There is a significant difference between cereal crops and floricultural crops in the concept of genetic modification. This is due to the difference in requirements for these agricultural products; wide variation of flower color, shape and fragrance for floricultural plants in contrast to productivity improvements or fine-tuning of physiological traits for crop plants. Bringing GM flowers to the market against their short product life involves the following factors: (1) a high-efficiency production and screening system of elite lines, (2) reliable methods to minimize biodiversity impact and evaluation of their efficacy, and (3) social receptivity based on intensive education and information sharing. We have developed an efficient system by merging the genetic information resources of the Arabidopsis genome and a transcription factor-based gene silencing system called CRES-T through the Flower CRES-T Project. In this project, we have demonstrated the applicability and general versatility of CRES-T in various plant species through experiments on eight different flower species with over 100 transcription factors. In addition to providing phenotypic information via the original database, we are attempting to improve public knowledge, for example, through the production of resin-embedded specimens of GM-flowers and reviving lost garden varieties of morning glory and using them in educational programs, to gain public acceptance of GM plants. Multi-petal cyclamens with complete sterility, which have been produced by suppressing a pair of floral-organ identity genes, will be released in the near future as the third GM commodity following Suntory's blue carnations and rose. This cyclamen represents the best example of a GM product with a low diversity impact ever produced.
FioreDB (http://www.cres-t.org/fiore/public_db/) is a database of phenotypes induced by Chimeric REpressor gene-Silencing Technology (CRES-T) in ornamental and model plants. CRES-T induces a loss-of-function phenotype of a transcription factor (TF) gene by expression of a chimeric repressor produced by fusion of a TF to the strong transcriptional repression domain (SRDX). The earlier version of FioreDB provided phenotypic information induced by many kinds of chimeric repressors in various plants including torenia, chrysanthemum, cyclamen, gentian, morning glory, lisianthus and Arabidopsis. Phenotypic information, however, was not linked with gene information. We report here the development of the new FioreDB that provides more than 300 phenotypic information of various plants, linked to more than 100 TFs. FioreDB also provides information about classification of TFs, putative repression motifs found in TFs and other proteins, and incorporates publicly available gene information such as sequences and microarray data for all Arabidopsis genes. The new FioreDB described here, will be a valuable resource for basic research of TFs and for the manipulation of traits of agronomically important plants by CRES-T, especially from the point of view of horticulture.
Chimeric REpressor gene-Silencing Technology (CRES-T) is a powerful gene-silencing tool to analyze the function of Arabidopsis transcription factors. To investigate whether CRES-T is also applicable to horticultural plants inadequate for genetic engineering because of their limited molecular biological characterization and polyploidy, we applied CRES-T to torenia and the hexaploid chrysanthemum and produced their transgenic plants expressing the chimeric repressor derived from the Arabidopsis TEOSINTE BRANCHED1, CYCLOIDEA, and PCF family transcription factor 3 (TCP3) fused with a plant-specific transcriptional repression domain named SRDX, consisting of 12 amino acids originated from the EAR-motif (TCP3-SRDX). Transgenic torenia and chrysanthemum expressing TCP3-SRDX exhibited fringed leaves and short pistils, while those expressing TCP3 fused with either the mutated repression domain (TCP3-mSRDX) or the overexpressor of TCP3 (TCP3-ox) did not exhibit phenotypic changes. In addition to fringed leaves, TCP3-SRDX transgenic torenia plants exhibited petals with fringed margins, distinctive color patterns, and reduced anthocyanin accumulation. In TCP3-SRDX transgenic chrysanthemum plants, floral organ development was suppressed as compared with the wild type. These results indicate that the Arabidopsis-derived TCP3-SRDX induced morphological changes in transgenic torenia and chrysanthemum although the observed phenotypes partially differ from each other. CRES-T may function in various plant species including polyploid species and modify their biological characteristics.
