (Gentiana lutea), carotenoid accumulation, the formation of chromoplasts and the upregulation of several carotenogenic genes are temporally coordinated. We investigated the regulatory mechanisms responsible for this coordinated expression by isolating five G. lutea carotenogenic gene (GlPDS , GlZDS , GlLYCB, GlBCH and GlLYCE ) promoters by inverse polymerase chain reaction (PCR). Each promoter was sufficient for developmentally regulated expression of the gusA reporter gene following transient expression in tomato (Solanum lycopersicum cv. Micro-Tom). Interestingly, the GlLYCB and GlBCH promoters drove high levels of gusA expression in chromoplastcontaining mature green fruits, but low levels in chloroplast-containing immature green fruits, indicating a strict correlation between promoter activity, tomato fruit development and chromoplast differentiation. As well as core promoter elements such as TATA and CAAT boxes, all five promoters together with previously characterized GlZEP promoter contained three common cis-regulatory motifs involved in the response to methyl jasmonate (CGTCA) and ethylene (ATCTA), and required for endosperm expression (Skn1_motif, GTCAT). These shared common cis-acting elements may represent binding sites for transcription factors responsible for co-regulation. Our data provide insight into the regulatory basis of the coordinated upregulation of carotenogenic gene expression during flower development in G. lutea.
Abbreviations – ABRE, abscisic acid responsive element; BCH, β-carotene hydroxylase; ERE, ethylene response element; GlBCH, Gentiana lutea β-carotene hydroxylase; GlLYCB, Gentiana lutea lycopene β-cyclase; GlPDS, Gentiana lutea phytoene desaturase; GlLYCE, Gentiana lutea lycopene -cyclase gene; GlZDS, Gentiana lutea ζ -carotene desaturase gene; GlZEP, Gentiana lutea zeaxanthin epoxidase; IMG, immature green; LA-PCR, long accurate polymerase chain reaction; LYCB, lycopene β-cyclase; MeJA, methyl jasmonate; MG, mature green; PCR, polymerase chain reaction; PDS, phytoene desaturase; PSY, phytoene synthase; UTR, untranslated region; ZDS, ζ -carotene desaturase; ZEP, zeaxanthin epoxidase. † These
authors contributed equally to this work.
Physiol. Plant. 150, 2014
493
Introduction The carotenoids are a large class of yellow, orange and red pigments derived from isoprenoid precursors. In higher plants, they accumulate in the chloroplasts of leaves and in the chromoplasts of many flowers and fruits. Carotenoids function as accessory pigments during photosynthesis and help to prevent photooxidation (Frank and Cogdell 1996, Demmig-Adams and Adams 2002). They are also precursors of the plant hormones abscisic acid (Creelman and Zeevart 1984) and strigolactone (Gomez-Roldan et al. 2008, Umehara et al. 2008). In flowers and fruits, the bright colors imparted by carotenoids help to attract pollinating insects and seed-dispersing animals (Bartley and Scolnik 1995), and the colors also provide agronomic value in fruit and vegetable crops as well as ornamental plants (Bartley and Scolnik 1995, Zhu et al. 2010). Carotenoids play a fundamental role in human nutrition as antioxidants and precursors of vitamin A, and a high dietary intake of carotenoids reduces the risk of several diseases (Fraser and Bramley 2004, Zhu et al. 2009, Bai et al. 2011, Farre et al. 2011). Genes encoding carotenogenic enzymes have therefore been cloned and used for metabolic engineering to increase the nutritional value and health-promoting properties of plants by improving the carotenoid content and composition (Sandmann et al. 2006, Zhu et al. 2007, 2013, Giuliano et al. 2008, Farre et al. 2010, 2011, Bai et al. 2011). However, the regulation of endogenous carotenoid formation in higher plants is poorly understood, restricting the extent to which the impact of metabolic engineering in crops can be predicted (Sandmann et al. 2006, Giuliano et al. 2008, Farre et al. 2010). The flowers in the Great Yellow Gentian (Gentiana lutea) contain abundant amounts of β-carotene and xanthophylls, making