Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A rice promoter containing both novel positive and negative cis -elements for regulation of green tissue-specific gene expression in transgenic plants

A rice promoter containing both novel positive and negative cis -elements for regulation of green... <h1>Introduction</h1> Several genes from non-plant or non-crop species have been demonstrated to be useful for crop improvement. For example, the Bt gene from Bacillus thuringiensis , encoding endotoxin, has been used to breed insect-resistant cotton, maize, potato and rice ( Tu et al ., 2000 ; Nester et al ., 2002 ). The Bar gene from Streptomyces hygroscopicus , encoding phosphinothricin acetyltransferase ( Block et al ., 1987 ), and the EPSPS gene from Agrobacterium tumefaciens , encoding 5-enolpyruvylshikimate-3-phosphate synthase ( Funke et al ., 2006 ), have been used to breed herbicide-resistant soybean, cotton, canola and maize ( Duke, 2005 ; Pline-Srnic, 2005 ). The seed albumin gene from Amaranthus hypochondriacus has been used to improve the nutritive value of potato ( Chakraborty et al ., 2000 ). A cold-tolerant pyruvate orthophosphate dikinase gene from Flaveria brownii has been introduced into maize to improve cold tolerance ( Ohta et al ., 2004 ), and the Arabidopsis DREM1A gene has been shown to improve the drought tolerance of wheat ( Pellegrineschi et al ., 2004 ). Although some of these transgenic crops carrying genes from non-plants or non-crops are commercially available, in many countries and areas public concern regarding the food safety of genetically modified organisms has limited the application of transgenic crops. Using tissue-specific promoters to avoid the expression of foreign genes in seed or fruit may help to relieve this public concern. Many of the genes that can be used for crop improvement, including those from plants and other organisms, act in a dosage-dependent manner, and plants expressing high levels of this type of gene have a better modified phenotype. For instance, the function of Bt is dosage dependent ( Roush, 1998 ; Bates et al ., 2005 ). Many plant abiotic- and biotic-responsive genes are also dosage dependent ( Hu et al ., 2006 ; Qiu et al ., 2007 ). Thus, strong constitutive expression promoters, such as the cauliflower mosaic virus 35S promoter and the maize ubiquitin gene promoter ( Odell et al ., 1985 ; Cornejo et al ., 1993 ), have been used widely to express the foreign genes in transgenic plants. However, constitutive expression of a foreign gene may be harmful to the host plant, causing sterility, retarded development, abnormal morphology, yield penalty, altered grain composition or transgene silencing ( Sinha et al ., 1993 ; Matzke et al ., 2000 ; Kurek et al ., 2002 ; Cheon et al ., 2004 ; Xu et al ., 2006 ). Using a strong tissue-specific or inducible promoter to restrict gene expression to only the required tissue or at a particular time may solve this type of problem. Although a large number of tissue-specific promoters have been identified, strong tissue-specific promoters that can be used for various crop improvements are limited. Some cis -acting elements regulating tissue-specific gene expression have been identified. The ACGTROOT1, ROOTMOTIFTAPOX1, WUSATAg, OSE1ROOTNODULE and OSE2ROOTNODULE elements promote root-specific gene expression ( Salinas et al ., 1992 ; Elmayan and Tepfer, 1995 ; Kamiya et al ., 2003 ; Vieweg et al ., 2004 ). DPBFCOREDCDC3, DOFCOREZM, TGACGTVMAMY and ESP activate embryo-, endosperm- or cotyledon-specific gene expression ( Kim et al ., 1997 ; Vicente-Carbajosa et al ., 1997 ; Yanagisawa and Schmidt, 1999 ; Yamauchi, 2001 ; Vickers et al ., 2006 ). CANBNNAP functions as a repressor in embryo-specific gene expression and as an activator in endosperm- and leaf-specific gene expression ( Ellerstrom et al ., 1996 ). POLLEN1LELAT52 promotes pollen-specific gene expression ( Bate and Twell, 1998 ), and RGATAOS activates phloem-specific gene expression ( Yin et al ., 1997 ). An AC-rich element stimulates gene expression in xylem tissue ( Lauvergeat et al ., 2002 ). An element in the –199 to –186 region of the bean GRP 1.8 gene promoter represses gene expression in vascular tissue ( Keller and Baumgartner, 1991 ). RYREPEATBNNAPA and TGTCACACMCUCUMISIN are activators for seed- and fruit-specific gene expression, respectively ( Ezcurra et al ., 1999 ; Yamagata et al ., 2002 ). The regulatory mechanism of tissue-specific gene expression remains unclear, however, and information concerning cis -regulatory elements is still limited, and cannot explain the regulation of various types of tissue-specific gene expression. In this study, we isolated and characterized the rice green tissue-specific promoter P D54O . Five novel tissue-specific cis -acting elements were identified from P D54O . They functioned as activators, repressors, or both in the regulation of leaf-, stem-, root- and panicle-specific gene expression. Using P D54O , truncated P D54O , mutated P D54O or its cis -elements to regulate the cry1Ac gene or a reporter gene resulted in insect-resistant rice and transgenic rice with various tissue-specific expression patterns of the reporter gene, indicating that the original and modified promoters and their cis -elements may prove to be useful in rice breeding programmes. <h1>Results</h1> <h2>Identification and isolation of a promoter not expressed in seed</h2> Using complementary DNA (cDNA) array technology, we identified a cDNA that was not expressed in seed, but was expressed in other tissue ( Figure 1a ). Sequence analysis indicated that this cDNA corresponded to LOC_Os08g10020, according to the rice genome annotation of the Institute for Genomic Research ( http://rice.tigr.org ), encoding a photosystem II 10-kDa polypeptide. This gene was named D54O . Further analysis showed that D54O was expressed in leaf and sheath, but not in root, stem or panicle, at the grain-filling stage ( Figure 1b ). A 2.1-kb fragment harbouring the –2134 to +41 region (translation initiation site of D54O as +1) upstream of D54O was isolated from rice variety Nongken 58. Sequence analysis showed that this fragment contained the basal regulatory elements, TATA box for RNA polymerase binding and CAAT box for transcription frequency regulation (Table S1 and Figure S1, see ‘Supplementary material’); this fragment also harboured several known cis -elements regulating tissue-specific gene expression, including root, phloem, pollen, embryo, endosperm, cotyledon, seed and fruit (Table S1 and Figure S1). Thus, this fragment was considered to harbour the full promoter of D54O and was designated as P D54O . <h2>Identification of the tissue-specific promoter region</h2> P D54O and a series of truncated P D54O were used to regulate the expression of the reporter gene ॆ-glucuronidase ( GUS ) in rice variety Mudanjiang 8 ( Figure 1c ). In transgenic plants carrying P D54O :GUS , histochemical staining revealed GUS expression in leaf, sheath, ligule and lemma palea of the young panicle, but not in lemma palea of the mature panicle, endosperm, embryo, stem and root ( Figure 2a ). In addition to the expression in leaf, sheath and young panicle, quantitative analysis revealed GUS expression in stem, endosperm/embryo and root of P D54O :GUS plants, but the expression levels in the three tissues were only 7.4%, 1.1% and 0.8%, respectively, of the GUS expression in leaf tissue ( Figure 2b ). The expression levels of P D54O :GUS were 44.5%, 52%, 5%, 5%, 1.2% and 1.5% of those of P 35S :GUS in leaf, sheath, young panicle, stem, root and endosperm/embryo, respectively ( Figure 2b ). GUS expression levels in transgenic plants carrying a single copy of P D54O :GUS or truncated P D54O :GUS constructs ( P D54O-1439 :GUS , P D54O-944 :GUS , P D54O-723 :GUS , P D54O-544 :GUS , P D54O-264 :GUS or P D54O-61 :GUS ; Figure 1c ) were compared by quantitative analysis. Plants carrying P D54O-1439 :GUS showed a similar GUS expression pattern to plants carrying P D54O :GUS , except that the GUS levels in the former were 22%, 21% and 24% lower in leaf, sheath and young panicle, respectively ( Figure 2b ). Compared with P D54O-1439 :GUS plants, GUS