Cyclamen persicum is among the most popular pot plants in the world. Creating new ruffled flower petals is an important breeding target in this plant. TCP transcription factors are involved in defining plant morphology in angiosperms. To modify flower shape in cyclamen, we isolated a TCP gene from cyclamen (CpTCP1), which is clustered in the same group as TCP3 (AtTCP3) and TCP4 (AtTCP4) in Arabidopsis and CINCINNATA (CIN) in Antirrhinum majus. A chimeric repressor gene construct for CpTCP1 driven by the cauliflower mosaic virus (CaMV) 35S promoter (35S:CpTCP1SRDX) was transformed into Arabidopsis to confirm whether CpTCP1 has a function similar to that of AtTCP3 in Arabidopsis. The phenotypic changes resembled those of the chimeric repressor derived from AtTCP3. The 35S:CpTCP1SRDX was introduced into cyclamen cv. Wink Pink. Finally, we produced and analyzed 106 independent transgenic plants. CpTCP1SRDX expression resulted in cyclamen with serrated wavy leaves and curled petals. Scanning electron microscopy showed that cells in petals and leaf margins of 35S:CpTCP1SRDX plants were smaller than those of controls, and these cells also appeared immature. These results suggest that cyclamen CpTCP1 possibly plays a role in regulating morphogenesis in floral and vegetative organs. We demonstrated that CpTCP1SRDX expression in cyclamen resulted in curly and ruffled flowers with a high ornamental value.
Chimeric repressor gene-silencing technology is a useful tool for changing morphology of ornamental plants. It has previously been demonstrated that the chimeric repressor TCP3SRDX, which consists of Arabidopsis TCP3 and an ERF-associated amphiphilic repression motif repression domain, perturbs the marginal morphology of Arabidopsis leaves and flowers. To obtain new rose cultivars that have ornamental values, we attempted to alter the morphology of Rosa×hybrida cv. Lavande with TCP3SRDX. The TCP3SRDX transgenic rose plants showed interesting phenotypes: the number of leaflets and the size of leaf teeth increased, the petals were wavy, and the sepals were compound-leafy. We succeeded in altering rose morphology using Arabidopsis TCP3 without the sequence information of a TCP3 homologue in the target plant species.
The Chimeric REpressor gene-Silencing Technology (CRES-T) system is a novel reverse genetic method using a chimeric transcriptional repressor fusing an EAR transcriptional repression domain called SRDX. We sought to change the flower shape of Pharbitis nil, a model ornamental flower, using an Arabidopsis transcription factor fused with SRDX. For the first trial modulating flower shape we transformed with the class-C MADS-box transcription factor AGAMOUS (AG) fused with SRDX (AGSRDX). Defects in class-C genes cause double flowers in Arabidopsis and Pharbitis. However, when AGSRDX was expressed under the CaMV 35S promoter (p35S), the transgenic Pharbitis bore a malformed flower with a protruding pistil. We then used DUPLICATED (DP), one of the class-C genes in Pharbitis. The p35S::DPSRDX-introduced callus were difficult to regenerate during transgenic steps, but occasionally made a perfect double flower bud showing severe growth defects. The flower buds never developed to flower opening stage. These results indicate that CRES-T is functional in Pharbitis but even using a conserved transcription factor, some species-specific variation might exist. To avoid these unwanted effects, we recruited inducible promoters to control expression of the chimeric transcription repressors in combination with the DNA-binding domain of GAL4 in yeast fused with SRDX and the GAL4 upstream activator sequence (UAS). Normal regeneration was observed by inducible repression of DPSRDX during in vitro redifferentiation, and the double-flowered Pharbitis was generated. We successfully induced a non-transformant (NT)-like flower in DPSRDX-expressing double-flowered transformants. Our approach will enable us to breed transgenic horticultural plants with inducible fertility.
Molecular analysis of cyclamen flower architecture has not been widely performed, therefore effective molecular breeding of this plant is not available. The ABCDE model is based on interactions between members of different classes of transcription factors that establish floral organ identity, most of which belong to the MADS-box family. To elucidate the mechanism involved in regulating cyclamen flower development, we isolated genes encoding putative MADS-box transcription factors and analyzed their expression patterns in cyclamen. We cloned full-length cDNAs using homology-based RT-PCR and RACE-PCR, and identified and characterized 10 putative cyclamen MADS-box genes. A phylogenetic tree reveals that these genes are related to the APETALA1, PISTILLATA, APETALA3, AGAMOUS, SEEDSTICK, and SEPALLATA transcription factor subfamilies. Respective genes are expressed in each whorl according to the ABCDE function, but there are slight differences in the expression of several genes in various tissues; class-A homologous gene CpAP1 is not expressed in petal. One of three identified class-B homologous gene is TM6-like but expressed broadly. Two closely related class-C homologous genes are differentially expressed in stamens and carpels. These data suggest that modified ABC model might uniquely evolved in cyclamen.