it a useful model to investigate the regulation of carotenoid biosynthesis in flowers (Zhu et al. 2002, 2003). The petals possess chromoplasts that originate from fully developed chloroplasts, and there is a temporal correlation among the accumulation of carotenoids, the formation of chromoplasts and the induction of the carotenogenic genes PSY (phytoene synthase), PDS (phytoene desaturase), ZDS (ζ -carotene desaturase), LYCB (lycopene β-cyclase), BCH (β-carotene hydroxylase) and ZEP (zeaxanthin epoxidase) (Zhu et al. 2002, 2003) (Fig. 1). Transcription of carotenogenic genes in G. lutea controls the accumulation of carotenoids during flower development (Zhu et al. 2002, 2003). The G. lutea zeaxanthin epoxidase (GlZEP ) promoter was recently cloned and characterized in transgenic tomato plants (Yang et al. 494
2012), where it was shown to drive the developmentally regulated expression of gusA, encoding the reporter β-glucuronidase (GUS). High levels of gusA expression were observed in chromoplast-containing fruits and petals. In contrast, only low levels of expression was seen in the immature green (IMG) chloroplast-containing fruits and leaves. GlZEP promoter activity was strictly associated with fruit development and chromoplast differentiation (Yang et al. 2012). Here we describe the isolation and analysis of five G. lutea carotenogenic gene promoters (Gentiana lutea phytoene desaturase gene (GlPDS ), Gentiana lutea ζ -carotene desaturase gene (GlZDS ), Gentiana lutea lycopene β-cyclase gene (GlLYCB), Gentiana lutea β-carotene hydroxylase (GlBCH ) and Gentiana lutea lycopene cyclase gene (GlLYCE ) promoters, the latter controlling the expression of lycopene -cyclase gene). We investigated the regulatory basis of coordinated upregulation during G. lutea flower development and searched for common cis-regulatory elements that could provide insight into the hierarchal control of carotenoid biosynthesis in flower petals.
Materials and methods Plant material
Gentiana lutea leaves were obtained from the Hokkaido Experimental Institute of Health Science, Japan. The tissues were frozen in liquid nitrogen immediately after harvesting and stored at −80◦ C. Tomato (Solanum lycopersicum cv. Micro-Tom) plants were grown in the greenhouse at 25◦ C with a 16-h photoperiod. Cloning the promoter sequences
Gentiana lutea Genomic DNA was extracted from 5 g of leaf tissue according to Edwards et al. (1991). Genomic DNA (20 μg) was completely digested with a restriction enzyme appropriate to the corresponding promoter sequence (Table 1) and self-ligated using 10 Weiss units of T4 DNA Ligase (Invitrogen, Carlsbad, CA) to generate circular molecules. These were used as templates for amplification by the long accurate polymerase chain reaction (LA-PCR), following the recommendations provided with the Takara LA-PCR Kit (Takara, Shuzo, Japan). The products were cloned into vector PCR® II TOPO® (TA Cloning Kit, Invitrogen, Carlsbad, CA) for sequencing using the Big Dye Terminator v3.1 Cycle Sequencing Kit on a 3130x1 Genetic Analyzer (Applied Biosystems, Foster City, CA). The restriction enzymes and primer sequences required for each of the five promoters are listed in Table 1.
Physiol. Plant. 150, 2014
Fig. 1. Carotenoid biosynthesis pathway in plants. Abbreviations: CRTISO, carotenoid isomerase; CYP97C, heme-containing cytochrome P450 carotene -ring hydroxylase; GGPP, geranylgeranyl diphosphate; HYDB, β-carotene hydroxylase [non-heme di-iron β-carotene hydroxylase (BCH) and heme-containing cytochrome P450 β-ring hydroxylases (CYP97A and CYP97B)]; LYCB, lycopene β-cyclase; LYCE, lycopene -cyclase; PDS, phytoene desaturase; PSY, phytoene synthase; VDE, violaxanthin de-epoxidase; ZDS, ζ -carotene desaturase; ZEP, zeaxanthin epoxidase; Z-ISO, ζ -carotene isomerase. This figure was modified based on Zhu et al. (2010, 2013).
Physiol. Plant. 150, 2014
495
Table 1. PCR primers used to clone gene promoters, to construct promoter-gusA fusion plasmids and to clone genomic DNA to confirm the isolated promoter sequences.