activities in plants carrying P D54O-944 :GUS were approximately six- and five-fold higher in stem and root, respectively, but showed no significant difference in other tissues. Compared with plants carrying P D54O-944 :GUS , GUS activity in P D54O-723 :GUS plants was approximately 29% lower in leaf and sheath, showed no significant difference in stem and endosperm/embryo, and was 8.5- and 4.1-fold higher in young panicle and root, respectively. Plants carrying P D54O-544 :GUS had 3.5-, 3.7-, 2.8-, 1.8- and 5.0-fold higher GUS activities compared with plants carrying P D54O-723 :GUS , and 1.7-, 1.8-, 18.3-, 7.6- and 59.4-fold higher GUS activities compared with plants carrying P D54O :GUS in leaf, sheath, young panicle, stem and root, respectively, but showed no GUS activity in endosperm/embryo. No GUS activity was detected in endosperm/embryo in P D54O-264 :GUS or P D54O-61 :GUS plants ( Figure 2b ). Compared with plants carrying P D54O :GUS , the GUS levels in P D54O-264 :GUS plants were 70%, 71%, 59% and 72% lower in leaf, sheath, young panicle and stem, respectively, and 15-fold higher in root. GUS activity was nearly undetectable in histochemical staining assays of P D54O-61 :GUS plants, which showed 95%, 95%, 87% and 92% lower GUS activities in leaf, sheath, young panicle and stem, respectively, but no significant difference in root, compared with plants carrying P D54O :GUS . These results suggest that the –2134 to –545 region of P D54O contains tissue-specific cis -elements that positively regulate gene expression in leaf, sheath and endosperm/embryo and negatively regulate gene expression in leaf, sheath, young panicle, stem and root. In addition, the –544 to –62 region contains cis -elements that positively regulate gene expression in leaf, sheath, young panicle and stem, and the –264 to –62 region contains cis -elements that negatively regulate gene expression in root. The positive regulatory elements in P D54O activated gene expression only slightly in stem, root and endosperm/embryo. <h2>Identification of cis -elements controlling tissue-specific expression</h2> Fifty-eight double-stranded DNA probes (Table S2, see ‘Supplementary material’) covering the different regions of P D54O were used to study DNA–protein binding and to determine the tissue-specific sites of the promoter. Twenty 50-bp probes were first used to interact with equal amounts of nuclear extracts from leaf, root or young panicle/stem mixture. Gel mobility shift assays showed that probes D54O-10, D54O-22, D54O-24 and D54O-26 bound intensely to all the proteins from leaf, young panicle/stem mixture and root; D54O-4, D54O-6, D54O-12, D54O-16, D54O-20 and D54O-44 bound only to the proteins from leaf and young panicle/stem mixture, but not root; D54O-42 bound only to the proteins from leaf and root, but not young panicle/stem mixture; D54O-28, D54O-30, D54O-38 and D54O-40 bound only to the proteins from young panicle/stem mixture, but not leaf and root (Figure S2a, see ‘Supplementary material’). The other five 50-bp probes showed no distinct binding to the proteins from any tissue. Thirty-eight double-stranded short probes of 25–29 bp and 12–13 bp (Table S2) were subsequently used to further define the tissue-specific elements in the promoter regions covered by the probes D54O-6, D54O-10, D54O-16, D54O-20, D54O-22, D54O-24, D54O-26, D54O-28, D54O-30 and D54O-38 (Figure S2b,c). Gel mobility shift assays showed two unreported tissue-specific cis -elements, designated as LPSE1 (leaf and panicle/stem-specific element 1), represented by probe D54O-6-3-1 (5′-ATTGAGCTGCC located at –96 to –106), and PSE1 (panicle/stem-specific element 1), represented by probe D54O-30-3-1 (5′-TTTATCTATTTCC located at –696 to –708). LPSE1 bound specifically to proteins from leaf and young panicle/stem, and PSE1 bound specifically to proteins from young panicle/stem ( Figure 3 ). Three novel elements, represented by probes D54O-24-3-3, D54O-22-3-1 and D54O-26-1-1, were also identified, although they showed tissue-nonspecific binding to proteins from the tissues examined in the present experiments ( Figure 3 ). D54O-24-3-3 (5′-TTAGATAATGGA located at –534 to –545) was designated as LPSRE2 (leaf, panicle/stem and root element 2) and further studied. Four probes harbouring known tissue-specific cis -elements were also identified, but showed different tissue-specific activities than the harboured elements. Three of the four probes, D54O-16-1-1 (5′-TTGAT ATATT TGT located at –371 to –383) harbouring cis -element ROOTMOTIFTAPOX1 (italic sites), D54O-28-1-3 (5′-T CTTT GGCAGAG located at –659 to –670) harbouring element DOFCOREZM (complementary sequence of italic sites) and D54O-38-3-1 (5′-GTCCAAA ACGTCA located at –896 to –908) harbouring element TGACGTVMAMY (complementary sequence of italic sites) (Figure S1), showed tissue-specific protein binding ( Figure 3 ). D54O-16-1-1, designated as LPSE2, bound specifically to proteins from leaf and young panicle/stem, but not root. D54O-28-1-3 (designated as PSE2) and D54O-38-3-1 (designated as PSE3) bound specifically to proteins from young panicle/stem, but not leaf and root. Probe D54O-10-1-3 (5′-CGGCGC GCCAC G located at –209 to –220) harbouring element SORLIP1AT (italic sites) showed tissue-nonspecific binding to proteins from the tissues examined in the present experiments ( Figure 3 ); it was designated as LPSRE1. <h2>Tissue-specific cis -elements functioning negatively or positively on gene expression</h2> Six site-deleted promoters were constructed and fused with the GUS gene to study the effects of the identified cis -elements, LPSE1, LPSRE1, LPSE2, LPSRE2, PSE2 and PSE1, on gene expression. P D54O-LPSE1 , P D54O-LPSRE1 , P D54O-LPSE2 , P D54O-LPSRE2 , P D54O-PSE2 and P D54O-PSE1 were mutated promoters with partial or complete deletion of LPSE1, LPSRE1, LPSE2, LPSRE2, PSE2 and PSE1 elements, respectively (Table S3, see ‘Supplementary material’). GUS expression levels driven by all the mutated promoters were significantly influenced in at least one tissue examined ( Figure 4 ). Compared with plants carrying P D54O :GUS, deletion of LPSE1 reduced GUS expression by 39% and 54% in leaf and young panicle, respectively. In P D54O-LPSRE1 plants, the GUS level was 38% lower in leaf and 2.7-fold higher in stem. Deletion of LPSE2 reduced GUS expression by 44% in leaf and increased it by 7.8- and 1.4-fold in root and young panicle, respectively. The GUS levels were 1.3-, 12.8-, 6.2- and 7.6-fold higher in leaf, root, young panicle and stem, respectively, in plants carrying P D54O-LPSRE2 . Deletion of PSE2 decreased GUS activity by 26.8% in leaf and increased it by 3.3-fold in young panicle and stem. GUS expression was six-fold higher in young panicle and 5.8-fold higher in stem in P D54O-PSE1 plants. These results suggest that all six cis -elements are tissue-specific regulators; LPSE1 functions only as a positive regulator and LPSRE2 and PSE1 only as negative regulators, whereas LPSRE1, LPSE2 and PSE2 function as positive regulators in one tissue but negative regulators in another. <h2>Application of P D54O and fragment of P D54O harbouring tissue-specific cis -elements</h2> P D54O was used to regulate the expression of the Bt gene cry1Ac in rice variety Mudanjiang 8. Twenty independent positive transgenic plants were obtained. All the transgenic plants were highly resistant to leaf-folders under natural infestation conditions in the field compared with the wild-type ( Figure 5a ). Seven transgenic plants showed no infestation on the leaves, eight plants had only one leaf infested, four plants had two leaves infested, and one plant had three leaves infested. In contrast, each wild-type plant had, on average, eight leaves infested. The level of resistance in the transgenic plants positively correlated with the level of Cry1Ac in the leaves, but not the copy number of transgenic plants ( Table 1 ). Plants containing over 3.0 µg/g fresh leaf of Cry1Ac showed a superior resistance effect to leaf-folders and, at an expression level of Cry1Ac of 4.28 µg/g fresh leaf, plants were completely resistant to leaf-folders. No Cry1Ac was detected in the endosperm/embryo analysed by enzyme-linked