Chimeric repressor gene-silencing technology (CRES-T) is an efficient gene suppression system in a wide variety of dicots and monocots. In this study, we demonstrated that the CRES-T system functions in Japanese gentian. A chimeric repressor of the anthocyanin biosynthetic regulator gene GtMYB3, under the control of the Arabidopsis actin2 promoter, was introduced into blue-flowered gentian. Of 12 transgenic lines, 2 exhibited a picotee flower phenotype with a lack of pigmentation in the lower part of the petal. HPLC analysis showed that the petals of these lines contained less anthocyanin and more flavone than the wild-type, suggesting competitive accumulation of these two types of compounds. The expressions of ‘late’ flavonoid biosynthetic genes, including F3H, F3′5′H, DFR and ANS, were strongly suppressed in petals of these transgenic plants. In contrast, the ‘early’ flavonoid biosynthetic genes, such as CHS and FNSII, were not affected. Since FNSII is expressed more strongly in the lower part of petals than in the upper part, the absence of pigmentation in the lower parts might be induced by flavone synthesis. These results demonstrated that the suppression of anthocyanin biosynthetic genes by CRES-T was successfully applied to Japanese gentian to change petal color; therefore, this system could be useful for generating novel flower colors and patterns. Transgenic plants produced in this study might be utilized as elite materials in the breeding of Japanese gentian in the near future.
To establish an efficient way to create novel floral traits in horticultural flowers, we have introduced many chimeric repressors of Arabidopsis transcription factors into torenia. Among them, we found a transgenic torenia exhibiting unopened flower buds and glossy dark green leaves with curled margins as a consequence of overexpression of Arabidopsis MYB24 with a transcriptional repression domain (MYB24-SRDX). Petals inside the flower buds exhibited a distinct coloration pattern. To bring out this favorable petal trait without inducing the unfavorable phenotypes due to the constitutive expression of chimeric repressors by the cauliflower mosaic virus 35S (35S) promoter, we tested the ability of a floral organ-specific Arabidopsis APETALA1 (AP1) promoter, which was found to be active in both petals and flower buds of torenia. As expected, AP1 pro:MYB24-SRDX transgenic torenias resulted in the opening of flowers and a normal leaf phenotype. Furthermore, these AP1 pro:MYB24-SRDX torenias exhibited wavy petals with a characteristic configuration. This is a good example of the utilization of a floral organ-specific promoter for creating distinct flower phenotypes without causing unfavorable morphological and physiological changes in other organs.
Molecular breeding with genetic modification enables the production of novel floral traits in floricultural plants that could not be obtained by traditional breeding. To facilitate novel flower production, we collectively introduced 2 sets of 42 and 50 chimeric repressors of Arabidopsis transcription factors into Agrobacterium and then used these to co-transform torenia (Torenia fournieri). We generated 750 transgenic torenias, and identification of the transgenes revealed that more than 80% of the transgenic torenias had a single transgene. A total of 264 plants showed phenotypic modification, and 91.2% displayed modified flower colors and/or shapes, such as altered color patterns, curled petal margins, and wavy petals. These results indicated that the collective transformation system can be applied to molecular breeding of flowers. Detailed analysis of the phenotypes revealed that PETAL LOSS could control blotch sizes and that modification of cell shape could change the texture of petals. We found that the chimeric repressors of functionally unknown transcription factors also induced novel floral traits, and therefore, the transgenic torenias provide an understanding of the functions of transcription factors that could not be revealed by previous studies in Arabidopsis.
Chimeric repressor gene-silencing technology (CRES-T) is a powerful tool that has recently been developed for the functional analysis of plant transcription factors and for the genetic manipulation of plant traits. For CRES-T, a transcription factor is converted to a strong repressor by fusion with an SRDX repression domain, which is then expressed in plants to induce a loss-of-function phenotype. However, the traditional CRES-T vectors are inconvenient for gene cloning and promoter exchange. In this study, we developed new CRES-T vectors that are efficient and convenient to use by employing the Gateway system, a new vector backbone and a terminator derived from the heat shock protein 18.2 (HSP) gene. Our test experiments revealed that the CRES-T vector containing the Gateway linker sequence within the transcribed region showed reduced efficiency of CRES-T when compared with the traditional CRES-T vector. However, the HSP terminator compensated for the negative effect of the Gateway sequence and improved the efficiency of CRES-T in all cases tested and resulted in the highest efficiency achieved to date. We found that the HSP terminator increased transcription efficiency or transcript stability; in contrast, these factors were negatively affected by the Gateway linker sequence in our vector system. We, therefore, propose that the appropriate CRES-T vector should be chosen depending on situations and purposes.