Gene PDS
Promoter cloning (digested with EcoRI) Promoter-GUS construct
Promoter cloning (digested with BamHI) Promoter-GUS construct Genomic DNA cloning
LYCB
Promoter cloning (digested with EcoRI) Promoter-GUS construct Genomic DNA cloning
BCH
Promoter cloning (digested with PstI) Promoter-GUS construct Genomic DNA cloning
LYCE
Accession numbers for cDNA or genomic DNA (gDNA)
F1: 5 -GGACACATATCTGCTGTTAACATAGGTAGGCAAGG-3 R1: 5 -GGTAGGATAAAATTCACTAAGTTGAAGGTGAAAGGG-3 F2: 5 -GTCGACAATTCATGAGTTCAAACCCGTGATTCGTTC-3 R2: 5 -GGATCCATATCAAAGCTGGTACCAAACAGAGCAAAC-3 F3: 5 -AATTCATGAGTTCAAACCCGTGATTCGTTC-3 R3: 5 -GGTAGGATAAAATTCACTAAGTTGAAGGTG-3 F1: 5 -AGGCTTGTTTCCACCGGAACCTGAACATTATCGG-3
+145 to +179 +1 to +36 −1077 to −1048 −1 to −30 −1077 to −1048 +7 to +36 +415 to +448
EF203257 (cDNA)
Purposes
Genomic DNA cloning ZDS
Primers
Positions (+1 is the first nucleotide of cDNA)
Promoter cloning (digested with BglII) Promoter-GUS construct Genomic DNA cloning
R1: 5 -TCAATGCAATCCTGAAGAACTTAGACCATTGCTGT-3 F2: 5 -GTCGACGATCCGTTTAACGCTTAGATCGTCGTCATC-3 R2: 5 -GGATCCTTAGATTATGTTAAAACAAGCATCAAAGCT-3 F3: 5 -GATCCGTTTAACGCTTAGATCGTCGTCATC-3 R3: 5 -TCAACGCAATCCTGAAGAACTTAGACCATT-3 F1: 5 -CACCCTTTATGTGGGTTTGTTGATAAAGTCTGCTCC-3
+123 to +157 −766 to −737 −1 to −30 −766 to −737 +128 to +157 +385 to +420
R1: 5 -CCAATTGTTGCCCAGAAAGCAAACTGAATTTCGAG-3 F2: 5 -GTCGACAATTCAACATCAAATGGCTAGTTGGACTTT-3 R2: 5 -GGATCCTCTGCGACACGTCTGACGGTGGAGTTAATT-3 F3: 5 -AATTCAACATCAAATGGCTAGTTGGACTTT-3 R3: 5 -CTTTTCTCTATTGGGTGTACTTTATTTCTT-3 F1: 5 -TTGGTCTCCGGTAGAAACAGCAACATTCATTGCCGT-3
+259 to +293 −1506 to −1477 −1 to −30 −1506 to −1477 +313 to +342 +135 to +170
R1: 5 -GTGACCGGAAACCGGAGAAACAGCTGAGAGGAGAG-3 F2: 5 -GTCGACGCCTAGGCCTACAATGCACAACCTAATACC-3 R2: 5 -GGATCCACCACATGTTTTCTTTGGATTGTGAAACTC-3 F3: 5 -GCCTAGGCCTACAATGCACAACCTAATACC-3 R3: 5 -TACCGGAAAATGGTGACCGGAAACCGGAGA-3 F1: 5 -TATTCGAGACGATCAAGAAGAAGGAGGATTCTCAGTG-3
+73 to +107 −1744 to −1715 −1 to −30 −1744 to −1715 +90 to +119 +310 to +346
R1: 5 -GTTGTTTGAAGTGTGAGCGCAGGGCAGCCGCTAAAG-3 F2: 5 -GTCGACGATCTTAAGTTATGTTCTAGTAAGTAAAGC-3 R2: 5 -GGATCCTCGTTTCAAGAGTTGCCACGTGGCTTGAAA-3 F3: 5 -GATCTTAAGTTATGTTCTAGTAAGTAAAGC-3 R3: 5 -GTTTGGCCCTTTATTTGTGTAGTTTCCGTG-3
+31 to +66 −938 to −909 −1 to −30 −938 to −909 +226 to +255
Promoter-gusA constructs The promoter sequences were fused to the gusA gene in vector pBI101 (Clontech Laboratories, Mountain View, CA) (Jefferson et al. 1987) by amplifying each promoter using primers containing additional sequences to provide appropriate restriction sites. For example, the full-length GlPDS promoter region was amplified from G. lutea genomic DNA using forward primer F2 (5 -GTC GAC AAT TCA TGA GTT CAA ACC CGT GAT TCG TTC-3 , introducing a Sal I site in italics) and reverse primer R2 (5 -GGA TCC ATA TCA AAG CTG GTA CCA AAC AGA GCA AAC-3 , introducing a BamHI site in italics). The 1077-bp amplified promoter fragment was then transferred to the PCR® II TOPO® vector using the Invitrogen TA Cloning® Kit, to yield intermediate vector pCR-GlPDSPro. The pCR-GlPDSPro and pBI101 vectors were digested with Sal I and BamHI, allowing 496
DQ226992 (gDNA) DQ226992 (gDNA) EF203258 (cDNA)
JQ417647 (gDNA) JQ417647 (gDNA) EF203253 (cDNA)
JQ417648 (gDNA) JQ417648 (gDNA) EF203255 (cDNA)
EF203261 (gDNA) EF203261 (gDNA) EF203256 (cDNA)
EU592045 (gDNA) EU592045 (gDNA)
the GlPDSPro fragment to be inserted upstream of gusA, yielding the final construct pBI-GlPDSPro-GUS (PDSgusA). The constructs pBI-GlZDSPro-GUS (ZDS-gusA), pBI-GlLYCBPro-GUS (LYCB-gusA), pBI-GlBCHPro-GUS (BCH-gusA) and pBI-GlLYCEPro-GUS (LYCE-gusA) were constructed in an analogous manner using the specific primers listed in Table 1. The integrity of all intermediate and final constructs was confirmed by sequencing.