immunosorbent assay (ELISA). The P D54O DNA fragments D54O-S (–1439 to –924 bp), D54O-P (–944 to –524 bp) and D54O-S/P (–1439 to –524 bp), which contained negative cis -regulatory elements based on analyses using truncated promoters, were fused with the constitutive 35S promoter ( Figure 1d ). Compared with plants carrying P 35S :GUS , P D54O-S+35S :GUS plants showed 57% lower GUS activity in stem and P D54O-P+35S :GUS plants showed 78% lower GUS activity in young panicle, although GUS activity was also reduced by 12%–41% in other tissues ( Figure 5b ). The GUS activity in plants carrying P D54O-S/P+35S :GUS was significantly lower than in P 35S :GUS plants: 79%, 74%, 91% and 96% in leaf, stem, young panicle and root, respectively. These results suggest that D54O-S carries strong negative element(s) that repress gene expression in stem, and D54O-P carries strong negative element(s) that repress gene expression in panicle. These two DNA fragments also harbour weak negative elements that repress gene expression in leaf, stem, panicle and/or root. <h1>Discussion</h1> D54O is a putative photosynthesis-related gene with a tissue-specific expression pattern. It is expressed largely in the green tissues of rice, which fits its putative role in photosynthesis. However, the molecular mechanisms of tissue-specific gene expression are still unclear. In this study, we identified seven tissue-specific cis -elements in P D54O , indicating that the tissue-specific expression of D54O is regulated by multiple cis -elements. These results suggest that P D54O can be used as an effective transgenic tool to engineer the tissue-specific expression of genes in rice breeding programmes and can provide candidates for further in-depth analysis of the cis -elements and trans -factors regulating tissue-specific expression. Characterization of the D54O promoter also provides candidates to study the factors that function directly upstream of D54O , which may help to unravel the molecular regulation of photosynthesis in rice. <h2>P D54O contains multiple novel tissue-specific positive and negative cis -elements</h2> P D54O is highly expressed in leaf and sheath, as determined by the GUS activity under its regulation in transgenic plants, and shows similar expression levels in these two tissues. P D54O expression in leaf was approximately 14-, 90-, 14- and 129-fold higher than that in stem, endosperm/embryo, young panicle and root, respectively. Thus, P D54O can be considered as a green tissue-specific promoter. The tissue-specific expression of P D54O is regulated by both positive and negative regulatory cis -elements. Comparative analysis of the functions of native and truncated promoters indicated that P D54O contains cis -elements which both positively and negatively regulate gene expression in leaf, sheath, young panicle, stem and root. The site-deleted promoter analyses supported this finding. LPSE1 activates gene expression in leaf and young panicle. LPSRE2 suppresses gene expression in leaf, root, young panicle and stem, and PSE1 suppresses gene expression in young panicle and stem. LPSRE1, LPSE2 and PSE2 appear to have dual roles in the regulation of tissue-specific gene expression. These three elements all functioned as positive elements in leaf, but LPSRE1 functioned as a negative element in stem, LPSE2 as a negative element in young panicle and root, and PSE2 as a negative element in young panicle and stem. From a consideration of the results of site-directed deletion assays and DNA protein binding assays, it appears that these cis -elements may function differently in terms of the requirement of tissue-specific protein binding for the regulation of gene expression. The tissue-specific function of these elements required protein binding in most cases examined. However, the negative role of LPSE2 in root and the positive role of PSE2 in leaf do not appear to require protein binding, because no root- or leaf-specific protein binding to the two elements was detected. This may be a result of the deletion of the elements influencing protein binding to other cis -elements ( Sakamoto et al ., 2001 ), which results in suppression or activation of gene expression. LPSE1, PSE1 and LPSRE2 are newly reported tissue-specific cis -elements. Although LPSRE1 harbours the known root-specific cis -element SORLIP1AT ( Hudson and Quail, 2003 ), it did not regulate gene expression in root, indicating that the functioning site is not SORLIP1AT. Thus, LPSRE1 is also a novel tissue-specific element. LPSE2 harbours element ROOTMOTIFTAPOX1, which was identified in Agrobacterium and can activate root-specific gene expression in tobacco ( Elmayan and Tepfer, 1995 ). However, LPSE2 functioned as an activator in leaf and as a repressor in root and young panicle, suggesting that the functioning site is not ROOTMOTIFTAPOX1 and that LPSE2 is also a novel tissue-specific element. PSE2 harbours element DOFCOREZM for binding Dof protein, which positively regulates endosperm-specific gene expression ( Vicente-Carbajosa et al ., 1997 ; Yanagisawa and Schmidt, 1999 ). Although further study is needed to determine whether the function of PSE2 is caused by DOFCOREZM, the present results suggest that PSE2 functions in leaf, stem and young panicle as both an activator and repressor. PSE3 harbours element TGACGTVMAMY for cotyledon-specific expression in Vigna mungo ( Yamauchi, 2001 ); the cotyledon of V. mungo corresponds to the endosperm of the monocotyledon Oryza sativa . Our results show that PSE3 binds only to the mixture of proteins from young panicle and stem, suggesting that the functional site of PSE3 may be TGACGTVMAMY. <h2>Potential use of P D54O , mutated P D54O and tissue-specific elements in rice breeding programmes</h2> Leaf-folders and stemborers are two major lepidopteran pests in rice that cause severe yield loss. However, researchers have yet to identify rice genes or genes from other plants that effectively protect rice from infestation by these pests. A series of Bt genes are considered to be the most effective genes for breeding pest-resistant rice ( High et al ., 2004 ), but Bt proteins have always been a concern to the public, in spite of the fact that scientific research has confirmed that these proteins are safe to humans. Using tissue-specific promoters to breed crops with Bt-free fruit and/or seed may at least partly solve this problem. P D54O is expressed largely in leaf, the invasion site of leaf-folders ( Khan et al ., 1988 ; Pathak and Khan, 1994 ), and our study confirmed P D54O to be a potential tool for breeding leaf-folder resistance in rice. Likewise, stemborers invade plants at the stem ( Teng and Revilla, 1996 ). The truncated P D54O-544 , which is highly expressed in leaf, sheath and stem, but not in seed, could be used to breed rice that is resistant to stemborers and leaf-folders. In addition, leaf, sheath and panicle are the major invasion sites of Rhizoctonia solani , Magnaporthe grisea and Xanthomonas oryzae pv. oryzae , which cause sheath blight, blast and bacterial blight, respectively, the three most devastating rice diseases. Thus, P D54O-544 may also be a valuable tool for breeding disease-resistant rice. Our results and those of other studies ( Roush, 1998 ; Bates et al ., 2005 ) have shown that Bt genes act in a dosage-dependent manner: plants that express high levels of Bt are more resistant to stemborers and leaf-folders than those that express low levels of Bt . In addition, abiotic and biotic defence-related genes frequently have dosage effects ( Hu et al ., 2006 ; Qiu et al ., 2007 ). Thus, constitutive strong promoters are frequently used to regulate the expression of these genes. The average expression level of 35S , which is a strong promoter in rice leaf tissue, is 2.3-fold higher than the expression level of P D54O . However, the 35S promoter is constitutively expressed in all tissues, including the fruit of plants ( Battraw and Hall, 1990 ). Our results suggest that a fragment of P D54O (D54O-P harbouring elements PSE1, PSE2, PSE3 and LPSRE2) can be used in combination with 35S to direct tissue-specific gene expression. This fragment suppressed approximately 78% of 35S -regulated