SUMMARYThe natural bicolor floral traits of the horticultural petunia (Petunia hybrida) cultivars Picotee and Star are caused by the spatial repression of the chalcone synthase A (CHS-A) gene, which encodes an anthocyanin biosynthetic enzyme. Here we show that Picotee and Star petunias carry the same short interfering RNA (siRNA)-producing locus, consisting of two intact CHS-A copies, PhCHS-A1 and PhCHS-A2, in a tandem headto-tail orientation. The precursor CHS mRNAs are transcribed from the two CHS-A copies throughout the bicolored petals, but the mature CHS mRNAs are not found in the white tissues. An analysis of small RNAs revealed the accumulation of siRNAs of 21 nucleotides that originated from the exon 2 region of both CHS-A copies. This accumulation is closely correlated with the disappearance of the CHS mRNAs, indicating that the bicolor floral phenotype is caused by the spatially regulated post-transcriptional silencing of both CHS-A genes. Linkage between the tandemly arranged CHS-A allele and the bicolor floral trait indicates that the CHS-A allele is a necessary factor to confer the trait. We suppose that the spatially regulated production of siRNAs in Picotee and Star flowers is triggered by another putative regulatory locus, and that the silencing mechanism in this case may be different from other known mechanisms of post-transcriptional gene silencing in plants. A sequence analysis of wild Petunia species indicated that these tandem CHS-A genes originated from Petunia integrifolia and/or Petunia inflata, the parental species of P. hybrida, as a result of a chromosomal rearrangement rather than a gene duplication event.

Three ethylene receptor genes, DC-ERS1, DC-ERS2 and DC-ETR1, were previously identified in carnation (Dianthus caryophyllus L.). Here, the presence of mRNAs for respective genes in flower tissues and their changes during flower senescence are investigated by Northern blot analysis. DC-ERS2 and DC-ETR1 mRNAs were present in considerable amounts in petals, ovaries and styles of the flower at the full-opening stage. In the petals the level of DC-ERS2 mRNA showed a decreasing trend toward the late stage of flower senescence, whereas it increased slightly in ovaries and was unchanged in styles throughout the senescence period. However, DC-ETR1 mRNA showed no or little changes in any of the tissues during senescence. Exogenously applied ethylene did not affect the levels of DC-ERS2 and DC-ETR1 mRNAs in petals. Ethylene production in the flowers was blocked by treatment with 1,1-dimethyl-4-(phenylsulphonyl)semicarbazide (DPSS), but the mRNA levels for DC-ERS2 and DC-ETR1 decreased in the petals. DC-ERS1 mRNA was not detected in any cases. These results indicate that DC-ERS2 and DC-ETR1 are ethylene receptor genes responsible for ethylene perception and that their expression is regulated in a tissue-specific manner and independently of ethylene in carnation flowers during senescence.

Anthocyanin O-methyltransferase (OMT) is one of the key enzymes for anthocyanin modification and flower pigmentation. We previously bred a novel red-purple-flowered fragrant cyclamen (KMrp) from the purple-flowered fragrant cyclamen 'Kaori-no-mai' (KM) by ion-beam irradiation. Since the major anthocyanins in KMrp and KM petals were delphinidin 3,5-diglucoside and malvidin 3,5-diglucoside, respectively, inactivation of a methylation step in the anthocyanin biosynthetic pathway was indicated in KMrp. We isolated and compared OMT genes expressed in KM and KMrp petals. RT-PCR analysis revealed that CkmOMT2 was expressed in the petals of KM but not in KMrp. Three additional CkmOMTs with identical sequences were expressed in petals of both KM and KMrp. Genomic PCR analysis revealed that CkmOMT2 was not amplified from the KMrp genome, indicating that ion-beam irradiation caused a loss of the entire CkmOMT2 region in KMrp. In vitro enzyme assay demonstrated that CkmOMT2 catalyzes the 3' or 3',5' O-methylation of the B-ring of anthocyanin substrates. These results suggest that CkmOMT2 is functional for anthocyanin methylation, and defective expression of CkmOMT2 is responsible for changes in anthocyanin composition and flower coloration in KMrp.

Genetic improvement of flower vase life is an important breeding target. A breeding program aimed to improve the vase life of carnations (Dianthus caryophyllus L.) that carried out by repeatedly crossing and selecting promising progenies was effective. Thus, we developed many carnation lines with extremely long vase life by using the conventional cross-breeding technique. Selected lines from second-, third-and fourthgeneration have vase lives that are 2.7 to 4.1 times longer than the control cultivar, 'White Sim'. The mean vase life of line 108-44 was 23.6 days in 2003 and 19.1 days in 2004 (414% and 341% of 'White Sim' flower longevity, respectively). All selected lines showed low ethylene production in whole flowers during senescence. In addition, the petals and the gynoecia of three selected lines 908-46, 702-21, and 006-13, produced only trace amounts of ethylene during senescence, indicating that the ethylene biosynthesis pathway in these lines was almost completely blocked during senescence. Although ethylene sensitivity of the selected lines was generally high on day 0, immediately after harvesting, ethylene sensitivity after anthesis rapidly decreased with age in the three selected lines. These lines became completely ethylene-insensitive or showed low sensitivity at the end of senescence. Autocatalytic ethylene production by the gynoecia and petals of the three selected lines also decreased on days 3 and 6, respectively, after harvest.

The yellow coloration mechanism of Camellia chrysantha flowers has been a mystery because the pigments which produce their deep yellow color could not be found in the flower. We sought to solve the mystery regarding the characteristic accumulation of aluminum ions by camellia plants. Deep yellow C. chrysantha flowers contained aluminum ions at three times the concentration found in pale yellow flowers. Three quercetin derivatives, 3-rutinoside, 3-glucoside, and 7-glucoside, were identified as major flavonoids in both flowers. There were no significant differences in their flavonoid content or pH, which was 5.8. The deep yellow flowers of C. chrysantha contained flavonoids and aluminum ions in a ratio of 1 to 0.5. When quercetin 3-rutinoside solution, which was originally almost colorless and had an absorption maximum around 350 nm at pH 5.8, was prepared with an aluminum ion concentration similar to the endogenous ratio, the solution developed a deep yellow color. This mixture also had an absorption spectrum like that of C. chrysantha petals, which had an absorption maximum around 420 nm. After removing the cations using an ion exchange column, the yellow coloration of the C. chrysantha petal extract changed to a paler color; the coloration was restored by the addition of aluminum ions. We concluded that the deep yellow color of the C. chrysantha flower is generated by the chelation of aluminum ions by quercetin derivatives, which is a unique yellow 'dyeing' system of these flowers.