Hong-Gyu Kang

T-DNA Insertional Mutagenesis for Activation Tagging in Rice

We have developed a new T-DNA vector, pGA2715, which can be used for promoter trapping and activation tagging of rice (Oryza sativa) genes. The binary vector contains the promoterlessβ-glucuronidase (GUS) reporter gene next to the right border. In addition, the multimerized transcriptional enhancers from the cauliflower mosaic virus 35S promoter are located next to the left border. A total of 13,450 T-DNA insertional lines have been generated using pGA2715. Histochemical GUS assays have revealed that the GUS-staining frequency from those lines is about twice as high as that from lines transformed with the binary vector pGA2707, which lacks the enhancer element. This result suggests that the enhancer sequence present in the T-DNA improves the GUS-tagging efficiency. Reverse transcriptase-PCR analysis of a subset of randomly selected pGA2715 lines shows that expression of the genes immediately adjacent to the inserted enhancer is increased significantly. Therefore, the large population of T-DNA-tagged lines transformed with pGA2715 could be used to screen for promoter activity using the gusreporter, as well as for creating gain-of-function mutants.

IDENTIFICATION OF CLASS B AND CLASS C FLORAL ORGAN IDENTITY GENES FROM RICE PLANTS

The functions of two rice MADS-box genes were studied by the loss-of-function approach. The first gene, OsMADS4, shows a significant homology to members in the PISTILLATA (PI) family, which is required to specify petal and stamen identity. The second gene, OsMADS3, is highly homologous to the members in the AGAMOUS (AG) family that is essential for the normal development of the internal two whorls, the stamen and carpel, of the flower. These two rice MADS box cDNA clones were connected to the maize ubiquitin promoter in an antisense orientation and the fusion molecules were introduced to rice plants by the Agrobacterium-mediated transformation method. Transgenic plants expressing antisense OsMADS4 displayed alterations of the second and third whorls. The second-whorl lodicules, which are equivalent to the petals of dicot plants in grasses, were altered into palea/lemma-like organs, and the third whorl stamens were changed to carpel-like organs. Loss-of-function analysis of OsMADS3 showed alterations in the third and fourth whorls. In the third whorl, the filaments of the transgenic plants were changed into thick and fleshy bodies, similar to lodicules. Rather than making a carpel, the fourth whorl produced several abnormal flowers. These phenotypes are similar to those of the agamous and plena mutants in Arabidopsis and Antirrhinum, respectively. These results suggest that OsMADS4 belongs to the class B gene family and OsMADS3 belongs to the class C gene family of floral organ identity determination.

Generation and Analysis of End Sequence Database for T-DNA Tagging Lines in Rice

We analyzed 6,749 lines tagged by the gene trap vector pGA2707. This resulted in the isolation of 3,793 genomic sequences flanking the T-DNA. Among the insertions, 1,846 T-DNAs were integrated into genic regions, and 1,864 were located in intergenic regions. Frequencies were also higher at the beginning and end of the coding regions and upstream near the ATG start codon. The overall GC content at the insertion sites was close to that measured from the entire rice (Oryza sativa) genome. Functional classification of these 1,846 tagged genes showed a distribution similar to that observed for all the genes in the rice chromosomes. This indicates that T-DNA insertion is not biased toward a particular class of genes. There were 764, 327, and 346 T-DNA insertions in chromosomes 1, 4 and 10, respectively. Insertions were not evenly distributed; frequencies were higher at the ends of the chromosomes and lower near the centromere. At certain sites, the frequency was higher than in the surrounding regions. This sequence database will be valuable in identifying knockout mutants for elucidating gene function in rice. This resource is available to the scientific community at http://www.postech.ac.kr/life/pfg/risd.

Identification of a rice APETALA3 homologue by yeast two-hybrid screening.

