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

Learn More →

Evaluation of tissue specificity and expression strength of rice seed component gene promoters in transgenic rice

Evaluation of tissue specificity and expression strength of rice seed component gene promoters in... <h1>Introduction</h1> Plant biotechnology has been applied not only to the improvement of agronomic properties, such as pest/pathogen resistance, abiotic stress resistance, yield and quality, but also to the production of pharmaceutical and industrial materials. Several recombinant industrial enzymes and pharmaceuticals have been commercialized already, and the number is expected to increase in the near future ( Hood and Jilka, 1999 ; Hood and Woodard, 2002 ). Plant production of recombinant proteins has many advantages, the most prominent of which are cost savings relative to animal transgenic systems, easy control of production scale in response to market size and freedom from animal-derived pathogens, including viruses and prions (Fischer and Emans, 2000; Giddings et al ., 2000 ; Daniell et al ., 2001 ). For the production of recombinant protein in plants, seed systems have recently been proven to be more advantageous than systems using leaf or root ( Delaney, 2002 ; Howard and Hood, 2002 ). The seed is a storage organ within which a small number of storage proteins are highly and stably accumulated in specific organelles designated as protein bodies. Taking advantage of these properties, the seed has been utilized as an ideal bio-reactor for the production of recombinant proteins. Recombinant proteins accumulated in seed are highly stable, and do not require processing or purification, allowing direct oral delivery. When antibody or vaccine is expressed in seed, it has been reported to be highly stable, with no degradation for years even if stored at room temperature. Furthermore, vaccine delivered via seed is expected to induce antibody production through the mucosal immune system without purification and processing ( Walmsley and Arntzen, 2000 ). Rice is one of the most important crops and food resources in the world. Conventional breeding has long been employed to improve seed yield and quality. As a production system, rice seed has several advantages over other cereal crops, such as wheat, barley and rye, including easier storage and processing, greater biomass (yield per unit area) and lower producer costs ( Stoger et al ., 2002 ). In addition, transformation systems have already been established for rice ( Hiei et al ., 1998 ), and whole genome sequence information is now available ( Goff et al ., 2002 ). For rice seed to be utilized as a platform for the production of recombinant proteins, it is important to use promoters tailored to the requirements of each particular protein and biotechnology application, because the promoter controls not only the time and place but also the level of expression. However, analyses of the cis -regulatory elements involved in endosperm-specific expression have been limited to a few glutelin genes, using heterologous transgenic tobacco and homologous transgenic rice ( Takaiwa et al ., 1991a, 1996 ; Zheng et al ., 1993 ; Zhao et al ., 1994 ; Croissant-Sych and Okita, 1996 ; Yoshihara et al ., 1996 ; Wu et al ., 1998a, 2000 ). For a few other rice storage protein promoters, only the spatial expression patterns have been observed ( Wu et al ., 1998b ). In this study, the tissue-specific expression of 10 seed storage protein genes and five genes for enzymes involved in carbohydrate and nitrogen metabolism was examined by introducing them into rice via Agrobacterium -mediated transformation. These genes included: three glutelin genes, GluB-1 , GluB-2 and GluB-4 ( Masumura et al ., 1989a ; Takaiwa et al ., 1991b ); three prolamin genes, 10 kDa, 13 kDa ( PG5a ) and 16 kDa ( Masumura et al ., 1989b ; Nakase et al ., 1996a ; Mitsukawa et al ., 1999 ); a 26 kDa globulin gene ( Glb-1 ) ( Nakase et al ., 1996b ); a rice gene encoding the small subunit of ADP-glucose pyrophosphorylase (AGPase) ( Anderson et al ., 1991 ); a rice pyruvate orthophosphate dikinase gene (PPDK) ( Imaizumi et al ., 1997 ); a rice alanine aminotransferase gene (AlaAT) ( Kikuchi et al ., 1999 ); a rice embryo globulin gene (REG-2) ( Sun et al ., 1996 ); a rice 18 kDa oleosin gene (Ole18) ( Wu et al ., 1998c ); a rice glutamate synthase gene (GOGAT) ( Goto et al ., 1998 ); and a rice starch branching enzyme gene (SBE1) ( Kawasaki et al ., 1993 ). The soybean ॆ-conglycinin gene ( Doyle et al ., 1986 ) and maize ubiquitin gene promoters ( Christensen et al ., 1992 ) were introduced as controls, representing a typical seed protein gene and constitutive promoter, respectively. The nucleotide sequences of these 15 promoters will appear in the EMBL/ GenBank /DDBJ Nucleotide Sequence Databases with accession numbers AY427562 (REG-2), AY427563 (Ole18), AY427564 (AlaAT), AY427565 (GOGAT), AY427566 (AGPase small subunit), AY427567 (PPDK), AY427568 ( SBE1 ), AY427569 ( GluB-1 ), AY427570 ( GluB-2 ), AY427571 ( GluB-4 ), AY427572 (10 kDa prolamin), AY427573 (13 kDa prolamin), AY427574 (16 kDa prolamin), AY427575 ( Glb-1 ) and AB008679 (soybean ॆ-conglycinin). The major objective of this investigation was to characterize promoters that have the potential to drive high expression or specific spatial and temporal patterns of expression during rice seed maturation, in order to prepare a promoter set useful for the production of high value recombinant proteins in rice. <h1>Results</h1> <h2>Construction of promoter-GUS chimeric genes and isolation of transgenic plants</h2> The aim of this study was not to investigate the putative regulatory elements, but to characterize the expression pattern and promoter activities of several genes expressed in seed. Fifteen promoters with sizes ranging from 0.8 to 2.4 kb were isolated by polymerase chain reaction (PCR) using genomic DNA or genomic clones as templates. The isolated promoters were linked to the GUS reporter gene ( Figure 1 ) on the binary vector pGPTV-35S-HPT to make transcriptional fusions ( Jefferson, 1987 ), which were subsequently introduced into rice ( Oryza sativa L. cv. Kitaake) via Agrobacterium tumefaciens -mediated transformation. The presence of the desired promoter fusion was confirmed by PCR using genomic DNA isolated from leaves of independent transgenic rice lines, and positive lines (6 to 21) were subjected to analysis of the promoter properties. <h2>Promoters of seed storage protein genes direct GUS expression in endosperm or embryo</h2> To determine the sites of GUS reporter gene expression directed by the seed storage protein promoters, seeds of transgenic rice were examined by histochemical staining. Seeds collected at the stage of 17 days after flowering (DAF) from transgenic plants were longitudinally sectioned and stained with X-Gluc. The expression patterns are shown in Figure 2 . The rice glutelin promoters (1.3 kb and 2.3 kb GluB-1 , GluB-2 and GluB-4 ; Figure 2a–d ) and prolamin promoters (10 kDa, 13 kDa and 16 kDa; Figure 2e–g ) directed the GUS gene to express in endosperm. GUS expression from glutelin and prolamin promoters was observed in the aleurone and subaleurone, but not in the embryo. Closer inspection of maturing seed of transgenic rice containing the glutelin and prolamin promoters revealed that GUS activity was strong in the outer portion and weak in the inner portion of the endosperm. The GluB-1 promoters (both 1.3 kb and 2.3 kb) showed much higher activity in the region of the endosperm close to the embryo. The GUS expression directed by the 13 kDa prolamin promoter ( PG5a ) was strictly limited to the outer portion of the endosperm. The 26 kDa globulin Glb-1 promoter directed the GUS expression in inner starchy endosperm tissue ( Figure 2h ). The GUS expression driven by the embryo storage protein promoters (REG-2, Ole18 and ॆ-conglycinin; Figure 2i–k ) was restricted to the embryo and aleurone with no expression in the endosperm. There was little difference among the GUS expression patterns mediated by these embryo storage protein promoters. It is interesting to note that the ॆ-conglycinin promoter, from the dicot soybean, retained embryo-specific expression in the