Flavonoids are important for male gametophyte development. Here we report on the flavonoid components and their biosynthetic regulation in gentian anthers. Among flavonoids, flavonols, including kaempferol, quercetin and isorhamnetin derivatives, accumulated abundantly in gentian anthers. However, flavones and anthocyanins, which are the main flavonoids accumulating in petals, were not detected. Northern blot analysis of nine flavonoid biosynthetic genes showed that the ‘early’ flavonoid biosynthetic genes were expressed in both anthers and petals, and that flavonol synthase (FLS) transcripts were restricted to anthers. In contrast, flavone synthase II (FNSII) and ‘late’ flavonoid biosynthetic genes were expressed specifically in gentian petals. To confirm anther-specific expression of FLS, the 5′-upstream region of FLS (GtFLSpro) was cloned by inverse PCR and fused to the uidA (GUS) reporter gene. Tobacco, Arabidopsis and gentian plants, transformed with the GtFLSproGUS construct, exhibited anther-specific GUS expression. Expression was observed in the tapetum and in pollen at late stages of anther development in transgenic plants. These results revealed that flavonol accumulation in gentian anthers was regulated by the spatial expression of GtFLS. Our results also suggest that anther-specific regulation of FLS is conserved among higher plants and the GtFLS promoter is useful for induction of specific gene expression in anthers.
We have found that cauliflower mosaic virus (CaMV) 35S promoter-specific transgene silencing is mediated by DNA methylation in gentian (Gentiana triflora × G. scabra). De novo methylation of asymmetric cytosines (CpHpH; where H is A, C, or T) sequence has been detected at the enhancer region (−148 to −85) of the 35S promoter in transgenic gentians, and is thought to be responsible for the silencing mechanism. To clarify the concept of de novo methylation, the present study examined the detailed DNA methylation profile of the entire T-DNA sequence (ca. 4 kb) integrated into transgenic gentians. Although highly methylated cytosines at CpG and CpWpG (W is A or T) sequences were broadly distributed, except in the sGFP coding region, highly methylated cytosines at CpHpH and CpCpG sequences were mainly limited to the 35S enhancer region. In addition to the previously identified de novo methylation peak (−148 to −85), another peak was discovered at −298 to −241. Electrophoretic mobility shift assays showed that gentian nuclear extracts could bind to the corresponding probes (−149 to −124 and −275 to −250), and that the probes could compete with one another for binding. Thus, a nuclear factor might be involved in the de novo methylation of the two regions. In addition, the present data indicated that the methylation patterns at CpCpG sites could be categorized as CpHpH methylation rather than CpWpG methylation.
4,2′,4′,6′-Tetrahydroxychalcone (THC) 2′-glucoside that confers yellow color to the petals of carnation, cyclamen, and catharanthus is biosynthesized by the action of UDP-glucose-dependent THC 2′-glucosyltransferase (THC2′GT). We isolated 18 types of full-length cDNA encoding GT-like sequences from carnation petals. Expression of these cDNA in Escherichia coli identified five cDNAs encoding THC2′GT that were different from the previously isolated THC2′GT. We also isolated a cDNA encoding THC2′GT from both catharanthus and cyclamen. These THC2′GT cDNAs were introduced to petunia. Transgenic petunia that expressed three of the GTs produced THC 2′-glucoside, which indicated that they function as THC2′GT in vivo. These cDNAs could be useful molecular tools to yield yellow flower color, although the amount accumulated in the transgenic petals was too small to alter the flower color in this study.
Rose petals contain 3-glucosylated anthocyanidins and flavonols. We isolated three flavonoid 3-glucosyltransferase (UF3GT) homolog genes (RhUF3GT1, RhUF3GT2, and RhUF3GT3) from rose. RhUF3GT1 encoded an amino acid sequence that is almost identical to the reported rose partial UF3GT homologs and highly homologous to strawberry and apple UF3GTs. Recombinant RhUF3GT1 expressed in yeast catalyzes 3-glucosylation of anthocyanidins but not flavonols. RhUF3GT1 was not expressed in the petals of many cultivars even when anthocyanin biosynthesis was active, while it was expressed in the mature petals of cultivars that synthesize cyanidin 3-glucoside in the mature petals. RhUF3GT2 and RhUF3GT3, sharing 79% identity, exhibit only 42% and 41% identities to RhUF3GT1, respectively, and are distantly related to strawberry and apple UF3GTs. They were expressed in coordination with the flavonol synthase gene in the petal. The recombinant RhUF3GT2 expressed in yeast catalyzed 3-glucosylation of flavonol much more efficiently than that of anthocyanidins. We suggest that RhUF3GT2 catalyzes flavonol 3-glucosylation in rose petals and that it also contributes to accumulation of anthocyanidin 3-glucoside in the petals.