gene expression in panicle, but had no significant influence on gene expression in stem and root, and only reduced 35S -regulated gene expression in leaf by approximately 34%. Furthermore, P D54O-544 can also be used to regulate dosage-dependent genes when breeding rice with transgene-free seed. The activity of this truncated promoter was 1.7–59.4-fold higher in leaf, sheath, young panicle, stem and root than the native promoter, but showed no activity in endosperm and embryo ( Figure 2b ). Over-expression of a dosage-dependent gene using a constitutive promoter frequently causes fitness costs, resulting in transgenic plants with abnormal morphology or decreased fertility ( Sinha et al ., 1993 ; Matzke et al ., 2000 ; Kurek et al ., 2002 ; Cheon et al ., 2004 ; Xu et al ., 2006 ). Thus, over-expression of the target gene only in required tissue(s) is one way to solve the problem of fitness costs in breeding programmes. Previous studies have reported that the same cis -element connected in tandem can regulate the function of the 35S promoter ( Eagle et al ., 1994 ; Qi et al ., 2007 ). The various tissue-specific negative regulatory cis -elements identified in the present study provide new choices for altering constitutive promoter-regulated gene expression. <h1>Experimental procedures</h1> <h2>cDNA array and RNA gel blot analyses</h2> The preparation of cDNA arrays, array hybridization and identification of differentially expressed genes were performed as described previously ( Zhou et al ., 2002 ). Aliquots (20 µg) of total RNA were used for RNA gel blot analysis as described previously ( Zhou et al ., 2002 ). <h2>Sequence analysis</h2> The promoter region of D54O was predicted using the computer programs TSSP, provided at the Softberry website ( http://www.softberry.com ), and PROSCAN ( http://bimas.dcrt.nih.gov/molbio/proscan ). Regulatory elements in the promoter region were analysed using the computer program PLACE (a database of plant cis -acting regulatory DNA elements) from Signal Scan ( http://www.dna.affrc.go.jp/PLACE/signalscan.html ). <h2>Isolation of promoter P D54O and construction of 5′-truncated promoters and P D54O :cry1Ac</h2> A bacterial artificial chromosome clone, 119H12, from rice cultivar Nongken 58 ( Oryza sativa ssp. japonica ) contained the D54O gene. The promoter of D54O was obtained by digestion of 119H12 using restriction enzymes Sph I and Sma I, and ligated to vector pCAMBIA1381 to drive the expression of the reporter gene GUS ( Figure 1c ), and vector pUC19 to form an intermediate vector. The primers 119h12-2 (5′-AT GAATTC ACGTTAGCATTATCCGAGAC-3′), 119H12-3 (5′-TTA GAATTC TATTCCAGTTTCTACATAGA-3′), 119H12-4 (5′-AGA GAATTC CAGTGTCCGAGTCTTAAATA-3′), 119H12-5 (5′-GTC GAATTC ATCTATTACCTAATTTTGGA-3′), 119H12-6 (5′-AA GAATTC GAGGATAACGTTCTGGCACT-3′) and 119H12-7 (5′-CA GAATTC CAATTATATTGAGCTGCCAT-3′) were specific to P D54O and contained the digestion site (italic) of restriction enzyme Hin dIII. Each primer was used in combination with the 3′-anchored primer pucF (5′-GTTTTCCCAGTCACGACGTTG-3′) from the multicloning sites of pUC19 for polymerase chain reaction (PCR) amplification of 5′-truncated P D54O employing the intermediate vector containing P D54O as template. The PCR products were digested with Hin dIII and Eco RI and ligated to vector pCAMBIA1381 to form truncated promoter and GUS fusions ( Figure 1c ). P D54O was introduced into a modified pCAMBIA1381 vector in which the GUS reporter gene was deleted. The cry1Ac gene (kindly provided by Dr Illimar Altosaar, University of Ottawa, Ottawa, ON, Canada) was ligated with the modified pCAMBIA1381 under the control of P D54O (Figure S3). <h2>Plant transformation</h2> The promoter: GUS constructs were transferred into Agrobacterium tumefaciens strain EHA105 by electroporation. These constructs were then transferred into rice variety Mudanjiang 8 ( O. sativa ssp. japonica ) by Agrobacterium- mediated transformation ( Lin and Zhang, 2005 ). The positive transformants were selected by PCR using the GUS -specific primers GusF (5′-CCAGGCAGTTTTAACGATCAGTTCGC-3′) and GusR (5′-GAGTGAAGATCCCTTTCTTGTTACC). <h2>Histochemical and fluorometric analysis of GUS activity</h2> Histochemical staining of GUS activity was conducted as described previously ( Wu et al ., 2003 ). Quantitative analyses of GUS activity were performed essentially as reported by Jefferson (1987 ). In brief, tissues were suspended in a grinding buffer [50 m m NaHPO 4 at pH 7.0, 10 m m ethylenediaminetetraacetic acid (EDTA), 0.1% sodium laurylsarcosine, 0.1% Triton X-100 and 10 m m ॆ-mercaptoethanol] and homogenized for 2 min with a minishaker homogenizer. The mixture was centrifuged to collect the supernatant. The total protein concentration in the supernatant was quantified using the Bradford assay ( Bradford, 1976 ). GUS protein in the supernatant was determined fluorometrically with a DyNA QUANT 200 Fluorometer (Hoefer Pharmacia Biotech Inc., San Francisco, CA, USA). GUS activity was determined fluorometrically by measuring the amount of 4-methylumbelliferone (Mu) produced under the catalysis of GUS in 1 mg of total protein per minute. <h2>Gel mobility shift assay</h2> Rice tissues from Mudanjiang 8 were harvested at the booting stage to isolate the nuclear extract using the method reported by Qiu et al . (2007 ). The gel mobility shift assay was applied essentially as described previously ( Urao et al ., 1993 ). DNA probes (2 pmol) were obtained by annealing of two complementary single-stranded DNA fragments and labelled with 20 µCi [α 32 P]dCTP using the Klenow fragment (DNA polymerase I) in a 20-µL reaction volume. The binding reaction was performed in a binding buffer [10 m m tris(hydroxymethyl)aminomethane (Tris) at pH 7.5, 50 m m NaCl, 1 m m dithiothreitol (DTT), 1 m m EDTA, 5% glycerol and 1 m m MgCl 2 ] with 3 µg poly(dI-dC), 10 µg of nuclear extract and 5000 cpm labelled probe for about 20 min at room temperature. The samples were then loaded on to a native gel of 5% arcylamide/bis-acrylamide in a running buffer (380 m m glycine, 2 m m EDTA and 25 m m Tris) and run for about 1 h. After electrophoresis, the gel was dried on Whatman paper and kept in the dark for autoradiography overnight. <h2>Site-directed mutagenesis</h2> Site-directed mutagenesis was performed using a GeneTailor Site-Directed Mutagenesis System (Invitrogen Life Technologies, Carlsbad, CA, USA). In brief, the pUC19 plasmid containing P D54O , used as the PCR template, was methylated before use. A mutagenic primer pair (Table S3), in which the putative regulatory elements were deleted and a restriction enzyme site was introduced to the primer pair for facility of identification, was used to amplify the promoter containing the target mutation. The PCR product was transferred into a special host Escherichia coli strain DH5α-T1, in which the methylated plasmid could not replicate, to help to select the mutated construct. <h2>Quantification of the Cry1Ac protein</h2> The amount of Cry1Ac protein in the leaves and endosperm/embryo of transgenic plants was measured by ELISA using the QuantiPlate Kit for Cry1Ab/Cry1Ac (Envirologix Inc., Portland, ME, USA) according to the manufacturer's instructions. <h2>Fusion promoters</h2> Three DNA fragments harbouring negative cis -elements regulating tissue-specific expression in P D54O were obtained by PCR amplification using primer pairs 119H12-2/D54O-2 (5′-CGAAGCTTAAGATACATCTTTGACCTTA-3′), 119H12-3/D54O-3 (5′-ATAAGCTTCAGGTTTTAATCCATTATCT-3′) and 119H12-2/D54O-3. Fragments D54O-S (–1439 to –924 bp), D54O-P (–944 to –524 bp) and D54O-S/P (–1439 to –524 bp) negatively regulate gene expression in stem, panicle and both stem and panicle, respectively. After digestion with Hin dIII and Eco RI, the three fragments were ligated to vector pCAMBIA1301 and fused with the 35S promoter. The fusion promoters were designated as P D54O-S+35S , P D54O-P+35S and P D54O-S/P+35S ( Figure 1d ). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant Biotechnology Journal Wiley

A rice promoter containing both novel positive and negative cis -elements for regulation of green tissue-specific gene expression in transgenic plants

Loading next page...
