The main interest of the group is in understanding the biology of the dynamic genome, namely, genome organization and reorganization and its impact on gene expression and regulation. We are also characterizing various aspects of genetic and epigenetic gene regulations particularly on flower pigmentation of morning glories. In addition, we are undertaking reverse genetic approaches in order to elucidate the nature of dynamic genome in plants.

I. Spontaneous mutants in morning glories.

Considerable attention has recently been paid to the morning glory genus Ipomoea because of the experimental versatility of its floral biology including the genetics of floral variation, flavonoid biosynthesis, and transposon-induced mutations. The genus Ipomoea includes about 600 species distributed on a worldwide scale that exhibit various flower morphologies and pigmentation patterns. A large number of Ipomoea species can be found in the Americas, particularly in Mexico. Among the genus Ipomoea, three morning glories, Ipomoea nil (the Japanese morning glory), Ipomoea purpurea (the common morning glory), and Ipomoea tricolor, were domesticated well as floricultural plants, and many mutants displaying various flower pigmentation patterns were isolated. Among these morning glories, I. nil and I. purpurea belong to the same subgenera Ipomoea, whereas I. tricolor was classified into another subgenera, Quamoclit. Both I. nil and I. tricolor display blue flowers (see Figure 2A) that contain the peonidin (3’ methoxyl cyanidin) derivative named Heavenly Blue Anthocyanin (HBA), and I. purpurea produces dark purple flowers containing a cyanidin derivative that lacks one glucose molecule and a methyl residue from HBA. All of them produce dark-brown seeds. As floricultural plants, spontaneous mutants of I. purpurea and I. tricolor displaying either white or reddish flowers were obtained while various spontaneous mutants of I. nil exhibiting many different flower colorations were generated.

Although I. nil has long been believed to have originated from Southeast Asia, a hypothesis that I. nil may have arrived in Asia from tropical America was also recently proposed. In either case, the plant had been introduced into Japan from China approximately in the 8th Century as a medicinal herb. The seeds were used as a laxative, and the plant became a traditional floricultural plant in Japan around the 17th Century. I. nil has an extensive history of genetic studies, and a number of its spontaneous mutants related to the color and shape of the flowers have been isolated. It was also used extensively for physiological studies of the photoperiodic induction of flowering.

The wild-type I. purpurea plant, which originated from Central America, was probably introduced to Europe in the 17th Century, and cultivars with white and red flowers were already recorded in the late 18th Century. Like I. purpurea, I. tricolor originated from Central America, and several spontaneous mutants exhibiting various flower pigmentations were obtained in the mid-20th Century.

As Figure 1 shows, the spontaneous mutants of I. nil, I. purpurea, and I. tricolor carrying the magenta, pink, and fuchsia alleles, respectively, produce reddish flowers containing pelargonidin derivatives, and all of them are deficient in the gene for flavonoid 3’-hydroxylase (F3’H). The magenta allele in I. nil is a nonsense mutation caused by a single C to T base transition generating the stop codon TGA, and the cultivar Violet carries the same mutation. Several tested pink mutants in I. purpurea carry inserts of the 0.55-kb DNA transposable element Tip201 belonging to the Ac/Ds superfamily at the identical site. No excision of Tip201 from the F3’H gene could be detected, and both splicing and polyadenylation patterns of the F3’H transcripts were affected by the Tip201 integration. The fuchsia allele in I. tricolor is a single T insertion generating the stop codon TAG, and the accumulation of the F3’H transcripts was drastically reduced by the nonsense-mediated RNA decay.

The I. tricolor spontaneous mutant Blue Star carrying the mutable ivory seed-variegated allele exhibits pale-blue flowers with a few fine blue spots and ivory seeds with tiny dark-brown spots (Figure 2). The mutable allele is caused by an intragenic tandem duplication of 3.3 kb within a gene for transcriptional activator containing a bHLH DNA-binding motif. Each of the tandem repeats is flanked by a 3-bp sequence AAT, indicating that the 3-bp microhomology is used to generate the tandem duplication. The transcripts in the pale-blue flower buds of the mutant contain an internal 583-bp tandem duplication that results in the production of a truncated polypeptide lacking the bHLH domain. The mRNA accumulation of most of the structural genes encoding enzymes for anthocyanin biosynthesis in the flower buds of the mutant was significantly reduced. The transcripts identical to the wild-type mRNAs for the transcriptional activator were present abundantly in blue spots of the variegated flowers, whereas the transcripts containing the 583-bp tandem duplication were predominant in the pale-blue background of the same flowers. The flower and seed variegations are caused by somatic homologous recombination between an intragenic tandem duplication in the bHLH gene, whereas various flower variegations in Ipomoea are known to be caused by excision of DNA transposons inserted into pigmentation genes.

Fig.1. Flower phenotypes and structures of the F3’H genes in I. nil, I. purpurea, and I. tricolor exhibiting reddish flowers. A. I. nil carrying the magenta allele. B. I. purpurea carrying the pink allele. C. I. tricolor carrying the fuchsia allele.

Fig.2. Flower and seed pigmentation phenotypes and structures of the bHLH transcriptional regulatortor gene for anthocyanin pigmentation in the I. tricolor wild-type cultivar, Heavenly Blue (A and C), and its mutant, Blue Star (B, D and E).

II. Targeted gene disruption by homologous recombination in rice.

Rice (Oryza sativa L.) is an important staple food for more than half of the world’s population and a model plant for other cereal species. We have developed a large-scale Agrobacterium-mediated transformation procedure with a strong positive-negative selection and succeeded in efficient and reproducible targeting of the Waxy gene by homologous recombination without concomitant occurrence of ectopic events, which must be an important first step for developing a precise modification system of the genomic sequences in rice. By improving our transformation procedure further, we are attempting to modify Adh genes, which belong to a small multigene family and reside adjacent to repetitive retroelements.

III. Characterization of mutable virescent allele in rice.

Leaves of seedlings in the virescent mutant of rice are initially pale yellow green due to partial deficient in chlorophyll and gradually become green with the growth of the mutant. We have been characterizing a spontaneous mutable virescent allele pale yellow leaf-variegated (pyl-v), conferring pale yellow leaves with dark green sectors in its seedlings (Figure 3). The pyl-v mutant was isolated among progeny of a hybrid between indica and japonica rice plants. The leaf variegation is regarded as a recurrent somatic mutation from the recessive pale yellow allele to the dark green revertant allele. The availability of the genomic sequences of both japonica and indica subspecies facilitates map-based cloning of the pyl-v allele. Mapping data indicate the mutation resides in the short arm of the chromosome 3, and excision of a new DNA transposon from the pyl gene appears to be responsible for conferring the leaf variegation. It is important to emphasize here that tissue culture is necessary in all of the currently available rice reverse genetic approaches. No somaclonal variation is likely to occur in mutant lines induced by our newly characterizing endogenous element, because no tissue culture has been involved in its activation. Using a newly identified DNA transposon, therefore, we are attempting to develop a novel transposon tagging system in rice.

Fig.3. A. Leaf phenotype of the mutable pyl-v allele. B. Identification of the pyl-v allele. The pyl-v allele was mapped between the SSR markers RM251 and RM282 on chromosome 3 and subsequently located between the SLP markers CH30524-1 and MMCAPS-1 on a single BAC clone. The horizontal pentagonals represent predicted genes.