An important part of the assessment of the potential environmental impact from the introduction of a genetically modified (GM) plant is an evaluation of the potential for gene flow from the GM plant to related wild species. This information is needed as part of the risk-assessment process, in the context of whether gene flow to wild species is possible. One method for evaluating gene flow is to use molecular techniques to identify genes in wild species populations that may have originated from a cultivated species. An advantage of this method is that a phenotypic marker or trait is not required to measure gene flow. In the present study we analyzed the seedlings of seeds from three wild native Rosa species (R. multiflora Thunb., R. luciae Rochebr. et Franch. ex Crép. and R. rugosa Thunb.) selected from several locations across Japan where the wild rose was growing in close proximity to cultivated rose plants (Rosa×hybrida). To determine whether gene flow from cultivated rose had occurred, young leaves of 1,296 seedlings from the wild Rosa plants were analyzed by PCR for the presence of the KSN locus. This locus originated from a sport of R. chinensis Jacq. var. spontanea (Rehd. et Wils.) Yu et Ku and is involved in the recurrent flowering phenotype observed for cultivated rose hybrids, but is absent in Japanese species roses. The KSN locus was absent in all seedlings sampled, indicating no gene flow to wild Rosa species from the cultivated rose had occurred, and providing evidence that the probability of gene flow from cultivated to wild Rosa species in Japan is low or non-existent.
The release of genetically modified plants into the environment can only occur after permission is obtained from the relevant regulatory authorities. This permission will only be obtained after extensive risk assessment shows comparable risk of impact to the environment and biodiversity as compared to non-transgenic host plants. Two transgenic rose (Rosa×hybrida) lines, whose flowers were modified to a bluer colour as a result of accumulation of delphinidin-based anthocyanins, have been trialed in greenhouses and the field in both Japan and Australia. Flower colour modification was due to expression of genes of a viola flavonoid 3′,5′-hydroxylase and a torenia anthocyanin 5-acyltransferase. In all trials it was shown that the performance of the two transgenic lines, as measured by their growth characters, was comparable to the host untransformed variety. Biological assay showed that the transgenic lines did not produce allelopathic compounds. In Japan, seeds from wild rose species that had grown in close proximity to the transgenic roses did not carry either a Rosa×hybrida specific marker gene or the transgenes. In hybridization experiments using transgenic rose pollen and wild rose female parents, the transgenes were not detected in the seed obtained, though there was a low frequency of seed set. The transgene was also not transmitted when Rosa×hybrida cultivars were used as females. In in situ hybridization analysis transgene transcripts were only detected in the epidermal cells in the petals of the transgenic roses. In combination, the breeding and in situ analysis results show that the transgenic roses contain the transgene only in the L1 layer cells and not in the L2 layer cells that generate reproductive cells. General release permissions have been granted for both transgenic lines in Japan and one is now commercially produced.
While torenia (Torenia fournieri Lind.) is a useful model flower for molecular biological studies of floral architecture, the maintenance of plant materials and resulting transgenic plants requires vegetative propagation due to its heterozygous nature. Reduction of labor and costs for maintaining thousands of in vitro torenia cultures is therefore a critical issue. We found that substituting trehalose for sucrose drastically extended the culture period to 70 days, which is more than twice as long as for the common, sucrose-based medium, without reduction in plant viability. Comparative measurement of the plant mass indicated that the increased survival benefit of the trehalose-based medium might be on account of improvement in the rhizosphere environment through reduction of root density in the culture, rather than by reduced plant growth. No harmful effects arising from the trehalose-based medium were observed in 1,800 laboratory lines during the bimonthly subculture for over 12 months, except for a wilting on the first transfer to the trehalose-based medium. In conjunction with the use of the commercial food additive, Okome-ni-TREHA® rather than reagent-grade trehalose, we have succeeded in reducing the costs and labor associated with the culture medium to less than one third of those for the sucrose-based system.