 
/lp/wiley/a-rice-promoter-containing-both-novel-positive-and-negative-cis-FQ5LivBxZT

References (47)

Publisher
Wiley
Copyright
© 2007 The Authors Journal compilation © 2007 Blackwell Publishing Ltd
ISSN
1467-7644
eISSN
1467-7652
DOI
10.1111/j.1467-7652.2007.00271.x
pmid
17596180
Publisher site
See Article on Publisher Site

Abstract

<h1>Introduction</h1> Several genes from non-plant or non-crop species have been demonstrated to be useful for crop improvement. For example, the Bt gene from Bacillus thuringiensis , encoding endotoxin, has been used to breed insect-resistant cotton, maize, potato and rice ( Tu et al ., 2000 ; Nester et al ., 2002 ). The Bar gene from Streptomyces hygroscopicus , encoding phosphinothricin acetyltransferase ( Block et al ., 1987 ), and the EPSPS gene from Agrobacterium tumefaciens , encoding 5-enolpyruvylshikimate-3-phosphate synthase ( Funke et al ., 2006 ), have been used to breed herbicide-resistant soybean, cotton, canola and maize ( Duke, 2005 ; Pline-Srnic, 2005 ). The seed albumin gene from Amaranthus hypochondriacus has been used to improve the nutritive value of potato ( Chakraborty et al ., 2000 ). A cold-tolerant pyruvate orthophosphate dikinase gene from Flaveria brownii has been introduced into maize to improve cold tolerance ( Ohta et al ., 2004 ), and the Arabidopsis DREM1A gene has been shown to improve the drought tolerance of wheat ( Pellegrineschi et al ., 2004 ). Although some of these transgenic crops carrying genes from non-plants or non-crops are commercially available, in many countries and areas public concern regarding the food safety of genetically modified organisms has limited the application of transgenic crops. Using tissue-specific promoters to avoid the expression of foreign genes in seed or fruit may help to relieve this public concern. Many of the genes that can be used for crop improvement, including those from plants and other organisms, act in a dosage-dependent manner, and plants expressing high levels of this type of gene have a better modified phenotype. For instance, the function of Bt is dosage dependent ( Roush, 1998 ; Bates et al ., 2005 ). Many plant abiotic- and biotic-responsive genes are also dosage dependent ( Hu et al ., 2006 ; Qiu et al ., 2007 ). Thus, strong constitutive expression promoters, such as the cauliflower mosaic virus 35S promoter and the maize ubiquitin gene promoter ( Odell et al ., 1985 ; Cornejo et al ., 1993 ), have been used widely to express the foreign genes in transgenic plants. However, constitutive expression of a foreign gene may be harmful to the host plant, causing sterility, retarded development, abnormal morphology, yield penalty, altered grain composition or transgene silencing ( Sinha et al ., 1993 ; Matzke et al ., 2000 ; Kurek et al ., 2002 ; Cheon et al ., 2004 ; Xu et al ., 2006 ). Using a strong tissue-specific or inducible promoter to restrict gene expression to only the required tissue or at a particular time may solve this type of problem. Although a large number of tissue-specific promoters have been identified, strong tissue-specific promoters that can be used for various crop improvements are limited. Some cis -acting elements regulating tissue-specific gene expression have been identified. The ACGTROOT1, ROOTMOTIFTAPOX1, WUSATAg, OSE1ROOTNODULE and OSE2ROOTNODULE elements promote root-specific gene expression ( Salinas et al ., 1992 ; Elmayan and Tepfer, 1995 ; Kamiya et al ., 2003 ; Vieweg et al ., 2004 ). DPBFCOREDCDC3, DOFCOREZM, TGACGTVMAMY and ESP activate embryo-, endosperm- or cotyledon-specific gene expression ( Kim et al ., 1997 ; Vicente-Carbajosa et al ., 1997 ; Yanagisawa and Schmidt, 1999 ; Yamauchi, 2001 ; Vickers et al ., 2006 ). CANBNNAP functions as a repressor in embryo-specific gene expression and as an activator in endosperm- and leaf-specific gene expression ( Ellerstrom et al ., 1996 ). POLLEN1LELAT52 promotes pollen-specific gene expression ( Bate and Twell, 1998 ), and RGATAOS activates phloem-specific gene expression ( Yin et al ., 1997 ). An AC-rich element stimulates gene expression in xylem tissue ( Lauvergeat et al ., 2002 ). An element in the –199 to –186 region of the bean GRP 1.8 gene promoter represses gene expression in vascular tissue ( Keller and Baumgartner, 1991 ). RYREPEATBNNAPA and TGTCACACMCUCUMISIN are activators for seed- and fruit-specific gene expression, respectively ( Ezcurra et al ., 1999 ; Yamagata et al ., 2002 ). The regulatory mechanism of tissue-specific gene expression remains unclear, however, and information concerning cis -regulatory elements is still limited, and cannot explain the regulation of various types of tissue-specific gene expression. In this study, we isolated and characterized the rice green tissue-specific promoter P D54O . Five novel tissue-specific cis -acting elements were identified from P D54O . They functioned as activators, repressors, or both in the regulation of leaf-, stem-, root- and panicle-specific gene expression. Using P D54O , truncated P D54O , mutated P D54O or its cis -elements to regulate the cry1Ac gene or a reporter gene resulted in insect-resistant rice and transgenic rice with various tissue-specific expression patterns of the reporter gene, indicating that the original and modified promoters and their cis -elements may prove to be useful in rice breeding programmes. <h1>Results</h1> <h2>Identification and isolation of a promoter not expressed in seed</h2> Using complementary DNA (cDNA) array technology, we identified a cDNA that was not expressed in seed, but was expressed in other tissue ( Figure 1a ). Sequence analysis indicated that this cDNA corresponded to LOC_Os08g10020, according to the rice genome annotation of the Institute for Genomic Research ( http://rice.tigr.org ), encoding a photosystem II 10-kDa polypeptide. This gene was named D54O . Further analysis showed that D54O was expressed in leaf and sheath, but not in root, stem or panicle, at the grain-filling stage ( Figure 1b ). A 2.1-kb fragment harbouring the –2134 to +41 region (translation initiation site of D54O as +1) upstream of D54O was isolated from rice variety Nongken 58. Sequence analysis showed that this fragment contained the basal regulatory elements, TATA box for RNA polymerase binding and CAAT box for transcription frequency regulation (Table S1 and Figure S1, see ‘Supplementary material’); this fragment also harboured several known cis -elements regulating tissue-specific gene expression, including root, phloem, pollen, embryo, endosperm, cotyledon, seed and fruit (Table S1 and Figure S1). Thus, this fragment was considered to harbour the full promoter of D54O and was designated as P D54O . <h2>Identification of the tissue-specific promoter region</h2> P D54O and a series of truncated P D54O were used to regulate the expression of the reporter gene ॆ-glucuronidase ( GUS ) in rice variety Mudanjiang 8 ( Figure 1c ). In transgenic plants carrying P D54O :GUS , histochemical staining revealed GUS expression in leaf, sheath, ligule and lemma palea of the young panicle, but not in lemma palea of the mature panicle, endosperm, embryo, stem and root ( Figure 2a ). In addition to the expression in leaf, sheath and young panicle, quantitative analysis revealed GUS expression in stem, endosperm/embryo and root of P D54O :GUS plants, but the expression levels in the three tissues were only 7.4%, 1.1% and 0.8%, respectively, of the GUS expression in leaf tissue ( Figure 2b ). The expression levels of P D54O :GUS were 44.5%, 52%, 5%, 5%, 1.2% and 1.5% of those of P 35S :GUS in leaf, sheath, young panicle, stem, root and endosperm/embryo, respectively ( Figure 2b ). GUS expression levels in transgenic plants carrying a single copy of P D54O :GUS or truncated P D54O :GUS constructs ( P D54O-1439 :GUS , P D54O-944 :GUS , P D54O-723 :GUS , P D54O-544 :GUS , P D54O-264 :GUS or P D54O-61 :GUS ; Figure 1c ) were compared by quantitative analysis. Plants carrying P D54O-1439 :GUS showed a similar GUS expression pattern to plants carrying P D54O :GUS , except that the GUS levels in the former were 22%, 21% and 24% lower in leaf, sheath and young panicle, respectively ( Figure 2b ). Compared