The Horticulture Journal
Online ISSN : 2189-0110
Print ISSN : 2189-0102
ISSN-L : 2189-0102
ORIGINAL ARTICLES
Conversion of Abaxial to Adaxial Petal in a Torenia (Torenia fournieri Lind. ex Fourn.) Mutant Appeared in Selfed Progeny of the Mutable Line “Flecked”
Tomoya NikiKatsutomo SasakiMasahito ShikataTakako Kawasaki-NarumiNorihiro OhtsuboTakaaki Nishijima
Author information
JOURNAL OPEN ACCESS FULL-TEXT HTML

2016 Volume 85 Issue 4 Pages 351-359

Details
Abstract

We isolated a torenia mutant “Begonia” from selfed progeny of the mutable line “Flecked,” in which the ventral petal of the flower was converted into dorsal petal. In normal-type (NT) flowers, dorsal petals were pale violet and limbs of lateral and ventral petals were violet, whereas the ventral petal had a yellow nectar guide. In contrast, the ventral petal of mutant-type (MT) flowers changed to pale violet, and the nectar guide disappeared. These altered pigmentation patterns were observed from the early stage of corolla pigmentation. Expression analyses of the floral symmetry genes CYCLOIDEA (CYC), RADIALIS (RAD), and DIVARICATA (DIV) showed that TfCYC1, TfCYC2, TfCYC3, and TfRAD1 were mainly expressed in dorsal petals of NT flowers, but in the mutant, these genes were expressed in the ventral petal similar to dorsal ones. These results suggest that conversion of ventral to dorsal petal in the “Begonia” mutant is caused by high expression of TfCYC1, TfCYC2, TfCYC3, and TfRAD1 in the ventral petal, comparable to their expression in dorsal petals.

Introduction

For floricultural plants, variation in flower morphology, such as single- and double-flower and various flower sizes, is one of the most important means of increasing market size. Novel flower morphology developed by breeding is attractive to consumers. Flower symmetry is an important morphological factor influencing the appearance and impression of flowers (Endress, 1999; Hileman, 2014), and is classified by the number of symmetry axes passing through the corolla center. Actinomorphic flowers have multiple (at least two) symmetry axes, zygomorphic flowers have a single axis, and asymmetrical flowers have no symmetry axis (Hileman, 2014). Zygomorphic flowers evolved from actinomorphic flowers to attract specialized pollinators in a step that may have occurred independently in many species of angiosperms (Donoghue et al., 1998; Endress, 1999). Torenia (Torenia fournieri Lind. ex Fourn.) bears zygomorphic flowers. No cultivars with actinomorphic flowers have been bred. There are neither double-flowered nor large-flowered cultivars. There is, thus, little variation in flower morphology in torenia.

The mechanisms regulating flower symmetry, especially dorsoventral asymmetry, have been well studied in Antirrhinum majus, and include a regulatory network of related genes (Preston and Hileman, 2009). Identity of dorsal petals is established by genes encoding two closely-related TCP transcription factors: CYCLOIDEA (CYC) and DICHOTOMA (DICH), and a MYB transcription factor, RADIALIS (RAD), which are expressed specifically in dorsal petals (Corley et al., 2005; Costa et al., 2005; Luo et al., 1996, 1999). The functions of these genes have been identified by mutant analyses. Mutation of cyc converts a part of dorsal and lateral petals into lateral and ventral petals, respectively (Luo et al., 1996). Although flower phenotype is unchanged in dich single mutants, all petals are ventralized in cyc:dich double mutants and rad mutants (Corley et al., 2005; Luo et al., 1999). Identity of the ventral petal is established by DIVARICATA (DIV), which encodes another MYB transcription factor (Galego and Almeida, 2002; Rose et al., 1999). This is shown by the observation that the ventral petal is converted into a lateral one in div mutants (Almeida et al., 1997; Galego and Almeida, 2002). RAD is induced by CYC (Costa et al., 2005), while RAD competitively inhibits DIV function (Cui et al., 2010; Galego and Almeida, 2002). Thus, in dorsal petals, inhibition of DIV function resulting from induction of RAD caused by CYC establishes dorsal petal identity.