Transient expression of promoter-gusA constructs in tomato Plasmids pBI101, pBI121 (35S-gusA), pBI-GlPDSProGUS (PDS-gusA), pBI-GlZDSPro-GUS (ZDS-gusA), pBI-GlLYCBPro-GUS (LYCB-gusA), pBI-GlBCHPro-GUS (BCH-gusA) and pBI-GlLYCEPro-GUS (LYCE-gusA) were introduced into Agrobacterium tumefaciens
Physiol. Plant. 150, 2014
strain LBA 4404 by electroporation (Mattanovich et al. 1989). Individual colonies were seeded into 5-ml aliquots of YEM medium (0.5% beef extract, 0.1% yeast extract, 0.5% peptone, 0.5% sucrose, 2 mM MgSO4 , pH 7.2) containing 50 μg ml−1 kanamycin and 25 μg ml−1 rifampicin, and were shaken at 300 rpm, 28◦ C overnight. Each culture was then used to inoculate 50 ml of induction medium (YEM medium supplemented with 20 μM acetosyringone, 10 mM MES, pH 5.6) containing 50 μg ml−1 kanamycin and 25 μg ml−1 rifampicin, followed by incubation as above. Bacteria were recovered by centrifugation (2700 g ), resuspended in infiltration medium (10 mM MgCl2 , 10 mM MES, 200 μM acetosyringone, pH 5.6) to an OD600 of approximately 1.0, and then incubated at room temperature with gentle agitation (20 rpm) for 3 h. Approximately 450 μl of the infiltration medium was then injected into fruits at the immature green stage (fruit diameter 0.8–1.0 cm, about half full size, 15–20 days after anthesis. Fruits were not fully expanded and still green, defined as ‘immature green’) (Akihiro et al. 2008), and 600 μl was injected into fruits at the mature green (MG) stage (fruit diameter 1.5–2.0 cm, 30–35 days after anthesis. Fruits were fully expanded and green, defined as ‘mature green’) (Akihiro et al. 2008), in both cases through the stylar apex using a 1-ml syringe with needle (Orzaez et al. 2006). Injected fruits were left on the vine for 3 days before harvesting and sectioning for histochemical staining. All the experiments were repeated six times in six independent Micro-Tom plants. Histochemical GUS assay Histochemical GUS assays were carried out according to Jefferson et al. (1987) with minor modifications. Tissues were incubated at 37◦ C for 12 h in the dark in 1 mM XGluc (5-bromo-4-chloro-3-indolyl-β-D-glucuronide) in 100 mM sodium phosphate (pH 7.0), 10 mM EDTA (ethylenediaminetetraacetic acid), 0.5 mM potassium ferricyanide, 0.5 mM potassium ferrocyanide, 0.3% (v/v) Triton X-100 and 20% (v/v) methanol to eliminate endogenous GUS expression (Kosugi et al. 1990). After staining, the tissues were destained in an ethanol series (50, 70, 80 and 95%) to remove chlorophyll, and then stored in 70% (v/v) ethanol, and photographed with a digital camera.