with P D54O-1439 :GUS plants, GUS activities in plants carrying P D54O-944 :GUS were approximately six- and five-fold higher in stem and root, respectively, but showed no significant difference in other tissues. Compared with plants carrying P D54O-944 :GUS , GUS activity in P D54O-723 :GUS plants was approximately 29% lower in leaf and sheath, showed no significant difference in stem and endosperm/embryo, and was 8.5- and 4.1-fold higher in young panicle and root, respectively. Plants carrying P D54O-544 :GUS had 3.5-, 3.7-, 2.8-, 1.8- and 5.0-fold higher GUS activities compared with plants carrying P D54O-723 :GUS , and 1.7-, 1.8-, 18.3-, 7.6- and 59.4-fold higher GUS activities compared with plants carrying P D54O :GUS in leaf, sheath, young panicle, stem and root, respectively, but showed no GUS activity in endosperm/embryo. No GUS activity was detected in endosperm/embryo in P D54O-264 :GUS or P D54O-61 :GUS plants ( Figure 2b ). Compared with plants carrying P D54O :GUS , the GUS levels in P D54O-264 :GUS plants were 70%, 71%, 59% and 72% lower in leaf, sheath, young panicle and stem, respectively, and 15-fold higher in root. GUS activity was nearly undetectable in histochemical staining assays of P D54O-61 :GUS plants, which showed 95%, 95%, 87% and 92% lower GUS activities in leaf, sheath, young panicle and stem, respectively, but no significant difference in root, compared with plants carrying P D54O :GUS . These results suggest that the –2134 to –545 region of P D54O contains tissue-specific cis -elements that positively regulate gene expression in leaf, sheath and endosperm/embryo and negatively regulate gene expression in leaf, sheath, young panicle, stem and root. In addition, the –544 to –62 region contains cis -elements that positively regulate gene expression in leaf, sheath, young panicle and stem, and the –264 to –62 region contains cis -elements that negatively regulate gene expression in root. The positive regulatory elements in P D54O activated gene expression only slightly in stem, root and endosperm/embryo. <h2>Identification of cis -elements controlling tissue-specific expression</h2> Fifty-eight double-stranded DNA probes (Table S2, see ‘Supplementary material’) covering the different regions of P D54O were used to study DNA–protein binding and to determine the tissue-specific sites of the promoter. Twenty 50-bp probes were first used to interact with equal amounts of nuclear extracts from leaf, root or young panicle/stem mixture. Gel mobility shift assays showed that probes D54O-10, D54O-22, D54O-24 and D54O-26 bound intensely to all the proteins from leaf, young panicle/stem mixture and root; D54O-4, D54O-6, D54O-12, D54O-16, D54O-20 and D54O-44 bound only to the proteins from leaf and young panicle/stem mixture, but not root; D54O-42 bound only to the proteins from leaf and root, but not young panicle/stem mixture; D54O-28, D54O-30, D54O-38 and D54O-40 bound only to the proteins from young panicle/stem mixture, but not leaf and root (Figure S2a, see ‘Supplementary material’). The other five 50-bp probes showed no distinct binding to the proteins from any tissue. Thirty-eight double-stranded short probes of 25–29 bp and 12–13 bp (Table S2) were subsequently used to further define the tissue-specific elements in the promoter regions covered by the probes D54O-6, D54O-10, D54O-16, D54O-20, D54O-22, D54O-24, D54O-26, D54O-28, D54O-30 and D54O-38 (Figure S2b,c). Gel mobility shift assays showed two unreported tissue-specific cis -elements, designated as LPSE1 (leaf and panicle/stem-specific element 1), represented by probe D54O-6-3-1 (5′-ATTGAGCTGCC located at –96 to –106), and PSE1 (panicle/stem-specific element 1), represented by probe D54O-30-3-1 (5′-TTTATCTATTTCC located at –696 to –708). LPSE1 bound specifically to proteins from leaf and young panicle/stem, and PSE1 bound specifically to proteins from young panicle/stem ( Figure 3 ). Three novel elements, represented by probes D54O-24-3-3, D54O-22-3-1 and D54O-26-1-1, were also identified, although they showed tissue-nonspecific binding to proteins from the tissues examined in the present experiments ( Figure 3 ). D54O-24-3-3 (5′-TTAGATAATGGA located at –534 to –545) was designated as LPSRE2 (leaf, panicle/stem and root element 2) and further studied. Four probes harbouring known tissue-specific cis -elements were also identified, but showed different tissue-specific activities than the harboured elements. Three of the four probes, D54O-16-1-1 (5′-TTGAT ATATT TGT located at –371 to –383) harbouring cis -element ROOTMOTIFTAPOX1 (italic sites), D54O-28-1-3 (5′-T CTTT GGCAGAG located at –659 to –670) harbouring element DOFCOREZM (complementary sequence of italic sites) and D54O-38-3-1 (5′-GTCCAAA ACGTCA located at –896 to –908) harbouring element TGACGTVMAMY (complementary sequence of italic sites) (Figure S1), showed tissue-specific protein binding ( Figure 3 ). D54O-16-1-1, designated as LPSE2, bound specifically to proteins from leaf and young panicle/stem, but not root. D54O-28-1-3 (designated as PSE2) and D54O-38-3-1 (designated as PSE3) bound specifically to proteins from young panicle/stem, but not leaf and root. Probe D54O-10-1-3 (5′-CGGCGC GCCAC G located at –209 to –220) harbouring element SORLIP1AT (italic sites) showed tissue-nonspecific binding to proteins from the tissues examined in the present experiments ( Figure 3 ); it was designated as LPSRE1. <h2>Tissue-specific cis -elements functioning negatively or positively on gene expression</h2> Six site-deleted promoters were constructed and fused with the GUS gene to study the effects of the identified cis -elements, LPSE1, LPSRE1, LPSE2, LPSRE2, PSE2 and PSE1, on gene expression. P D54O-LPSE1 , P D54O-LPSRE1 , P D54O-LPSE2 , P D54O-LPSRE2 , P D54O-PSE2 and P D54O-PSE1 were mutated promoters with partial or complete deletion of LPSE1, LPSRE1, LPSE2, LPSRE2, PSE2 and PSE1 elements, respectively (Table S3, see ‘Supplementary material’). GUS expression levels driven by all the mutated promoters were significantly influenced in at least one tissue examined ( Figure 4 ). Compared with plants carrying P D54O :GUS, deletion of LPSE1 reduced GUS expression by 39% and 54% in leaf and young panicle, respectively. In P D54O-LPSRE1 plants, the GUS level was 38% lower in leaf and 2.7-fold higher in stem. Deletion of LPSE2 reduced GUS expression by 44% in leaf and increased it by 7.8- and 1.4-fold in root and young panicle, respectively. The GUS levels were 1.3-, 12.8-, 6.2- and 7.6-fold higher in leaf, root, young panicle and stem, respectively, in plants carrying P D54O-LPSRE2 . Deletion of PSE2 decreased GUS activity by 26.8% in leaf and increased it by 3.3-fold in young panicle and stem. GUS expression was six-fold higher in young panicle and 5.8-fold higher in stem in P D54O-PSE1 plants. These results suggest that all six cis -elements are tissue-specific regulators; LPSE1 functions only as a positive regulator and LPSRE2 and PSE1 only as negative regulators, whereas LPSRE1, LPSE2 and PSE2 function as positive regulators in one tissue but negative regulators in another. <h2>Application of P D54O and fragment of P D54O harbouring tissue-specific cis -elements</h2> P D54O was used to regulate the expression of the Bt gene cry1Ac in rice variety Mudanjiang 8. Twenty independent positive transgenic plants were obtained. All the transgenic plants were highly resistant to leaf-folders under natural infestation conditions in the field compared with the wild-type ( Figure 5a ). Seven transgenic plants showed no infestation on the leaves, eight plants had only one leaf infested, four plants had two leaves infested, and one plant had three leaves infested. In contrast, each wild-type plant had, on average, eight leaves infested. The level of resistance in the transgenic plants positively correlated with the level of Cry1Ac in the leaves, but not the copy number of transgenic plants ( Table 1 ). Plants containing over 3.0 µg/g fresh leaf of Cry1Ac showed a superior resistance effect to leaf-folders and, at an expression level of Cry1Ac of 4.28 µg/g fresh leaf, plants were completely resistant to leaf-folders. No Cry1Ac was detected in the