monocot rice, even though there have been numerous reports of differences in expression between dicotyledonous and monocotyledonous plants ( Rathaus et al ., 1993 ; Chowdhury et al ., 1997 ). It is notable that GUS expression directed by the ॆ-conglycinin promoter was very weak in rice, in marked contrast with the high expression observed in embryo and cotyledon when the same promoter was introduced into the dicot tobacco. Generally, no GUS activity was detected in leaf, leaf sheath, stalk or roots of transgenic rice carrying fusions to seed storage protein promoters (data not shown). The one exception was the 10 kDa prolamin promoter, which did show some expression in vegetative organs ( Figure 3 ). These results support the conclusion that the endosperm storage protein genes (except for 10 kDa prolamin) are expressed in an endosperm-specific manner and that expression of the embryo storage protein genes is restricted to the embryo and aleurone layer. Although seed storage protein promoters showed either endosperm- or embryo-specific expression, non-storage protein promoters exhibited different expression patterns ( Figure 2 ). The GUS gene directed by the AlaAT promoter was expressed in the central part of the starchy endosperm with stronger activity in the endosperm near the embryo ( Figure 2l ). The expression pattern of the PPDK-GUS transgene was similar to that of the endosperm storage protein ( Figure 2o ). The GUS gene directed by the AGPase small subunit promoter was expressed throughout the seed, including the pericarp, but especially in the inner starchy endosperm and the embryo ( Figure 2n ). In contrast, the GOGAT and SBE promoters directed GUS gene expression mainly in the scutellum (at the boundary between the embryo and endosperm) ( Figure 2m,p ). <h2>Patterns of GUS expression in vegetative organs</h2> Most promoters tested here exhibited endosperm- or embryo-specific expression of the GUS gene. However, GUS activity was also detected in vegetative tissues in the transgenic rice carrying the 10 kDa prolamin, PPDK and AGPase small subunit promoters ( Figure 3 ), as well as the AlaAT promoter ( Kikuchi et al ., 1999 ). In these transgenic rice plants, GUS activity was detected in the phloem of large vascular bundles in leaf, leaf sheath and stalk, in addition to the endosperm or the whole seed ( Figure 3a–c ). GUS activity was also detected in the root endodermis in these transgenic rice plants. However, the expression pattern driven by the AGPase small subunit promoter differed slightly from that of the PPDK and 10 kDa prolamin promoters; specifically, the former promoter drove strong GUS expression in the apical meristem, whereas the latter promoters resulted in root tissue that was uniformly stained. Furthermore, the GUS activities in the root differed, i.e. the AGPase small subunit promoter was stronger than the PPDK and 10 kDa prolamin promoters. The intact 10 kDa prolamin gene is normally expressed in maturing endosperm and not detected in vegetative tissues. The ectopic expression of the GUS fusion observed here could be reverted back to the normal endosperm-specific expression by replacing the Nos terminator with the 0.3 kb region downstream of the stop codon of the native 10 kDa prolamin gene (data not shown). It should be noted that this replacement of the 3′ terminator region had little effect on promoter activities. These results indicate that the endosperm-specific expression of the 10 kDa prolamin gene requires both the 5′ and 3′ flanking regions. <h2>Promoter activities during seed development</h2> The distribution of transgene expression within developing seed was examined by histological staining of longitudinal sections of seed harvested at the developing stages of 7, 12 and 17 DAF. In Kitaake, an early heading variety, morphogenesis of the embryo and cellularization of the endosperm finish before 5 DAF; the endosperm cells then start to accumulate storage reserves (starch and storage proteins). The endosperm cells gradually expand by deposition of storage reserves and reach a maximum size at 12 DAF (mid-maturation stage); the fresh weight of seed reaches a maximum level at 17 DAF (late maturation stage). Seed completely maturates and desiccates around 23 DAF. The expression pattern during seed maturation was examined in every transgenic line, and the results for one representative line for each seed promoter are shown in Figure 4 . It is interesting to note that the site of the first detectable GUS expression was different for each construct. In the case of the glutelin and prolamin promoters, blue GUS staining was first observed in the peripheral region of the endosperm, i.e. in the aleurone and subaleurone. For the glutelin and 16 kDa prolamin promoters, but not for the 10 kDa and 13 kDa prolamin promoters, the staining then spread to the inner starchy endosperm as the seed matured (17 DAF) ( Figure 4a–g ). This expression pattern was in remarkable contrast with that observed in the 26 kDa Glb-1 promoter, in which blue GUS staining was first detected in inner starch endosperm cells close to the embryo and the expression pattern was not changed during seed development ( Figure 4h ). GUS gene expression directed by the promoters of the REG-2, Ole18 and ॆ-conglycinin genes was observed by 7 DAF. The activity tended to start in the aleurone layer, followed by expression in the embryo. The expression of these promoters was restricted to the aleurone and embryo ( Figure 4i–k ). The temporal patterns during seed maturation for transgenic lines representative of non-storage protein promoters are shown in Figure 4(l–p) . GUS expression from the AlaAT promoter appeared first in the inner starchy endosperm tissue and eventually spread throughout the whole endosperm, but the embryos remained unstained ( Figure 4l ). The GUS activity driven by the SBE1 promoter was also limited to the inner starchy endosperm tissue, especially to the tissue near the embryo ( Figure 4p ). However, because of the very low GUS activity, blue staining could not be detected until 12 DAF. In contrast, when the AGPase small subunit gene promoter fusion was introduced, GUS staining initiated in the embryo and then spread to the central portion of the endosperm. Blue GUS staining was finally observed in the whole seed tissues towards maturation, with stronger staining in the embryo ( Figure 4n ). This expression profile during seed development was very similar to that observed for the ubiquitin promoter ( Figure 4q ). On the other hand, the PPDK promoter exhibited an expression pattern similar to that of the glutelin and prolamin gene promoters ( Figure 4o ). GUS activity directed by the GOGAT promoter was restricted to the scutellum with little change during seed development, except that GUS activity could not be detected at 7 DAF ( Figure 4m ). <h2>Quantitative analysis of promoter strength</h2> To evaluate the potential strength of the various promoters, GUS expression levels were determined in the maturing seed (17 DAF) from independent transgenic plants for each construction. As shown in Figure 5 , the promoter activities varied widely. On the basis of the promoter strength, the seed promoters analysed here can be classified into four groups. The high GUS activity group has four members: the GluB-4 , 10 kDa prolamin, 16 kDa prolamin and Glb-1 promoters. The average GUS activities for these promoters were 44.8 ± 16.5, 38.8 ± 10.8, 27.1 ± 12.7 and 28.6 ± 11.8 pmol (4 m U/min/µg protein), respectively. The numbers of transgenic rice analysed for these four constructions were 19, 10, 8 and 20 lines, respectively. The medium GUS activity group has two members: the 2.3 kb GluB-1 and AGPase small subunit gene promoters. Their GUS activities are weaker than those observed in the high activity group, but are much stronger than those of the others. The average GUS activities for the 2.3 kb GluB-1 and AGPase small subunit promoters were 21.3 ± 7.0 and 10 ± 4.7 pmol (4 m U/min/µg protein) for 10 and 18 independent lines, respectively. Seven promoters, namely the 1.3 kb GluB-1 , GluB-2 , 13 kDa prolamin, REG-2, Ole18, AlaAT and PPDK promoters, were tentatively classified into the relatively low GUS activity group. The average GUS