Results Cloning the promoter sequences The promoters for GlPDS , GlZDS , GlLYCB, GlBCH and GlLYCE were cloned by inverse PCR using cleaved and
Physiol. Plant. 150, 2014
circularized G. lutea genomic DNA as the template and outward-facing primers based on the corresponding cDNA sequences (Zhu et al. 2002, 2003). The primers, GenBank accession numbers and restriction enzymes used to prepare the genomic templates are listed in Table 1. After sequencing the products, DNA fragments of 1113 bp (GlPDS ), 923 bp (GlZDS ), 1848 bp (GlLYCB), 1863 bp (GlBCH ) and 1193 bp (GlLYCE ) were isolated directly from genomic DNA using gene-specific primers (Primers and GenBank accession numbers are listed in Table 1). All the fragments comprised the upstream promoters and the 5 -untranslated region (UTR). The full-length promoter fragments were defined as 1077 bp (GlPDS ), 766 bp (GlZDS ), 1506 bp (GlLYCB), 1744 bp (GlBCH ) and 938 bp (GlLYCE ) with position +1 assigned to the first nucleotide of the cDNAs (Zhu et al. 2002, 2003). Construction of promoter-gusA fusion genes and transient expression in tomato fruits The full-length promoter regions were amplified from G. lutea genomic DNA using forward primers containing a Sal I restriction site (5 -GTCGAC-3 ) and reverse primers containing a BamHI restriction site (5 -GGATCC-3 ). The promoter fragments were inserted upstream of the gusA gene in vector pBI101, which had been digested with Sal I and BamHI. Tomato fruits at the IMG and MG stages were injected with bacterial cultures carrying the vectors pBI101 (promoterless gusA), pBI121 (35S-gusA, constitutive Cauliflower mosaic virus 35S promoter), pBI-GlPDSPro-GUS (PDS-gusA), pBI-GlZDSPro-GUS (ZDS-gusA), pBI-GlLYCBPro-GUS (LYCB-gusA), pBIGlBCHPro-GUS (BCH-gusA) and pBI-GlLYCEPro-GUS (LYCE-gusA). Fruits were harvested 3 days later and transverse sections were stained for GUS expression. As expected, both IMG fruits (Fig. 2) and MG fruits (Fig. 3) showed significant GUS expression when transformed with the 35S-gusA construct but no GUS expression when transformed with the promoterless control vector pBI101. We evaluated the five carotenoid constructs (PDS-gusA, ZDS-gusA, LYCB-gusA, BCH-gusA and LYCE-gusA) by transient expression in tomato fruits as above, using the 35S-gusA vector as a control. Histochemical staining revealed no GUS expression in IMG fruits transformed with PDS-gusA, ZDS-gusA and LYCEgusA, and with only low GUS expression levels in IMG fruits transformed with the LYCB-gusA and BCH-gusA constructs (Fig. 2) but higher GUS expression for these constructs, as well as PDS-gusA and LYCE-gusA, in MG fruits (Fig. 3). The ZDS-gusA construct produced low level GUS expression in the mesocarp of MG fruits 497
Fig. 2. Histochemical GUS staining of typical transgenic MicroTom IMG fruit expressing pBI101, 35S-gusA, PDS-gusA, ZDS-gusA, LYCB-gusA, BCH-gusA, LYCE-gusA, respectively. Abbreviations: pBI101, pBI101; 35S-gusA, pBI121; PDS-gusA, pBI-GlPDSPro-GUS; ZDS-gusA, pBI-GlZDSPro-GUS; LYCB-gusA, pBI-GlLYCBPro-GUS; BCH-gusA, pBIGlBCHPro-GUS (BCH-gusA); LYCE-gusA, pBI-GlLYCEPro-GUS.
(Fig. 3). Overall, the GUS expression of the five constructs during fruit development was similar, indicating that all cis-acting elements necessary to confer GUS expression in tomato fruits are contained within the isolated promoter sequences. Promoter analysis A search of the PlantCARE database of plant cis-regulatory elements (Lescot et al. 2002, http://bioinformatics. psb.ugent.be/webtools/plantcare/html) revealed potential TATA and CAAT boxes in all five promoters (Table 2; Figs S1–S5, Supporting Information). In addition to these core promoter elements, we found that all five promoters contained three common cis-regulatory motifs: CGTCAmotif which is linked to methyl jasmonate (MeJA) signaling (Rouster et al. 1997), and ATCTA-motif which was recently found in the AtPSY promoter and is known to interact with the ethylene response transcription factor RAP2.2 (Welsch et al. 2003, 2007). Another motif present was Skn-1_motif (GTCAT) that is required for endosperm expression (Takaiwa et al. 1991). Carotenoid biosynthesis is known to be regulated by light (Von Lintig et al. 1997, Simkin et al. 2003, Li et al. 498
Fig. 3. Histochemical GUS staining of typical transgenic MicroTom MG fruit expressing pBI101, 35S-gusA, PDS-gusA, ZDS-gusA, LYCB-gusA, BCH-gusA, LYCE-gusA, respectively. Abbreviations: pBI101, pBI101; 35S-gusA, pBI121; PDS-gusA, pBI-GlPDSPro-GUS; ZDS-gusA, pBI-GlZDSPro-GUS; LYCB-gusA, pBI-GlLYCBPro-GUS; BCH-gusA, pBIGlBCHPro-GUS (BCH-gusA); LYCE-gusA, pBI-GlLYCEPro-GUS.