endosperm/embryo analysed by enzyme-linked immunosorbent assay (ELISA). The P D54O DNA fragments D54O-S (–1439 to –924 bp), D54O-P (–944 to –524 bp) and D54O-S/P (–1439 to –524 bp), which contained negative cis -regulatory elements based on analyses using truncated promoters, were fused with the constitutive 35S promoter ( Figure 1d ). Compared with plants carrying P 35S :GUS , P D54O-S+35S :GUS plants showed 57% lower GUS activity in stem and P D54O-P+35S :GUS plants showed 78% lower GUS activity in young panicle, although GUS activity was also reduced by 12%–41% in other tissues ( Figure 5b ). The GUS activity in plants carrying P D54O-S/P+35S :GUS was significantly lower than in P 35S :GUS plants: 79%, 74%, 91% and 96% in leaf, stem, young panicle and root, respectively. These results suggest that D54O-S carries strong negative element(s) that repress gene expression in stem, and D54O-P carries strong negative element(s) that repress gene expression in panicle. These two DNA fragments also harbour weak negative elements that repress gene expression in leaf, stem, panicle and/or root. <h1>Discussion</h1> D54O is a putative photosynthesis-related gene with a tissue-specific expression pattern. It is expressed largely in the green tissues of rice, which fits its putative role in photosynthesis. However, the molecular mechanisms of tissue-specific gene expression are still unclear. In this study, we identified seven tissue-specific cis -elements in P D54O , indicating that the tissue-specific expression of D54O is regulated by multiple cis -elements. These results suggest that P D54O can be used as an effective transgenic tool to engineer the tissue-specific expression of genes in rice breeding programmes and can provide candidates for further in-depth analysis of the cis -elements and trans -factors regulating tissue-specific expression. Characterization of the D54O promoter also provides candidates to study the factors that function directly upstream of D54O , which may help to unravel the molecular regulation of photosynthesis in rice. <h2>P D54O contains multiple novel tissue-specific positive and negative cis -elements</h2> P D54O is highly expressed in leaf and sheath, as determined by the GUS activity under its regulation in transgenic plants, and shows similar expression levels in these two tissues. P D54O expression in leaf was approximately 14-, 90-, 14- and 129-fold higher than that in stem, endosperm/embryo, young panicle and root, respectively. Thus, P D54O can be considered as a green tissue-specific promoter. The tissue-specific expression of P D54O is regulated by both positive and negative regulatory cis -elements. Comparative analysis of the functions of native and truncated promoters indicated that P D54O contains cis -elements which both positively and negatively regulate gene expression in leaf, sheath, young panicle, stem and root. The site-deleted promoter analyses supported this finding. LPSE1 activates gene expression in leaf and young panicle. LPSRE2 suppresses gene expression in leaf, root, young panicle and stem, and PSE1 suppresses gene expression in young panicle and stem. LPSRE1, LPSE2 and PSE2 appear to have dual roles in the regulation of tissue-specific gene expression. These three elements all functioned as positive elements in leaf, but LPSRE1 functioned as a negative element in stem, LPSE2 as a negative element in young panicle and root, and PSE2 as a negative element in young panicle and stem. From a consideration of the results of site-directed deletion assays and DNA protein binding assays, it appears that these cis -elements may function differently in terms of the requirement of tissue-specific protein binding for the regulation of gene expression. The tissue-specific function of these elements required protein binding in most cases examined. However, the negative role of LPSE2 in root and the positive role of PSE2 in leaf do not appear to require protein binding, because no root- or leaf-specific protein binding to the two elements was detected. This may be a result of the deletion of the elements influencing protein binding to other cis -elements ( Sakamoto et al ., 2001 ), which results in suppression or activation of gene expression. LPSE1, PSE1 and LPSRE2 are newly reported tissue-specific cis -elements. Although LPSRE1 harbours the known root-specific cis -element SORLIP1AT ( Hudson and Quail, 2003 ), it did not regulate gene expression in root, indicating that the functioning site is not SORLIP1AT. Thus, LPSRE1 is also a novel tissue-specific element. LPSE2 harbours element ROOTMOTIFTAPOX1, which was identified in Agrobacterium and can activate root-specific gene expression in tobacco ( Elmayan and Tepfer, 1995 ). However, LPSE2 functioned as an activator in leaf and as a repressor in root and young panicle, suggesting that the functioning site is not ROOTMOTIFTAPOX1 and that LPSE2 is also a novel tissue-specific element. PSE2 harbours element DOFCOREZM for binding Dof protein, which positively regulates endosperm-specific gene expression ( Vicente-Carbajosa et al ., 1997 ; Yanagisawa and Schmidt, 1999 ). Although further study is needed to determine whether the function of PSE2 is caused by DOFCOREZM, the present results suggest that PSE2 functions in leaf, stem and young panicle as both an activator and repressor. PSE3 harbours element TGACGTVMAMY for cotyledon-specific expression in Vigna mungo ( Yamauchi, 2001 ); the cotyledon of V. mungo corresponds to the endosperm of the monocotyledon Oryza sativa . Our results show that PSE3 binds only to the mixture of proteins from young panicle and stem, suggesting that the functional site of PSE3 may be TGACGTVMAMY. <h2>Potential use of P D54O , mutated P D54O and tissue-specific elements in rice breeding programmes</h2> Leaf-folders and stemborers are two major lepidopteran pests in rice that cause severe yield loss. However, researchers have yet to identify rice genes or genes from other plants that effectively protect rice from infestation by these pests. A series of Bt genes are considered to be the most effective genes for breeding pest-resistant rice ( High et al ., 2004 ), but Bt proteins have always been a concern to the public, in spite of the fact that scientific research has confirmed that these proteins are safe to humans. Using tissue-specific promoters to breed crops with Bt-free fruit and/or seed may at least partly solve this problem. P D54O is expressed largely in leaf, the invasion site of leaf-folders ( Khan et al ., 1988 ; Pathak and Khan, 1994 ), and our study confirmed P D54O to be a potential tool for breeding leaf-folder resistance in rice. Likewise, stemborers invade plants at the stem ( Teng and Revilla, 1996 ). The truncated P D54O-544 , which is highly expressed in leaf, sheath and stem, but not in seed, could be used to breed rice that is resistant to stemborers and leaf-folders. In addition, leaf, sheath and panicle are the major invasion sites of Rhizoctonia solani , Magnaporthe grisea and Xanthomonas oryzae pv. oryzae , which cause sheath blight, blast and bacterial blight, respectively, the three most devastating rice diseases. Thus, P D54O-544 may also be a valuable tool for breeding disease-resistant rice. Our results and those of other studies ( Roush, 1998 ; Bates et al ., 2005 ) have shown that Bt genes act in a dosage-dependent manner: plants that express high levels of Bt are more resistant to stemborers and leaf-folders than those that express low levels of Bt . In addition, abiotic and biotic defence-related genes frequently have dosage effects ( Hu et al ., 2006 ; Qiu et al ., 2007 ). Thus, constitutive strong promoters are frequently used to regulate the expression of these genes. The average expression level of 35S , which is a strong promoter in rice leaf tissue, is 2.3-fold higher than the expression level of P D54O . However, the 35S promoter is constitutively expressed in all tissues, including the fruit of plants ( Battraw and Hall, 1990 ). Our results suggest that a fragment of P D54O (D54O-P harbouring elements PSE1, PSE2, PSE3 and LPSRE2) can be used in combination with 35S to direct tissue-specific gene expression. This fragment