activities for these promoters were 2.1 ± 1.2, 5.5 ± 2.2, 7.4 ± 5.5, 2.4 ± 1.2, 2 ± 4.6, 5.9 ± 4.0 and 4.0 ± 3.0 pmol (4 m U/min/µg protein), respectively. The GUS expression levels of these seven promoters were obtained by investigating 9, 12, 6, 12, 14, 6 and 21 independent lines, respectively. The remaining three promoters (the GOGAT, SBE1 and ॆ-conglycinin gene promoters) were grouped into the low GUS activity group. These promoters directed very weak GUS expression with activities less than 1 pmol (4 m U/min/µg protein), when the average GUS activities were examined with 18, 16 and 11 independent lines for the corresponding constructs. GUS activity expressed from the ubiquitin promoter, which was used as a control, averaged 7.4 ± 8.5 pmol (4 m U/min/µg protein) in maturing seed. This level is similar to that obtained from the relatively low GUS activity group promoters, even though the ubiquitin promoter has been used in many applications as a universal promoter. For comparison, the activities of the PPDK and AGPase small subunit promoters were also examined in vegetative tissues. The average GUS activities in leaf, stalk and sheath were 8.7 ± 6.8, 3.7 ± 3.6 and 16.3 ± 13.9 pmol (4 m U/min/µg protein), respectively, for the PPDK promoter, and 12.5 ± 5.0, 40.2 ± 28.5 and 23.2 ± 16.6 pmol (4 m U/min/µg protein), respectively, for the AGPase small subunit promoter. These promoter activities were similar to or higher than those obtained from maturing seeds. On the other hand, although vegetative expression was observed for the 10 kDa prolamin promoter, the GUS activities [leaf, 3.1 ± 1.1; stalk, 6.0 ± 2.9; and sheath, 2.3 ± 1.0 pmol (4 m U/min/µg protein)] were much lower than those observed in maturing seeds. Although the PPDK, AGPase small subunit and 10 kDa prolamin genes were constitutively expressed, the expression levels in different tissues varied depending on the gene. <h1>Discussion</h1> Protein and starch are the major storage reserves in rice seed. It is important to note that all the genes involved in the synthesis of these storage components are mainly expressed in seed. It has been reported that the expression pattern and transcript level are primarily regulated by the individual gene promoter, although the final amount of gene product is determined at both the transcriptional and post-transcriptional levels. Recent advances in plant biotechnology have made it possible not only to improve traditional agricultural products through increased pest/herbicide/disease resistance, yield and nutritional value, but also to use plants as bio-reactors for the production of a wide range of recombinant proteins for pharmaceutical and industrial uses. However, a shortage of suitable promoters is a major factor limiting the production of recombinant proteins in cereal seed, not only at the desired levels, but also in the desired spatial and temporal manner. Wu et al . (1998b ), using transgenic rice lines produced via electroporation, evaluated the endosperm specificity and activity of four kinds of rice seed storage protein gene promoter. As many genes that are highly expressed in maturing rice seed have now been identified by microarray analysis ( Zhu et al ., 2003 ) and serial analysis of gene expression (SAGE) ( Gibbings et al ., 2003 ), it is worthwhile extending this investigation to other seed component promoters. The major objective of this investigation was to use stable transgenic rice lines to characterize the expression pattern and to evaluate the potential promoter activity of several seed genes involved in metabolism and the biosynthesis of seed storage components. Based on their pattern of expression within the seed ( Figures 2 and 4 ), the tested promoters can be classified into four groups: (1) The Outer Endosperm-Specific Group containing eight members: 1.3 kb GluB-1 , 2.3 kb GluB-1 , GluB-2 , GluB-4 , 10 kDa prolamin, 13 kDa prolamin ( PG5a ), 16 kDa prolamin and PPDK; (2) The Inner Endosperm-Specific Group containing four members: Glb-1 , AlaAT, GOGAT and SBE1; (3) The Embryo-Specific Group containing three members: REG-2, Ole18 and ॆ-conglycinin; and (4) The Whole Seed Group containing only one member: AGPase small subunit promoter ( Figure 2 ). The developmental stage at which GUS activity began to be detectable varied among the promoters used here and was positively correlated with promoter activity, i.e. the stronger the promoter, the earlier GUS was expressed. In transgenic rice containing the glutelin, prolamin and PPDK promoter fusions, GUS activity could be detected at 7 DAF, initially in the aleurone, subaleurone and outer portion of the endosperm ( Figure 4 ), and then spreading to all tissues of the endosperm. In the case of the embryo storage protein promoters (REG-2 and Ole18), GUS activity was first detected in the aleurone at 7 DAF, followed by expression in the embryo at later stages of seed development. This temporal pattern of staining, first in the aleurone layer and then in the embryo, did not vary during seed development ( Figure 4 ). A change in the GUS expression pattern, which started at the embryo and then spread to the inner starchy endosperm, was observed for the AGPase small subunit promoter ( Figure 4 ). The temporal expression pattern of the AGPase small subunit promoter is unique among the promoters tested. The AlaAT promoter directed expression in the inner endosperm at the early maturation stage (7 DAF) and the expression level gradually increased towards 17 DAF, resulting in expression throughout the whole endosperm ( Figure 4 ). GUS expression directed by the GOGAT and SBE1 promoters could be detected weakly at mid-development (12 DAF), and was restricted to the endosperm and embryo boundary and to the inner endosperm near the embryo, respectively ( Figure 4 ), reflecting the weakness of the promoters, although the SBE1 enzyme has been reported to be expressed abundantly in the mid-maturation stage of developing seeds ( Kawasaki et al ., 1993 ). The GUS activities varied significantly among the promoters. A wide range of activities were also observed among the individual transformants for each construct, reflecting differences in insertion position and gene dosage effects ( Hobbs et al ., 1993 ). The highest GUS activities were obtained by using the GluB-4 , 10 kDa prolamin, 16 kDa prolamin and 26 kDa globulin promoters ( Figure 5 ). The GluB-1 , GluB-2 , AGPase small subunit, PPDK and AlaAT promoters also directed high GUS activity in the endosperm, but their activities were much lower than those of the four strongest promoters. Nevertheless, it is notable that these promoter activities, although in the relatively low activity group, are similar to that of the maize ubiquitin promoter, which is often used as a universal promoter. Taken together, the strong endosperm-specific promoters developed here are expected to perform well in the production of recombinant proteins whilst avoiding any detrimental effects on vegetative tissues. The embryo storage protein (REG-2, Ole18 and ॆ-conglycinin) promoters showed expression strictly limited to the embryo and aleurone layer ( Figure 2 ). This observation is particularly interesting, because the aleurone is a layer of cells that differentiates from the starchy endosperm during seed development. Although it was observed by histochemical analysis that these promoters drove strong expression of the GUS gene in the embryo and aleurone layer, the activities were not very high on a whole seed basis. This result might be accounted for by the fact that the expression is restricted to a small portion in the rice seed. These embryo storage protein promoters are expected to be an ideal choice for the engineering of seed lipid metabolism, which mainly occurs in the embryo and aleurone. In the transformants carrying fusions to the seed storage protein gene promoters ( GluB-1 , GluB-2 and GluB-4 glutelins, 13 kDa and 16 kDa prolamins, Glb-1 , REG-2 and Ole18), strong GUS activity was detected in maturing seeds, whereas the leaves and stems showed only background or very weak GUS activity (data not shown). These results indicated that the 5′ flanking regions of these genes were sufficient to determine endosperm or embryo specificity. This is in