A cDNA clone OsMADS16 was isolated from the rice young inflorescence cDNA expression library by the yeast two-hybrid screening method with OsMADS4 as bait. We have previously shown that the OsMADS4 gene is a member of the PI family and that the MADS-box gene is involved in controlling development of the second and third whorls of rice flowers. The sequence comparison indicated that OsMADS16 belongs to the AP3 family. The OsMADS16 protein contains a PI-derived motif, FAFRVVPSQPNLH, that is a conserved sequence in AP3 family genes at the C-terminal region. In addition, OsMADS16 contains a paleoAP3 motif, YGGNHDLRLG, downstream of the PI-derived motif. The paleoAP3 motif is a consensus sequence in the C-terminal region of the AP3 family genes of lower eudicot and magnolid dicot species. RNA blot analysis showed that the OsMADS16 gene was expressed in the second and third whorls, whereas the OsMADS4 transcripts were present in the second, third, and fourth whorls. These expression patterns of the OsMADS16 and OsMADS4 genes are very similar to those of AP3 and PI, respectively. In the yeast two-hybrid system, OsMADS4 interacted only with OsMADS16 among several rice MADS genes investigated, suggesting that OsMADS4 and OsMADS16 function as a heterodimer in specifying sepal and petal identities. The OsMADS16 protein displayed transcription activation ability in yeast, whereas AP3 did not. It was also shown in yeast that OsMADS16 interacted with PI whereas OsMADS4 did not interact with AP3. These differences between OsMADS16 and AP3 indicate that the functions of the AP3 family genes of monocots and dicots diverged during molecular evolution processes of the B function genes. Deletion analysis showed that the 155–200 amino acid region of the OsMADS16 protein plays an important role in the transcription activation ability.

Phenotypic alterations of petal and sepal by ectopic expression of a rice MADS box gene in tobacco

Floral organ development is controlled by a group of regulatory factors containing the MADS domain. In this study, we have isolated and characterized a cDNA clone from rice, OsMADS3, which encodes a MADS-domain containing protein. The OsMADS3 amino acid sequence shows over 60% identity to AG of Arabidopsis, PLE of Antirrhinum majus, and AG/PLE homologues of petunia, tobacco, tomato, Brassica napus, and maize. Homology in the MADS box region is most conserved. RNA blot analysis indicated that the rice MADS gene was preferentially expressed in reproductive organs, especially in stamen and carpel. In situ localization studies showed that the transcript was present primarily in stamen and carpel. The function of the rice OsMADS3 was elucidated by ectopic expression of the gene under the control of the CaMV 35S promoter in a heterologous tobacco plant system. Transgenic plants exhibited an altered morphology and coloration of the perianth organs. Sepals were pale green and elongated. Limbs of the corolla were split into sections which in some plants became antheroid structures attached to tubes that resembled filaments. The phenotypes mimic the results of ectopic expression of dicot AG gene or AG homologues. These results indicate that the OsMADS3 gene is possibly an AG homologue and that the AG genes appear to be structurally and functionally conserved between dicot and monocot.

Identification of the ADP-glucose pyrophosphorylase isoforms essential for starch synthesis in the leaf and seed endosperm of rice (Oryza sativa L.)

ADP-glucose pyrophosphorylase (AGP) catalyzes the first committed step of starch biosynthesis in higher plants. To identify AGP isoforms essential for this biosynthetic process in sink and source tissues of rice plants, we analyzed the rice AGP gene family which consists of two genes, OsAGPS1 and OsAGPS2, encoding small subunits (SSU) and four genes, OsAGPL1, OsAGPL2, OsAGPL3 and OsAGPL4, encoding large subunits (LSU) of this enzyme heterotetrameric complex. Subcellular localization studies using green fluorescent protein (GFP) fusion constructs indicate that OsAGPS2a, the product of the leaf-preferential transcript of OsAGPS2, and OsAGPS1, OsAGPL1, OsAGPL3, and OsAGPL4 are plastid-targeted isoforms. In contrast, two isoforms, SSU OsAGPS2b which is a product of a seed-specific transcript of OsAGPS2, and LSU OsAGPL2, are localized in the cytosol. Analysis of osagps2 and osagpl2 mutants revealed that a lesion of one of the two cytosolic isoforms, OsAGPL2 and OsAGPS2b, causes a shrunken endosperm due to a remarkable reduction in starch synthesis. In leaves, however, only the osagps2 mutant appears to severely reduce the transitory starch content. Interestingly, the osagps2 mutant was indistinguishable from wild type during vegetative plant growth. Western blot analysis of the osagp mutants and wild type plants demonstrated that OsAGPS2a is an SSU isoform mainly present in leaves, and that OsAGPS2b and OsAGPL2 are the major SSU and LSU isoforms, respectively, in the endosperm. Finally, we propose a spatiotemporal complex model of OsAGP SSU and LSU isoforms in leaves and in developing endosperm of rice plants.