2008, Welsch et al. 2008), and we identified at least one light-response motif in each promoter, e.g. Box I, Box II, ATC, GT-1, TCT, Sp1, G-box, ACE and AEbox (Table 3; Figs S1–S5) providing the basis for the regulation of carotenogenic gene expression according to day length and other cues involved in the control of flower development. At least one MYB binding site, involved in drought tolerance (Abe et al. 2003), was also found in each promoter (Table 3). We also identified a number of additional cis-acting elements shared among some of the promoters, including an ethylene response element (ERE) in the GlPDS and GlLYCE promoters, abscisic acid responsive element (ABRE) and auxin response element (TGA-element) in the GlBCH and GlLYCE promoters, and a W1-box that responds to fungal elicitors in the GlLYCB and GlLYCE promoters. The GlZDS , GlLYCB and GlLYCE promoters shared a GCN4 motif common to endosperm-specific promoters, whereas the GlLYCE and GlBCH promoters contained four TA-rich regions and an ATGCAAAT motif that binds GCN4 in rice, respectively (Table 2).
Discussion The carotenoid biosynthesis pathway (Fig. 1) has been completely elucidated through a combination of biochemical, genetic and transgenic approaches
Physiol. Plant. 150, 2014
Physiol. Plant. 150, 2014
499
TATA-box
CAAT-box
ERE
PDS −76 −624, −552 (CAAT); −287, −1036, −1003 (1077 bp) (TATATATA) −39 (CAAAT); −433, (ATTTCAAA) −168, −87 (CAATT) ZDS −47 (TAATA); −654, −626, −285 (CAAT); (766 bp) −16 (TTTA) −706, −580, −560, −501; −380, −60 (CAAAT); −415, −353, −157 (CAATT) LYCB −44 −926, −735, −685, −579, (1506 bp) (TATAAA) −517, −403, −145 (CAAT); −1496, −1361, −725, −308 (CAAAT); −871, −831, −776, −533, −381, −158, −151, −49 (CAATT) BCH −142 −1733, −1550, −1147, (1744 bp) (TAATA) −1136, −353, −22 (CAAT); −1390, −1310, −1163, −647, −483 (CAAAT); −1540, −1208, −1154, −935, −928, −264, −195, −188, −155, −137 (CAATT) LYCE −73 −834 (CAAT); −683, −599, −157 (938 bp) (TAATA) −93 (CAAAT) (ATTTCAAA)
Promoter
−1399 (AACGAC)
−908 (AACGAC)
−440, −22 (CACGTG); −21 (ACGTGGC)
TGA-element (auxin response)
−1670, −738 −1144 (CACGTG); (TGACG−70 TAA) (TACGTG)
ABRE
TGA-box (auxin response)
−696 (TTGACC)
−1123 (TTGACC)
Box-W1
−27 (CAAGCCA)
−626 (TGAGTCA)
−302 (TGT GTCA)
GCN4
−1166 (ATACAAAT)
ATGCAAAT motif
−391, −389, −387, −385 (TATATATA TATATATA TATATA)
TA-rich region
Table 2. Specific cis-acting regulatory elements found in different carotenogenic gene promoters from Gentiana lutea. The positions of the cis-acting regulatory elements are denoted relative to the positions of cDNA (+1 is the first nucleotide of cDNA). GenBank accession numbers for the promoter sequences analyzed are as follows: PDS, DQ226992; ZDS, JQ417647; LYCB, JQ417648; BCH, EF203261; ZEP, EF203262; LYCE, EU592045.
Table 3. The consensus cis-acting regulatory elements in different promoters of carotenogenic genes from Gentiana lutea. The positions of the cis-acting regulatory elements are denoted relative to the positions of cDNA (+1 is the first nucleotide of cDNA). GenBank accession numbers for the promoter sequences analyzed are as follows: PDS, DQ226992; ZDS, JQ417647; LYCB, JQ417648; BCH, EF203261; ZEP, EF203262; LYCE, EU592045.