contrast with the results for the 10 kDa prolamin promoter, which directed GUS gene expression in all tissues of the plant, except for the embryo ( Figures 2 and 3 ). This unusual ectopic expression was prevented by replacing the Nos terminator with the native 10 kDa prolamin terminator. This is the first example of a seed storage protein gene requiring both the 3′ non-coding region as well as the 5′ flanking region for endosperm-specific expression. This study also showed that, for the endosperm storage protein genes tested, the 5′ flanking regions contained all the elements required for endosperm-specific expression. The GCN4 and AACA motifs are conserved in all glutelin gene promoters isolated so far ( Takaiwa et al ., 1996 ). The importance of these two motifs in controlling gene expression in the endosperm has been demonstrated ( Takaiwa et al ., 1996 ; Yoshihara et al ., 1996 ). The GCN4 motif is also observed in the PPDK, SBE1 and 13 kDa ( PG5a ) and 16 kDa prolamin promoters. This motif acts as a key element conferring aleurone- and subaleurone-specific expression ( Wu et al ., 1998a ). The promoters of at least eight rice seed storage protein genes ( GluB-1 , GluB-2 , GluB-4 , GluA-1 , GluA-3 , 10 kDa prolamin, 13 kDa ( PG5a ) prolamin and 16 kDa prolamin) have been shown to be expressed predominantly in aleurone and subaleurone cells in transgenic plants. It is likely that the GCN4 motif conserved in these promoters may contribute greatly to this expression pattern. This hypothesis does not contradict the evidence that the embryo-specific REG-2, Ole18 and soybean ॆ-conglycinin gene promoters do not contain the GCN4 motif. The AGPase small subunit and GOGAT promoters also do not contain the GCN4 motif, but they contain various other potential cis -elements, such as the endosperm motif, −300 element and typical DNA repeats in the 5′ flanking regions ( Anderson et al ., 1991 ; Goto et al ., 1998 ). Further investigation will be required to fully understand the regulatory mechanisms underlying the endosperm specificity of these promoters. The activities of the PPDK, AGPase small subunit and AlaAT promoters were relatively high, whereas those of the GOGAT and SBE1 promoters were very low. The PPDK and AGPase small subunit promoters directed the GUS gene to express in vegetative organs as well as in seed ( Figure 3 ). These promoters might be useful for the production of transgenic rice where the final products are accumulated in vegetative tissues. It is interesting to note that the activity of the GluB-4 promoter was the highest amongst the promoters tested, despite its relatively short length of 1.4 kb. In light of the fact that about 250 bp of the upstream region overlaps with the coding region of the α-expansin gene, it is notable that the region of about 1 kb between the α-expansin and GluB-4 coding regions includes all the elements necessary to provide the strongest promoter activity. It has been demonstrated for the GluB-1 gene that at least three cis -elements containing the GCN4, ACGT and AACA motifs within the −197 bp proximal promoter are the minimal elements required to direct endosperm-specific expression ( Wu et al ., 1998a, 2000 ). It is noteworthy that, although the 2.3 kb GluB-1 and 1.3 kb GluB-1 promoters used in this study were derived from the same GluB-1 promoter and their expression patterns were similar, their promoter activities were quite different, i.e. the former was about 10-fold stronger than the latter ( Figure 5 ). The only notable difference between the two promoters is the presence of one more ACGT motif in the additional 1 kb fragment of the 2.3 kb promoter. The activity of the GluB-2 promoter was much lower than that of the 2.3 kb GluB-1 promoter, even though their 5′ flanking sequences are relatively conserved up to about 600 bp from the transcriptional start site ( Takaiwa et al ., 1991a ). This difference in activity may also be explained by the absence of the ACGT motif in the GluB-2 promoter, although it contains four GCN4 motifs and three AACA motifs which are required for endosperm-specific expression. Furthermore, it has been demonstrated by loss-of-function experiments that the removal of the ACGT motif in the proximal GluB-1 promoter causes a remarkable decrease in promoter activity ( Washida et al ., 1999 ; Wu et al ., 2000 ). Taking these results together, it seems that the ACGT motif might be involved in the quantitative regulation of gene expression. To date, many improvements in rice seed quality have been achieved by introducing deficient nutrients through genetic engineering ( Zheng et al ., 1995 ; Goto et al ., 1999 ; Katsube et al ., 1999 ; Ye et al ., 2000 ). Our results demonstrate various expression patterns and levels of activity among seed promoters, thus providing alternative choices for expressing foreign genes of interest at a desirable level in the targeted tissue. <h1>Experimental procedures</h1> <h2>Plasmid construction</h2> Promoters for the genes of interest were isolated from genomic DNA by PCR using specific primers designed according to information from public databases. The genes and the sizes of the corresponding promoters were as follows: 10 kDa prolamin, 0.8 kb; 13 kDa prolamin ( PG5a ), 0.9 kb; 16 kDa prolamin, 0.9 kb; 26 kDa globulin, 1.0 kb; GluB-4 , 1.4 kb; rice embryo globulin (REG-2), 1.3 kb; rice 18 kDa oleosin (Ole18), 1.3 kb; rice glutamate synthase gene (GOGAT), 0.8 kb; rice pyruvate orthophosphate dikinase (PPDK), 0.8 kb; rice ADP-glucose pyrophosphorylase small subunit (AGPase), 2.0 kb; rice starch branching enzyme (SBE1), 2.0 kb; and soybean ॆ-conglycinin, 1.0 kb. The 1.3 kb GluB-1 , 2.3 kb GluB-1 , 2.4 kb GluB-2 and 1.0 kb rice alanine aminotransferase (AlaAT) gene promoters were already available in our laboratory. Various promoter fragments were inserted into a modified binary vector, pGPTV-35S-HPT, containing the hygromycin phosphotransferase ( HPT ) gene as a selectable marker. The modified vector was constructed from the pGPTV-HPT binary vector ( Becker et al ., 1992 ) by exchanging the Nos promoter for the 0.8 kb CaMV 35S promoter as the promoter of the HPT gene. The seed promoters to be tested were introduced upstream of the UdiA gene which encodes ॆ-glucuronidase (GUS) in the modified binary vector. <h2>Production of transgenic plants</h2> Transgenic rice plants ( Oryza sativa cv. Kitaake, an early heading variety) were produced by Agrobacterium -mediated transformation. Plasmids constructed as described above were transferred into Agrobacterium tumefaciens strain EHA105 by electroporation. Five-week-old calli derived from a mature rice seed were treated with the transformed A. tumefaciens for 3 days. The infected calli were successively cultured for 4 weeks each in N6 selection and MS regeneration medium containing hygromycin. The regenerated seedlings were transplanted to a greenhouse as described previously ( Goto et al ., 1999 ). <h2>Analysis of GUS gene expression</h2> For histochemical analysis, maturing seeds at 7, 12 and 17 DAF were sectioned longitudinally with a razor blade and incubated in 50 m m sodium phosphate buffer (pH 7.0) containing 0.5 m m X-Gluc (5-brom-4-chloro-3-indolyl glucuronide) and 20% methanol at 37 °C. The optimum incubation times for the staining reaction varied from 30 min to overnight depending on the abundance of the GUS activity. Fluorometric assays of GUS activities were conducted according to Jefferson (1987 ). Maturing seeds at 17 DAF were homogenized in GUS extraction buffer (50 m m NaPO 4 , pH 7.0, 10 m m 2-mercaptoethanol, 10 m m Na 2 -EDTA, 0.1% sodium dodecylsulphate, 0.1% Triton X-100). After centrifugation, 10 µL of the supernatant was mixed with 90 µL of an assay buffer containing 1 m m 4-methyl umbelliferryl ॆ- d -glucuronide (MUG). After incubation at 37 °C for 1 h, the reaction was stopped by adding 900 µL of 0.2 m Na 2 CO 3 . Fluorometer values were compared with those of a 4-methyl umbelliferone (4 m U) dilution series. Protein content was determined with a Bio-Rad Protein Assay kit using bovine serum albumin as the standard. Three seeds from each independent transgenic plant were assayed. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Plant Biotechnology Journal Wiley