Characterization of two rice MADS box genes homologous to GLOBOSA

Abstract A group of regulatory factors containing the MADS box domain is playing an important role in controlling floral organ induction and development. In this study, we have isolated and characterized two cDNA clones from rice, OsMADS2 and OsMADS4, which encode MADS-domain containing proteins. The OsMADS2 amino acid sequence shows more than 50% identity to GLO of Antirrhinum majus, PI of Arabidopsis, and GLO PI homologs of petunia and tobacco. Also, these two rice proteins contain, in addition to the conserved MADS box sequence, two other conserved sequences in the same locations as are found specifically in the GLO PI family. RNA blot analysis showed that the rice MADS box containing genes were expressed throughtout flower development. While OsMADS2 expression increased strongly as the flower matured, OsMADS4 expression was at a higher level from early stages of flower development. RNA blot hybridization with total RNA from each floral organ showed that the expression of these genes was restricted to anther and carpel. In situ hybridization studies indicated that the transcripts were present mainly in pollen, tapetum and also in stigma. Since rice flowers consist of a single perianth which resembles sepal, GLO PI homologs in rice would be expected to be expressed predominantly in reproductive organs. These results suggest that the rice OsMADS genes may be members of the GLO PI family.

Generation of T-DNA tagging lines with a bidirectional gene trap vector and the establishment of an insertion-site database

We have developed a binary T-DNA vector, pGA2717, that contains the promoter-less β-glucuronidase (gus) gene adjacent to the right border and the promoter-less green fluorescence protein (gfp) gene next to the left border of the T-DNA. Therefore, inserting T-DNA into a gene can result in the activation of either gus or gfp. A total of 12 169 T-DNA insertional lines of japonica rice were generated using this binary vector. Out of 3140 lines examined, 0.5% of their mature seeds and 2.0% of the 3-day-old etiolated seedlings were GFP-positive. However, GUS assays of the same materials resulted in the identification of 151 (4.8%) GUS-positive lines. Using DNA gel blot and reverse transcription (RT)-PCR analyses, we confirmed that the GFP-positive lines were a true indication of gene trapping. A fusion transcript was also obtained between gfp and the trapped gene. We isolated 990 genomic sequences flanking T-DNA from our analysis of 2099 transgenic plants. Among the insertions, 625 T-DNAs were integrated into genic regions; 361 were located in intergenic regions. These tagging lines will be valuable in trapping and studying various genes for their expression patterns, as well as providing a useful tool for genetic approaches.

Overexpression of FTL1/DDF1, an AP2 transcription factor, enhances tolerance to cold, drought, and heat stresses in Arabidopsis thaliana.

Freezing temperatures control where and when plants can grow, and negatively influence crop quality and productivity. To identify key regulatory genes involved in cold adaptation, we screened activation-tagged Arabidopsis lines for mutants with greater freezing tolerance. One mutant, freezing tolerant line1 (ftl1-1D), manifested enhanced tolerance along with dwarfism and delayed flowering. This was caused by activation of DWARF AND DELAYED FLOWERING 1 (DDF1), a gene previously described as a regulatory component in salinity signaling. The induced gene encoded an AP2 transcription factor of the CBF/DREB1 subfamily. In addition to conferring tolerance to low temperatures and salt stress, ftl1-1D/ddf1 improved tolerance to drought and heat. Real-time PCR indicated that FTL1/DDF1 was up-regulated by those four types of stresses in wild-type Arabidopsis. Its increased expression in the mutant induced various stress-responsive genes under normal growing conditions, resulting in improved tolerances. However, phenotypes shown in the ftl1-1D/ddf1 were restored by treatment with exogenous gibberellin (GA₃), indicating the involvement of a GA pathway in FTL1/DDF1-mediated tolerance. Therefore, we conclude that FTL1/DDF1 plays a role in regulating responses to several abiotic stresses, perhaps via cross-talk in the pathways.

STM study of Ar+-induced defects produced by near-threshold energy collision

Abstract We have investigated graphite surfaces impacted with Ar + ions of near-threshold energies (40–100 eV) using STM. In order to understand the effect of Ar atoms trapped between the graphite basal planes on STM image, the trapped atoms are evaporated by heating the surface at 600°C after 50 eV Ar + impact. Ion impact creates hillocks of various size in STM images, which originate from the enhanced charge density at vacancy and interstitial defect sites. The yields for total defect production, vacancy production and Ar trapping upon ion impact are determined at an energy of 50–100 eV. The threshold energies for vacancy production and trapping are deduced from these data.