suppressed approximately 78% of 35S -regulated gene expression in panicle, but had no significant influence on gene expression in stem and root, and only reduced 35S -regulated gene expression in leaf by approximately 34%. Furthermore, P D54O-544 can also be used to regulate dosage-dependent genes when breeding rice with transgene-free seed. The activity of this truncated promoter was 1.7–59.4-fold higher in leaf, sheath, young panicle, stem and root than the native promoter, but showed no activity in endosperm and embryo ( Figure 2b ). Over-expression of a dosage-dependent gene using a constitutive promoter frequently causes fitness costs, resulting in transgenic plants with abnormal morphology or decreased fertility ( Sinha et al ., 1993 ; Matzke et al ., 2000 ; Kurek et al ., 2002 ; Cheon et al ., 2004 ; Xu et al ., 2006 ). Thus, over-expression of the target gene only in required tissue(s) is one way to solve the problem of fitness costs in breeding programmes. Previous studies have reported that the same cis -element connected in tandem can regulate the function of the 35S promoter ( Eagle et al ., 1994 ; Qi et al ., 2007 ). The various tissue-specific negative regulatory cis -elements identified in the present study provide new choices for altering constitutive promoter-regulated gene expression. <h1>Experimental procedures</h1> <h2>cDNA array and RNA gel blot analyses</h2> The preparation of cDNA arrays, array hybridization and identification of differentially expressed genes were performed as described previously ( Zhou et al ., 2002 ). Aliquots (20 µg) of total RNA were used for RNA gel blot analysis as described previously ( Zhou et al ., 2002 ). <h2>Sequence analysis</h2> The promoter region of D54O was predicted using the computer programs TSSP, provided at the Softberry website ( http://www.softberry.com ), and PROSCAN ( http://bimas.dcrt.nih.gov/molbio/proscan ). Regulatory elements in the promoter region were analysed using the computer program PLACE (a database of plant cis -acting regulatory DNA elements) from Signal Scan ( http://www.dna.affrc.go.jp/PLACE/signalscan.html ). <h2>Isolation of promoter P D54O and construction of 5′-truncated promoters and P D54O :cry1Ac</h2> A bacterial artificial chromosome clone, 119H12, from rice cultivar Nongken 58 ( Oryza sativa ssp. japonica ) contained the D54O gene. The promoter of D54O was obtained by digestion of 119H12 using restriction enzymes Sph I and Sma I, and ligated to vector pCAMBIA1381 to drive the expression of the reporter gene GUS ( Figure 1c ), and vector pUC19 to form an intermediate vector. The primers 119h12-2 (5′-AT GAATTC ACGTTAGCATTATCCGAGAC-3′), 119H12-3 (5′-TTA GAATTC TATTCCAGTTTCTACATAGA-3′), 119H12-4 (5′-AGA GAATTC CAGTGTCCGAGTCTTAAATA-3′), 119H12-5 (5′-GTC GAATTC ATCTATTACCTAATTTTGGA-3′), 119H12-6 (5′-AA GAATTC GAGGATAACGTTCTGGCACT-3′) and 119H12-7 (5′-CA GAATTC CAATTATATTGAGCTGCCAT-3′) were specific to P D54O and contained the digestion site (italic) of restriction enzyme Hin dIII. Each primer was used in combination with the 3′-anchored primer pucF (5′-GTTTTCCCAGTCACGACGTTG-3′) from the multicloning sites of pUC19 for polymerase chain reaction (PCR) amplification of 5′-truncated P D54O employing the intermediate vector containing P D54O as template. The PCR products were digested with Hin dIII and Eco RI and ligated to vector pCAMBIA1381 to form truncated promoter and GUS fusions ( Figure 1c ). P D54O was introduced into a modified pCAMBIA1381 vector in which the GUS reporter gene was deleted. The cry1Ac gene (kindly provided by Dr Illimar Altosaar, University of Ottawa, Ottawa, ON, Canada) was ligated with the modified pCAMBIA1381 under the control of P D54O (Figure S3). <h2>Plant transformation</h2> The promoter: GUS constructs were transferred into Agrobacterium tumefaciens strain EHA105 by electroporation. These constructs were then transferred into rice variety Mudanjiang 8 ( O. sativa ssp. japonica ) by Agrobacterium- mediated transformation ( Lin and Zhang, 2005 ). The positive transformants were selected by PCR using the GUS -specific primers GusF (5′-CCAGGCAGTTTTAACGATCAGTTCGC-3′) and GusR (5′-GAGTGAAGATCCCTTTCTTGTTACC). <h2>Histochemical and fluorometric analysis of GUS activity</h2> Histochemical staining of GUS activity was conducted as described previously ( Wu et al ., 2003 ). Quantitative analyses of GUS activity were performed essentially as reported by Jefferson (1987 ). In brief, tissues were suspended in a grinding buffer [50 m m NaHPO 4 at pH 7.0, 10 m m ethylenediaminetetraacetic acid (EDTA), 0.1% sodium laurylsarcosine, 0.1% Triton X-100 and 10 m m ॆ-mercaptoethanol] and homogenized for 2 min with a minishaker homogenizer. The mixture was centrifuged to collect the supernatant. The total protein concentration in the supernatant was quantified using the Bradford assay ( Bradford, 1976 ). GUS protein in the supernatant was determined fluorometrically with a DyNA QUANT 200 Fluorometer (Hoefer Pharmacia Biotech Inc., San Francisco, CA, USA). GUS activity was determined fluorometrically by measuring the amount of 4-methylumbelliferone (Mu) produced under the catalysis of GUS in 1 mg of total protein per minute. <h2>Gel mobility shift assay</h2> Rice tissues from Mudanjiang 8 were harvested at the booting stage to isolate the nuclear extract using the method reported by Qiu et al . (2007 ). The gel mobility shift assay was applied essentially as described previously ( Urao et al ., 1993 ). DNA probes (2 pmol) were obtained by annealing of two complementary single-stranded DNA fragments and labelled with 20 µCi [α 32 P]dCTP using the Klenow fragment (DNA polymerase I) in a 20-µL reaction volume. The binding reaction was performed in a binding buffer [10 m m tris(hydroxymethyl)aminomethane (Tris) at pH 7.5, 50 m m NaCl, 1 m m dithiothreitol (DTT), 1 m m EDTA, 5% glycerol and 1 m m MgCl 2 ] with 3 µg poly(dI-dC), 10 µg of nuclear extract and 5000 cpm labelled probe for about 20 min at room temperature. The samples were then loaded on to a native gel of 5% arcylamide/bis-acrylamide in a running buffer (380 m m glycine, 2 m m EDTA and 25 m m Tris) and run for about 1 h. After electrophoresis, the gel was dried on Whatman paper and kept in the dark for autoradiography overnight. <h2>Site-directed mutagenesis</h2> Site-directed mutagenesis was performed using a GeneTailor Site-Directed Mutagenesis System (Invitrogen Life Technologies, Carlsbad, CA, USA). In brief, the pUC19 plasmid containing P D54O , used as the PCR template, was methylated before use. A mutagenic primer pair (Table S3), in which the putative regulatory elements were deleted and a restriction enzyme site was introduced to the primer pair for facility of identification, was used to amplify the promoter containing the target mutation. The PCR product was transferred into a special host Escherichia coli strain DH5α-T1, in which the methylated plasmid could not replicate, to help to select the mutated construct. <h2>Quantification of the Cry1Ac protein</h2> The amount of Cry1Ac protein in the leaves and endosperm/embryo of transgenic plants was measured by ELISA using the QuantiPlate Kit for Cry1Ab/Cry1Ac (Envirologix Inc., Portland, ME, USA) according to the manufacturer's instructions. <h2>Fusion promoters</h2> Three DNA fragments harbouring negative cis -elements regulating tissue-specific expression in P D54O were obtained by PCR amplification using primer pairs 119H12-2/D54O-2 (5′-CGAAGCTTAAGATACATCTTTGACCTTA-3′), 119H12-3/D54O-3 (5′-ATAAGCTTCAGGTTTTAATCCATTATCT-3′) and 119H12-2/D54O-3. Fragments D54O-S (–1439 to –924 bp), D54O-P (–944 to –524 bp) and D54O-S/P (–1439 to –524 bp) negatively regulate gene expression in stem, panicle and both stem and panicle, respectively. After digestion with Hin dIII and Eco RI, the three fragments were ligated to vector pCAMBIA1301 and fused with the 35S promoter. The fusion promoters were designated as P D54O-S+35S , P D54O-P+35S and P D54O-S/P+35S ( Figure 1d ).

Journal

Plant Biotechnology JournalWiley

Published: Sep 1, 2007

Keywords: Bt gene; cis -element; GUS gene; Oryza sativa ; promoter; tissue-specific expression.

There are no references for this article.