Evaluation of tissue specificity and expression strength of rice seed component gene promoters in transgenic rice

Qu, Le Qing; Takaiwa, Fumio
Plant Biotechnology Journal , Volume 2 (2) – Mar 1, 2004
Download PDF
14 pages

Loading next page...
 
/lp/wiley/evaluation-of-tissue-specificity-and-expression-strength-of-rice-seed-Qo8Yxh5e00

References (45)

Publisher
Wiley
Copyright
© 2004 Blackwell Publishing Ltd
ISSN
1467-7644
eISSN
1467-7652
DOI
10.1111/j.1467-7652.2004.00055.x
pmid
17147604
Publisher site
See Article on Publisher Site

Abstract

<h1>Introduction</h1> Plant biotechnology has been applied not only to the improvement of agronomic properties, such as pest/pathogen resistance, abiotic stress resistance, yield and quality, but also to the production of pharmaceutical and industrial materials. Several recombinant industrial enzymes and pharmaceuticals have been commercialized already, and the number is expected to increase in the near future ( Hood and Jilka, 1999 ; Hood and Woodard, 2002 ). Plant production of recombinant proteins has many advantages, the most prominent of which are cost savings relative to animal transgenic systems, easy control of production scale in response to market size and freedom from animal-derived pathogens, including viruses and prions (Fischer and Emans, 2000; Giddings et al ., 2000 ; Daniell et al ., 2001 ). For the production of recombinant protein in plants, seed systems have recently been proven to be more advantageous than systems using leaf or root ( Delaney, 2002 ; Howard and Hood, 2002 ). The seed is a storage organ within which a small number of storage proteins are highly and stably accumulated in specific organelles designated as protein bodies. Taking advantage of these properties, the seed has been utilized as an ideal bio-reactor for the production of recombinant proteins. Recombinant proteins accumulated in seed are highly stable, and do not require processing or purification, allowing direct oral delivery. When antibody or vaccine is expressed in seed, it has been reported to be highly stable, with no degradation for years even if stored at room temperature. Furthermore, vaccine delivered via seed is expected to induce antibody production through the mucosal immune system without purification and processing ( Walmsley and Arntzen, 2000 ). Rice is one of the most important crops and food resources in the world. Conventional breeding has long been employed to improve seed yield and quality. As a production system, rice seed has several advantages over other cereal crops, such as wheat, barley and rye, including easier storage and processing, greater biomass (yield per unit area) and lower producer costs ( Stoger et al ., 2002 ). In addition, transformation systems have already been established for rice ( Hiei et al ., 1998 ), and whole genome sequence information is now available ( Goff et al ., 2002 ). For rice seed to be utilized as a platform for the production of recombinant proteins, it is important to use promoters tailored to the requirements of each particular protein and biotechnology application, because the promoter controls not only the time and place but also the level of expression. However, analyses of the cis -regulatory elements involved in endosperm-specific expression have been limited to a few glutelin genes, using heterologous transgenic tobacco and homologous transgenic rice ( Takaiwa et al ., 1991a, 1996 ; Zheng et al ., 1993 ; Zhao et al ., 1994 ; Croissant-Sych and Okita, 1996 ; Yoshihara et al ., 1996 ; Wu et al ., 1998a, 2000 ). For a few other rice storage protein promoters, only the spatial expression patterns have been observed ( Wu et al ., 1998b ). In this study, the tissue-specific expression of 10 seed storage protein genes and five genes for enzymes involved in carbohydrate and nitrogen metabolism was examined by introducing them into rice via Agrobacterium -mediated transformation. These genes included: three glutelin genes, GluB-1 , GluB-2 and GluB-4 ( Masumura et al ., 1989a ; Takaiwa et al ., 1991b ); three prolamin genes, 10 kDa, 13 kDa ( PG5a ) and 16 kDa ( Masumura et al ., 1989b ; Nakase et al ., 1996a ; Mitsukawa et al ., 1999 ); a 26 kDa globulin gene ( Glb-1 ) ( Nakase et al ., 1996b ); a rice gene encoding the small subunit of ADP-glucose pyrophosphorylase (AGPase) ( Anderson et al ., 1991 ); a rice pyruvate orthophosphate dikinase gene (PPDK) ( Imaizumi et al ., 1997 ); a rice alanine aminotransferase gene (AlaAT) ( Kikuchi et al ., 1999 ); a rice embryo globulin gene (REG-2) ( Sun et al ., 1996 ); a rice 18 kDa oleosin gene (Ole18) ( Wu et al ., 1998c ); a rice glutamate synthase gene (GOGAT) ( Goto et al ., 1998 ); and a rice starch branching enzyme gene (SBE1) ( Kawasaki et al ., 1993 ). The soybean ॆ-conglycinin gene ( Doyle et al ., 1986 ) and maize ubiquitin gene promoters ( Christensen et al ., 1992 ) were introduced as controls, representing a typical seed protein gene and constitutive promoter, respectively. The nucleotide sequences of these 15 promoters will appear in the EMBL/ GenBank /DDBJ Nucleotide Sequence Databases with accession numbers AY427562 (REG-2), AY427563 (Ole18), AY427564 (AlaAT), AY427565 (GOGAT), AY427566 (AGPase small subunit), AY427567 (PPDK), AY427568 ( SBE1 ), AY427569 ( GluB-1 ), AY427570 ( GluB-2 ), AY427571 ( GluB-4 ), AY427572 (10 kDa prolamin), AY427573 (13 kDa prolamin), AY427574 (16 kDa prolamin), AY427575 ( Glb-1 ) and AB008679 (soybean ॆ-conglycinin). The major objective of this investigation was to characterize promoters that have the potential to drive high expression or specific spatial and temporal patterns of expression during rice seed maturation, in order to prepare a promoter set useful for the production of high value recombinant proteins in rice. <h1>Results</h1> <h2>Construction of promoter-GUS chimeric genes and isolation of transgenic plants</h2> The aim of this study was not to investigate the putative regulatory elements, but to characterize the expression pattern and promoter activities of several genes expressed in seed. Fifteen promoters with sizes ranging from 0.8 to 2.4 kb were isolated by polymerase chain reaction (PCR) using genomic DNA or genomic clones as templates. The isolated promoters were linked to the GUS reporter gene ( Figure 1 ) on the binary vector pGPTV-35S-HPT to make transcriptional fusions ( Jefferson, 1987 ), which were subsequently introduced into rice ( Oryza sativa L. cv. Kitaake) via Agrobacterium tumefaciens -mediated transformation. The presence of the desired promoter fusion was confirmed by PCR using genomic DNA isolated from leaves of independent transgenic rice lines, and positive lines (6 to 21) were subjected to analysis of the promoter properties. <h2>Promoters of seed storage protein genes direct GUS expression in endosperm or embryo</h2> To determine the sites of GUS reporter gene expression directed by the seed storage protein promoters, seeds of transgenic rice were examined by histochemical staining. Seeds collected at the stage of 17 days after flowering (DAF) from transgenic plants were longitudinally sectioned and stained with X-Gluc. The expression patterns are shown in Figure 2 . The rice glutelin promoters (1.3 kb and 2.3 kb GluB-1 , GluB-2 and GluB-4 ; Figure 2a–d ) and prolamin promoters (10 kDa, 13 kDa and 16 kDa; Figure 2e–g ) directed the GUS gene to express in endosperm. GUS expression from glutelin and prolamin promoters was observed in the aleurone and subaleurone, but not in the embryo. Closer inspection of maturing seed of transgenic rice containing the glutelin and prolamin promoters revealed that GUS activity was strong in the outer portion and weak in the inner portion of the endosperm. The GluB-1 promoters (both 1.3 kb and 2.3 kb) showed much higher activity in the region of the endosperm close to the embryo. The GUS expression directed by the 13 kDa prolamin promoter ( PG5a ) was strictly limited to the outer portion of the endosperm. The 26 kDa globulin Glb-1 promoter directed the GUS expression in inner starchy endosperm tissue ( Figure 2h ). The GUS expression driven by the embryo storage protein promoters (REG-2, Ole18 and ॆ-conglycinin; Figure 2i–k ) was restricted to the embryo and aleurone with no expression in the endosperm. There was little difference among the GUS expression patterns mediated