Promoter
Cis-acting regulatory element involved in light responsiveness
ATCTA-motif (RAP2.2 motif in AtPSY)
PDS (1077 bp)
−1035, −1002 (Box I, TTTCAAA); −795, −607 (ATC-motif, AGTAATCT)
−603 (ATCTA)
ZDS (766 bp)
−420 (TCT-motif, TCTTAC); −397 (GT-1 motif, GGTTAAT); −49 (Box 4, ATTAAT) −1407 (TCT-motif, TCTTAC)
−6 (ATCTA)
LYCB (1506 bp) BCH (1744 bp)
ZEP (2225 bp)
LYCE (938 bp)
−1670, −738 (G-box, CACGTG); −1427 (G-box, CACGTT); −114 (G-box, TGACGTGG); −70 (G-box, TACGTG); −1516, −1254 (Box 4, ATTAAT); −1453 (AE-box, AGAAACTT); −1410 (AE-box, AGAAACAT); −112 (ACE, ACGTGGA); −1725, −1420 (MRE, AACCTAA); −33 (GAG-motif, AGAGAGT) −2217 (GT-1-motif, GCGGTAATT); −2160, −275 (Box I, TTTCAAA); −1613 (G-box, CACATGG); −1585, −1355 (G-box, CACGTC); −1567 (chs-CMA2a, TCACTTGA); −1068 (GAG-motif, AGAGAGT); −599, −262, −237, −215 (Box 4, ATTAAT) −614 (Sp1, GGGCGG); −531 (ATC-motif, AGCTATCCA); −464 (G-box, CACATGG); −440, −22 (G-Box, CACGTG); −156 (Box I, TTTCAAA); −23 (Box II, CCACGTGGC)
−597 (CGTCA-motif, CGTCA); −266 (TGACG-motif, TGACG) −743 (CGTCA-motif, CGTCA)
Skn-1_motif required for endosperm expression
MYB binding site involved in droughtinducibility
−596 (GTCAT)
−46 (MBS, TAACTG)
−742, −299 (GTCAT)
−546 (MBS, TAACTG)
−1352, −954 (ATCTA) −1637, −1617, −1529, −415 (ATCTA)
−18 (CGTCA-motif, CGTCA) −1144, −114 (TGACG-motif, TGACG); −817 (CGTCA-motif, CGTCA)
−623, −190 (GTCAT) −1287, −967, −816, −227, −207 (GTCAT)
−1372 (MBS, CAACTG) −1602 (MBS, TAACTG)
−2191, −1512 (ATCTA)
−2064, −1527 (CGTCA-motif, CGTCA); −950 (TGACG-motif, TGACG)
−658, −557 (GTCAT)
−1191, −997 (MBS, CAACTG)
−765, −83 (ATCTA)
−588 (TGACG-motif, TGACG); −485 (CGTCA-motif, CGTCA)
−539, −435 (GTCAT)
−524 (MBS, CAACTG)
(Zhu et al. 2007, 2010, 2013, Giuliano et al. 2008, Farre et al. 2010). During flower development in G. lutea, carotenoids accumulate in the petals, accompanied by a shift in profile from lutein derived from α-carotene to xanthophylls derived from β-carotene (Zhu et al. 2003). PSY and ZDS mRNA levels increase six- to sevenfold during this process, whereas PDS and other genes required for β-carotene formation and metabolism (LYCB, BCH and ZEP ) are induced approximately twofold (Zhu et al. 2002, 2003). In contrast, the abundance of the LYCE transcript is reduced to less than half of its normal value (Zhu et al. 2003). These 500
Cis-acting regulatory element involved in MeJAresponsiveness
antagonistic changes shift the relative abundance of carotenoids from the α to β branches. These data, combined with the developmental regulation of the GlZEP promoter during tomato fruit development and chromoplast differentiation (Yang et al. 2012), suggest that carotenogenesis in the flower petal is regulated by multiple developmental and environmental cues. It is clear that carotenogenesis in G. lutea petals during flower development is under transcriptional control (Zhu et al. 2002, 2003), but the mechanisms are largely unknown. The coordinated expression of several genes may reflect similarities between their promoters,
Physiol. Plant. 150, 2014
or potentially could reflect clustering within particular chromosomal domain (Caron et al. 2001, Winzer et al. 2012). Similar expression profiles may also be caused by the coordinated action of more than one set of transcription factors, in which case the promoter regions of co-regulated genes would be heterogeneous, e.g. the well-known C1 and R transcription factors which control the flavonoid biosynthesis pathway (Grotewold et al. 2000, and other references therein). To investigate the basis of coordinated carotenogenic gene expression during G. lutea flower development, we cloned promoter fragments upstream of GlPDS , GlZDS , GlLYCB, GlBCH and GlLYCE directly from G. lutea genomic DNA by inverse PCR. Sequencing identified putative TATA and CAAT boxes in all the promoters (Table 2). We investigated GUS expression driven by these promoters in ripening tomato fruits which provide as a useful model for studying the regulation of carotenogenic gene