by these embryo storage protein promoters. It is interesting to note that the ॆ-conglycinin promoter, from the dicot soybean, retained embryo-specific expression in the monocot rice, even though there have been numerous reports of differences in expression between dicotyledonous and monocotyledonous plants ( Rathaus et al ., 1993 ; Chowdhury et al ., 1997 ). It is notable that GUS expression directed by the ॆ-conglycinin promoter was very weak in rice, in marked contrast with the high expression observed in embryo and cotyledon when the same promoter was introduced into the dicot tobacco. Generally, no GUS activity was detected in leaf, leaf sheath, stalk or roots of transgenic rice carrying fusions to seed storage protein promoters (data not shown). The one exception was the 10 kDa prolamin promoter, which did show some expression in vegetative organs ( Figure 3 ). These results support the conclusion that the endosperm storage protein genes (except for 10 kDa prolamin) are expressed in an endosperm-specific manner and that expression of the embryo storage protein genes is restricted to the embryo and aleurone layer. Although seed storage protein promoters showed either endosperm- or embryo-specific expression, non-storage protein promoters exhibited different expression patterns ( Figure 2 ). The GUS gene directed by the AlaAT promoter was expressed in the central part of the starchy endosperm with stronger activity in the endosperm near the embryo ( Figure 2l ). The expression pattern of the PPDK-GUS transgene was similar to that of the endosperm storage protein ( Figure 2o ). The GUS gene directed by the AGPase small subunit promoter was expressed throughout the seed, including the pericarp, but especially in the inner starchy endosperm and the embryo ( Figure 2n ). In contrast, the GOGAT and SBE promoters directed GUS gene expression mainly in the scutellum (at the boundary between the embryo and endosperm) ( Figure 2m,p ). <h2>Patterns of GUS expression in vegetative organs</h2> Most promoters tested here exhibited endosperm- or embryo-specific expression of the GUS gene. However, GUS activity was also detected in vegetative tissues in the transgenic rice carrying the 10 kDa prolamin, PPDK and AGPase small subunit promoters ( Figure 3 ), as well as the AlaAT promoter ( Kikuchi et al ., 1999 ). In these transgenic rice plants, GUS activity was detected in the phloem of large vascular bundles in leaf, leaf sheath and stalk, in addition to the endosperm or the whole seed ( Figure 3a–c ). GUS activity was also detected in the root endodermis in these transgenic rice plants. However, the expression pattern driven by the AGPase small subunit promoter differed slightly from that of the PPDK and 10 kDa prolamin promoters; specifically, the former promoter drove strong GUS expression in the apical meristem, whereas the latter promoters resulted in root tissue that was uniformly stained. Furthermore, the GUS activities in the root differed, i.e. the AGPase small subunit promoter was stronger than the PPDK and 10 kDa prolamin promoters. The intact 10 kDa prolamin gene is normally expressed in maturing endosperm and not detected in vegetative tissues. The ectopic expression of the GUS fusion observed here could be reverted back to the normal endosperm-specific expression by replacing the Nos terminator with the 0.3 kb region downstream of the stop codon of the native 10 kDa prolamin gene (data not shown). It should be noted that this replacement of the 3′ terminator region had little effect on promoter activities. These results indicate that the endosperm-specific expression of the 10 kDa prolamin gene requires both the 5′ and 3′ flanking regions. <h2>Promoter activities during seed development</h2> The distribution of transgene expression within developing seed was examined by histological staining of longitudinal sections of seed harvested at the developing stages of 7, 12 and 17 DAF. In Kitaake, an early heading variety, morphogenesis of the embryo and cellularization of the endosperm finish before 5 DAF; the endosperm cells then start to accumulate storage reserves (starch and storage proteins). The endosperm cells gradually expand by deposition of storage reserves and reach a maximum size at 12 DAF (mid-maturation stage); the fresh weight of seed reaches a maximum level at 17 DAF (late maturation stage). Seed completely maturates and desiccates around 23 DAF. The expression pattern during seed maturation was examined in every transgenic line, and the results for one representative line for each seed promoter are shown in Figure 4 . It is interesting to note that the site of the first detectable GUS expression was different for each construct. In the case of the glutelin and prolamin promoters, blue GUS staining was first observed in the peripheral region of the endosperm, i.e. in the aleurone and subaleurone. For the glutelin and 16 kDa prolamin promoters, but not for the 10 kDa and 13 kDa prolamin promoters, the staining then spread to the inner starchy endosperm as the seed matured (17 DAF) ( Figure 4a–g ). This expression pattern was in remarkable contrast with that observed in the 26 kDa Glb-1 promoter, in which blue GUS staining was first detected in inner starch endosperm cells close to the embryo and the expression pattern was not changed during seed development ( Figure 4h ). GUS gene expression directed by the promoters of the REG-2, Ole18 and ॆ-conglycinin genes was observed by 7 DAF. The activity tended to start in the aleurone layer, followed by expression in the embryo. The expression of these promoters was restricted to the aleurone and embryo ( Figure 4i–k ). The temporal patterns during seed maturation for transgenic lines representative of non-storage protein promoters are shown in Figure 4(l–p) . GUS expression from the AlaAT promoter appeared first in the inner starchy endosperm tissue and eventually spread throughout the whole endosperm, but the embryos remained unstained ( Figure 4l ). The GUS activity driven by the SBE1 promoter was also limited to the inner starchy endosperm tissue, especially to the tissue near the embryo ( Figure 4p ). However, because of the very low GUS activity, blue staining could not be detected until 12 DAF. In contrast, when the AGPase small subunit gene promoter fusion was introduced, GUS staining initiated in the embryo and then spread to the central portion of the endosperm. Blue GUS staining was finally observed in the whole seed tissues towards maturation, with stronger staining in the embryo ( Figure 4n ). This expression profile during seed development was very similar to that observed for the ubiquitin promoter ( Figure 4q ). On the other hand, the PPDK promoter exhibited an expression pattern similar to that of the glutelin and prolamin gene promoters ( Figure 4o ). GUS activity directed by the GOGAT promoter was restricted to the scutellum with little change during seed development, except that GUS activity could not be detected at 7 DAF ( Figure 4m ). <h2>Quantitative analysis of promoter strength</h2> To evaluate the potential strength of the various promoters, GUS expression levels were determined in the maturing seed (17 DAF) from independent transgenic plants for each construction. As shown in Figure 5 , the promoter activities varied widely. On the basis of the promoter strength, the seed promoters analysed here can be classified into four groups. The high GUS activity group has four members: the GluB-4 , 10 kDa prolamin, 16 kDa prolamin and Glb-1 promoters. The average GUS activities for these promoters were 44.8 ± 16.5, 38.8 ± 10.8, 27.1 ± 12.7 and 28.6 ± 11.8 pmol (4 m U/min/µg protein), respectively. The numbers of transgenic rice analysed for these four constructions were 19, 10, 8 and 20 lines, respectively. The medium GUS activity group has two members: the 2.3 kb GluB-1 and AGPase small subunit gene promoters. Their GUS activities are weaker than those observed in the high activity group, but are much stronger than those of the others. The average GUS activities for the 2.3 kb GluB-1 and AGPase small subunit promoters were 21.3 ± 7.0 and 10 ± 4.7 pmol (4 m U/min/µg protein) for 10 and 18 independent lines, respectively. Seven promoters, namely the 1.3 kb GluB-1 , GluB-2 , 13 kDa prolamin, REG-2, Ole18, AlaAT and PPDK promoters, were tentatively classified into the relatively low GUS activity group. The average GUS activities for these promoters were 2.1 ± 1.2, 5.5 ± 2.2, 7.4 ± 5.5, 2.4 ± 1.2, 2 ± 4.6, 5.9 ± 4.0 and 4.0 ± 3.0 pmol (4 m U/min/µg protein), respectively. The GUS expression levels of these seven promoters were obtained by investigating 9, 12, 6, 12, 14, 6 and 21 independent lines, respectively. The remaining three promoters (the GOGAT, SBE1 and ॆ-conglycinin gene promoters) were grouped into the low GUS activity group. These promoters directed very weak GUS expression with activities less than 1 pmol (4 m U/min/µg protein), when the average GUS activities were examined with 18, 16 and 11 independent lines for the corresponding constructs. GUS activity expressed from the ubiquitin promoter, which was used as a control, averaged 7.4 ± 8.5 pmol (4 m U/min/µg protein) in maturing seed. This level is similar to that obtained from the relatively low GUS activity group promoters, even though the ubiquitin promoter has been used in many applications as a universal promoter. For comparison, the activities of the PPDK and AGPase small subunit promoters were also examined in vegetative tissues. The average GUS activities in leaf, stalk and sheath were 8.7 ± 6.8, 3.7 ± 3.6 and 16.3 ± 13.9 pmol (4 m U/min/µg protein), respectively, for the PPDK promoter, and 12.5 ± 5.0, 40.2 ± 28.5 and 23.2 ± 16.6 pmol (4 m U/min/µg protein), respectively, for the AGPase small subunit promoter. These promoter activities were similar to or higher than those obtained from maturing seeds. On the other hand, although vegetative expression was observed for the 10 kDa prolamin promoter, the GUS activities [leaf, 3.1 ± 1.1; stalk, 6.0 ± 2.9; and sheath, 2.3 ± 1.0 pmol (4 m U/min/µg protein)] were much lower than those observed in maturing seeds. Although the PPDK, AGPase small subunit and 10 kDa prolamin genes were constitutively expressed, the expression levels in different tissues varied depending on the gene. <h1>Discussion</h1> Protein and starch are the major storage reserves in rice seed. It is important to note that all the genes involved in the synthesis of these storage components are mainly expressed in seed. It has been reported that the expression pattern and transcript level are primarily regulated by the individual gene promoter, although the final amount of gene product is determined at both the transcriptional and post-transcriptional levels. Recent advances in plant biotechnology have made it possible not only to improve traditional agricultural products through increased pest/herbicide/disease resistance, yield and nutritional value, but also to use plants as bio-reactors for the production of a wide range of recombinant proteins for pharmaceutical and industrial uses. However, a shortage of suitable promoters is a major factor limiting the production of recombinant proteins in cereal seed, not only at the desired levels, but also in the desired spatial and temporal manner. Wu et al . (1998b ), using transgenic rice lines produced via electroporation, evaluated the endosperm specificity and activity of four kinds of rice seed storage protein gene promoter. As many genes that are highly expressed in maturing rice seed have now been identified by microarray analysis ( Zhu et al ., 2003 ) and serial analysis of gene expression (SAGE) ( Gibbings et al ., 2003 ), it is worthwhile extending this investigation to other seed component promoters. The major objective of this investigation was to use stable transgenic rice lines to characterize the expression pattern and to evaluate the potential promoter activity of several seed genes involved in metabolism and the biosynthesis of seed storage components. Based on their pattern of expression within the seed ( Figures 2 and 4 ), the tested promoters can be classified into four groups: (1) The Outer Endosperm-Specific Group containing eight members: 1.3 kb GluB-1 , 2.3 kb GluB-1 , GluB-2 , GluB-4 , 10 kDa prolamin, 13 kDa prolamin ( PG5a ), 16 kDa prolamin and PPDK; (2) The Inner Endosperm-Specific Group containing four members: Glb-1 , AlaAT, GOGAT and SBE1; (3) The Embryo-Specific Group containing three members: REG-2, Ole18 and ॆ-conglycinin; and (4) The Whole Seed Group containing only one member: AGPase small subunit promoter ( Figure 2 ). The developmental stage at which GUS activity began to be detectable varied among the promoters used here and was positively correlated with promoter activity, i.e. the stronger the promoter, the earlier GUS was expressed. In transgenic rice containing the glutelin, prolamin and PPDK promoter fusions, GUS activity could be detected at 7 DAF, initially in the aleurone, subaleurone and outer portion of the endosperm ( Figure 4 ), and then spreading to all tissues of the endosperm. In the case of the embryo storage protein promoters (REG-2 and Ole18), GUS activity was first detected in the aleurone at 7 DAF, followed by expression in the embryo at later stages of seed development. This temporal pattern of staining, first in the aleurone layer and then in the embryo, did not vary during seed development ( Figure 4 ). A change in the GUS expression pattern, which started at the embryo and then spread to the inner starchy endosperm, was observed for the AGPase small subunit promoter ( Figure 4 ). The temporal expression pattern of the AGPase small subunit promoter is unique among the promoters tested. The AlaAT promoter directed expression in the inner endosperm at the early maturation stage (7 DAF) and the expression level gradually increased towards 17 DAF, resulting in expression throughout the whole endosperm ( Figure 4 ). GUS expression directed by the GOGAT and SBE1 promoters could be detected weakly at mid-development (12 DAF), and was restricted to the endosperm and embryo boundary and to the inner endosperm near the embryo, respectively ( Figure 4 ), reflecting the weakness of the promoters, although the SBE1 enzyme has been reported to be expressed abundantly in the mid-maturation stage of developing seeds ( Kawasaki et al ., 1993 ). The GUS activities varied significantly among the promoters. A wide range of activities were also observed among the individual transformants for each construct, reflecting differences in insertion position and gene dosage effects ( Hobbs et al ., 1993 ). The highest GUS activities were obtained by using the GluB-4 , 10 kDa prolamin, 16 kDa prolamin and 26 kDa globulin promoters ( Figure 5 ). The GluB-1 , GluB-2 , AGPase small subunit, PPDK and AlaAT promoters also directed high GUS activity in the endosperm, but their activities were much lower than those of the four strongest promoters. Nevertheless, it is notable that these promoter activities, although in the relatively low activity group, are similar to that of the maize ubiquitin promoter, which is often used as a universal promoter. Taken together, the strong endosperm-specific promoters developed here are expected to perform well in the production of recombinant proteins whilst avoiding any detrimental effects on vegetative tissues. The embryo storage protein (REG-2, Ole18 and ॆ-conglycinin) promoters showed expression strictly limited to the embryo and aleurone layer ( Figure 2 ). This observation is particularly interesting, because the aleurone is a layer of cells that differentiates from the starchy endosperm during seed development. Although it was observed by histochemical analysis that these promoters drove strong expression of the GUS gene in the embryo and aleurone layer, the activities were not very high on a whole seed basis. This result might be accounted for by the fact that the expression is restricted to a small portion in the rice seed. These embryo storage protein promoters are expected to be an ideal choice for the engineering of seed lipid metabolism, which mainly occurs in the embryo and aleurone. In the transformants carrying fusions to the seed storage protein gene promoters ( GluB-1 , GluB-2 and GluB-4 glutelins, 13 kDa and 16 kDa prolamins, Glb-1 , REG-2 and Ole18), strong GUS activity was detected in maturing seeds, whereas the leaves and stems showed only background or very weak GUS activity (data not shown). These results indicated that the 5′ flanking regions of these genes were sufficient to determine endosperm or embryo specificity. This is in contrast with the