expression in chromoplastcontaining petals (Yang et al. 2012). We evaluated the constructs PDS-gusA, ZDS-gusA, LYCB-gusA, BCHgusA and LYCE-gusA by transient expression in tomato and found that all the promoters were sufficient for developmentally regulated reporter gene expression in the tomato fruits. The GlLYCB and GlBCH promoters sustained strong GUS expression in chromoplast-containing MG fruits (Fig. 3) but only low levels in chloroplast-containing IMG fruits (Fig. 2). This is similar to the endogenous gene expression profiles of the genes in G. lutea (Zhu et al. 2002, 2003) and the expression of the GlZEP promoter in transgenic tomatoes (Yang et al. 2012), confirming that carotenogenic gene promoter activity is strictly associated with tomato fruit development and chromoplast differentiation. The integration of expression profiles and promoter sequences can help to identify common and putative functionally relevant cis-acting elements (Werner 2001, Kim and Kim 2006, Yamamoto et al. 2011). Comparative bioinformatic studies on promoter regions of carotenoid genes may elucidate common binding motifs involved in carotenoid formation (Fraser and Bramley 2004). We therefore compared the isolated promoter sequences with the GlZEP promoter (Yang et al. 2012) using the Plant-CARE and PLACE databases (Higo et al. 1999, Lescot et al. 2002). All the promoters contained at least one CGTCA-motif involved in the response to MeJA, one ATCTA-motif that interacts with the ethylene response factor RAP2.2 in Arabidopsis thaliana (Welsch et al. 2003, 2007), one Skn-1_motif (GTCAT) that is required for endosperm expression, and consensus motifs involved in responses to light and drought (Table 3). These shared cis-acting elements could represent
Physiol. Plant. 150, 2014
binding sites for transcription factors responsible for co-regulation (Kreiman 2004, Haberer et al. 2006, Obayashi et al. 2007, Lenka et al. 2009). The existence of CGTCA and ATCTA motifs representing two hormone response pathways may give an example of functional cooperation or independent activities. The ATCTA motif in the AtPSY promoter binds RAP2.2, a member of the APETALA2/ERE-binding protein family. Overexpression of RAP2.2 induced only minor changes in the carotenoid profile of non-green Arabidopsis tissues, suggesting that additional factors may contribute to the regulation of PSY (Welsch et al. 2007). Single copy of the ATCTA motif is found in other carotenogenic gene promoters such as Arabidopsis DXS (encoding deoxyxylulose phosphate synthase) and PDS (Welsch et al. 2007), tomato and maize PDS (Welsch et al. 2007), and tomato CYC-B encoding lycopene βcyclase (Dalal et al. 2010). This motif is also present in several promoters involved in tocopherol biosynthesis (Welsch et al. 2007). MeJA response elements are often found in genes involved in plant pathogen interactions (Rouster et al. 1997). Lycopene synthesis in fruits treated with MeJA showed an inverted U-shaped dose response which significantly enhanced the lycopene content of the fruits and restored lycopene accumulation in mutants deficient in jasmonic acid (spr2 and def1) at a low concentration of 0.5 μM (Liu et al. 2012). The tomato DXS , GGPS (geranylgeranyl diphosphate synthase), PSY1 and PDS genes were up-regulated by MeJA treatment (Liu et al. 2012). Several other hormone response elements were identified in the promoters, including those responsible for the transduction of auxin and ethylene responses (Tables 2 and 3; Figs S1–S5). The GlLYCE promoter contained all the cis-regulatory elements found in other carotenogenic gene promoters except the ATGCAAAT motif present in the GlBCH promoter (Tables 2 and 3). The TA-rich region was unique to the GlLYCE promoter (Table 2). The diverse cis-regulatory elements of the GlLYCE gene promoter imply that diverse cis-acting elements may play crucial roles in determining the complex endogenous expression pattern in G. lutea when GlLYCE expression is first suppressed in petals from stage 1 (S1,