results for the 10 kDa prolamin promoter, which directed GUS gene expression in all tissues of the plant, except for the embryo ( Figures 2 and 3 ). This unusual ectopic expression was prevented by replacing the Nos terminator with the native 10 kDa prolamin terminator. This is the first example of a seed storage protein gene requiring both the 3′ non-coding region as well as the 5′ flanking region for endosperm-specific expression. This study also showed that, for the endosperm storage protein genes tested, the 5′ flanking regions contained all the elements required for endosperm-specific expression. The GCN4 and AACA motifs are conserved in all glutelin gene promoters isolated so far ( Takaiwa et al ., 1996 ). The importance of these two motifs in controlling gene expression in the endosperm has been demonstrated ( Takaiwa et al ., 1996 ; Yoshihara et al ., 1996 ). The GCN4 motif is also observed in the PPDK, SBE1 and 13 kDa ( PG5a ) and 16 kDa prolamin promoters. This motif acts as a key element conferring aleurone- and subaleurone-specific expression ( Wu et al ., 1998a ). The promoters of at least eight rice seed storage protein genes ( GluB-1 , GluB-2 , GluB-4 , GluA-1 , GluA-3 , 10 kDa prolamin, 13 kDa ( PG5a ) prolamin and 16 kDa prolamin) have been shown to be expressed predominantly in aleurone and subaleurone cells in transgenic plants. It is likely that the GCN4 motif conserved in these promoters may contribute greatly to this expression pattern. This hypothesis does not contradict the evidence that the embryo-specific REG-2, Ole18 and soybean ॆ-conglycinin gene promoters do not contain the GCN4 motif. The AGPase small subunit and GOGAT promoters also do not contain the GCN4 motif, but they contain various other potential cis -elements, such as the endosperm motif, −300 element and typical DNA repeats in the 5′ flanking regions ( Anderson et al ., 1991 ; Goto et al ., 1998 ). Further investigation will be required to fully understand the regulatory mechanisms underlying the endosperm specificity of these promoters. The activities of the PPDK, AGPase small subunit and AlaAT promoters were relatively high, whereas those of the GOGAT and SBE1 promoters were very low. The PPDK and AGPase small subunit promoters directed the GUS gene to express in vegetative organs as well as in seed ( Figure 3 ). These promoters might be useful for the production of transgenic rice where the final products are accumulated in vegetative tissues. It is interesting to note that the activity of the GluB-4 promoter was the highest amongst the promoters tested, despite its relatively short length of 1.4 kb. In light of the fact that about 250 bp of the upstream region overlaps with the coding region of the α-expansin gene, it is notable that the region of about 1 kb between the α-expansin and GluB-4 coding regions includes all the elements necessary to provide the strongest promoter activity. It has been demonstrated for the GluB-1 gene that at least three cis -elements containing the GCN4, ACGT and AACA motifs within the −197 bp proximal promoter are the minimal elements required to direct endosperm-specific expression ( Wu et al ., 1998a, 2000 ). It is noteworthy that, although the 2.3 kb GluB-1 and 1.3 kb GluB-1 promoters used in this study were derived from the same GluB-1 promoter and their expression patterns were similar, their promoter activities were quite different, i.e. the former was about 10-fold stronger than the latter ( Figure 5 ). The only notable difference between the two promoters is the presence of one more ACGT motif in the additional 1 kb fragment of the 2.3 kb promoter. The activity of the GluB-2 promoter was much lower than that of the 2.3 kb GluB-1 promoter, even though their 5′ flanking sequences are relatively conserved up to about 600 bp from the transcriptional start site ( Takaiwa et al ., 1991a ). This difference in activity may also be explained by the absence of the ACGT motif in the GluB-2 promoter, although it contains four GCN4 motifs and three AACA motifs which are required for endosperm-specific expression. Furthermore, it has been demonstrated by loss-of-function experiments that the removal of the ACGT motif in the proximal GluB-1 promoter causes a remarkable decrease in promoter activity ( Washida et al ., 1999 ; Wu et al ., 2000 ). Taking these results together, it seems that the ACGT motif might be involved in the quantitative regulation of gene expression. To date, many improvements in rice seed quality have been achieved by introducing deficient nutrients through genetic engineering ( Zheng et al ., 1995 ; Goto et al ., 1999 ; Katsube et al ., 1999 ; Ye et al ., 2000 ). Our results demonstrate various expression patterns and levels of activity among seed promoters, thus providing alternative choices for expressing foreign genes of interest at a desirable level in the targeted tissue. <h1>Experimental procedures</h1> <h2>Plasmid construction</h2> Promoters for the genes of interest were isolated from genomic DNA by PCR using specific primers designed according to information from public databases. The genes and the sizes of the corresponding promoters were as follows: 10 kDa prolamin, 0.8 kb; 13 kDa prolamin ( PG5a ), 0.9 kb; 16 kDa prolamin, 0.9 kb; 26 kDa globulin, 1.0 kb; GluB-4 , 1.4 kb; rice embryo globulin (REG-2), 1.3 kb; rice 18 kDa oleosin (Ole18), 1.3 kb; rice glutamate synthase gene (GOGAT), 0.8 kb; rice pyruvate orthophosphate dikinase (PPDK), 0.8 kb; rice ADP-glucose pyrophosphorylase small subunit (AGPase), 2.0 kb; rice starch branching enzyme (SBE1), 2.0 kb; and soybean ॆ-conglycinin, 1.0 kb. The 1.3 kb GluB-1 , 2.3 kb GluB-1 , 2.4 kb GluB-2 and 1.0 kb rice alanine aminotransferase (AlaAT) gene promoters were already available in our laboratory. Various promoter fragments were inserted into a modified binary vector, pGPTV-35S-HPT, containing the hygromycin phosphotransferase ( HPT ) gene as a selectable marker. The modified vector was constructed from the pGPTV-HPT binary vector ( Becker et al ., 1992 ) by exchanging the Nos promoter for the 0.8 kb CaMV 35S promoter as the promoter of the HPT gene. The seed promoters to be tested were introduced upstream of the UdiA gene which encodes ॆ-glucuronidase (GUS) in the modified binary vector. <h2>Production of transgenic plants</h2> Transgenic rice plants ( Oryza sativa cv. Kitaake, an early heading variety) were produced by Agrobacterium -mediated transformation. Plasmids constructed as described above were transferred into Agrobacterium tumefaciens strain EHA105 by electroporation. Five-week-old calli derived from a mature rice seed were treated with the transformed A. tumefaciens for 3 days. The infected calli were successively cultured for 4 weeks each in N6 selection and MS regeneration medium containing hygromycin. The regenerated seedlings were transplanted to a greenhouse as described previously ( Goto et al ., 1999 ). <h2>Analysis of GUS gene expression</h2> For histochemical analysis, maturing seeds at 7, 12 and 17 DAF were sectioned longitudinally with a razor blade and incubated in 50 m m sodium phosphate buffer (pH 7.0) containing 0.5 m m X-Gluc (5-brom-4-chloro-3-indolyl glucuronide) and 20% methanol at 37 °C. The optimum incubation times for the staining reaction varied from 30 min to overnight depending on the abundance of the GUS activity. Fluorometric assays of GUS activities were conducted according to Jefferson (1987 ). Maturing seeds at 17 DAF were homogenized in GUS extraction buffer (50 m m NaPO 4 , pH 7.0, 10 m m 2-mercaptoethanol, 10 m m Na 2 -EDTA, 0.1% sodium dodecylsulphate, 0.1% Triton X-100). After centrifugation, 10 µL of the supernatant was mixed with 90 µL of an assay buffer containing 1 m m 4-methyl umbelliferryl ॆ- d -glucuronide (MUG). After incubation at 37 °C for 1 h, the reaction was stopped by adding 900 µL of 0.2 m Na 2 CO 3 . Fluorometer values were compared with those of a 4-methyl umbelliferone (4 m U) dilution series. Protein content was determined with a Bio-Rad Protein Assay kit using bovine serum albumin as the standard. Three seeds from each independent transgenic plant were assayed.

Journal

Plant Biotechnology Journal – Wiley

Published: Mar 1, 2004

Keywords: embryo; endosperm; gene expression; promoter activity; specific expression

There are no references for this article.