Mutations caused by transposable elements contribute greatly to diversification in floricultural plants. For example, modified flower colors and variegated flowers are caused by transposon-induced mutations of genes associated with anthocyanin biosynthesis in petunia and Japanese morning glory (Inagaki et al., 1994; Matsubara et al., 2005; Nakajima et al., 2005). Furthermore, as a result of transposon-related mutations and their accumulation, outstanding morphological variations in flower, leaf, and stem, such as various degrees of double flower and of thin and twisted petals and leaves, are produced in Japanese morning glory (Nitasaka, 2003, 2007). In torenia, the mutant “Flecked” that has high transposition activity of transposon Ttf1 has been reported (Nishijima et al., 2013). Ttf1 belongs to the En/Spm family, and is a non-autonomous element without a DNA sequence encoding transposase (Nishijima et al., 2013). Novel mutants caused by transposition of Ttf1 have appeared in selfed progeny of “Flecked” (Nishijima and Niki, 2013a, b; Nishijima et al., 2015). Among these mutants, we found one in which pigmentation patterns of the ventral petal were converted into those of dorsal petals, resulting in dorsoventrally symmetric flowers resembling begonia flowers. Thus, we named the mutant “Begonia”. Mutations showing conversion of ventral into dorsal petal without any change in lateral petals have not been found in other plants, including model plants bearing zygomorphic flowers such as A. majus, Lotus japonicas, and pea (Feng et al., 2006; Wang et al., 2008). Thus, “Begonia” could shed light on regulatory networks involved in flower symmetry. “Begonia” will also serve as a material for torenia breeding, which could lead to new cultivars with dorsoventrally symmetric flowers.

In this study, we investigated the morphological and developmental characters of flowers in “Begonia” and analyzed their mode of inheritance. We also analyzed the expression profiles of genes involved in floral symmetry. Based on the results, we discuss the molecular mechanisms of petal identity conversion in the mutant.

Materials and Methods

Plant materials

S1 generations of the torenia (Torenia fournieri Lind. ex Fourn.) mutant “Flecked” were used for novel mutant selection. Seeds were germinated in 288-cell plug trays filled with horticultural soil (Prime-Mix; Sakata Seed Co., Kanagawa, Japan), and seedlings were grown in a glasshouse kept at 18°C–32°C. The seedlings were transplanted into 72-cell plug trays filled with the same horticultural soil, and grown until flowering. A mutant plant with the same morphological and color features in abaxial as in adaxial petals was selected from approximately 500 S1 plants. This mutant was planted in a mixture of Kureha-Engei-Baido (Kureha Chemical Industry Co., Tokyo, Japan) and Prime-Mix [1:1 (v/v)] in plastic pots (9 cm in diameter and 10 cm in depth), and grown in the greenhouse as described above. The plants were selfed, and S3–S4 generations were used for subsequent experiments.

Morphological analysis of flower bud development

For observation of morphological changes in developing flower buds, flower buds at the early corolla development stage (Stage 5) and the middle corolla development stage (Stage 6; Nishijima and Shima, 2006) were used. Flower buds were fixed in formalin-acetic acid-alcohol [FAA; 50% (v/v) ethanol, 10% (v/v) formaldehyde, 5% (v/v) acetic acid] at 4°C for more than 4 h, and then dehydrated through a graded 2-methyl-2-propanol (2M2P)/ethanol series [0%:30%, 0%:50%, 10%:50%, 20%:50%, 35%:50%, 50%:40%, and 75%:25% (v/v)]. Ethanol was then completely replaced with 2M2P. The samples were frozen in 2M2P, and then freeze-dried. The freeze-dried samples were analyzed using a scanning electron microscope (SEM; VE-7800; Keyence, Osaka, Japan).

Genetic analysis

Normal-type (NT) flower of “Flecked”, the parent line of “Begonia”, mutant-type (MT), and apparent somatic-revertant (SR) flowers of “Begonia” were self- or cross-pollinated by hand in the combinations shown in Table 1. Six flowers from each of the seed and pollen parent plants were used for crossing.

Table 1

Trait segregation of descendants derived from crosses among the normal-type (NT) flowersz, mutant-type (MT) flowers, and the flowers apparently with complete somatic-reversion (SR) of “Begonia”.

cDNA cloning of floral symmetry genes and phylogenetic analysis

For cDNA cloning, total RNA was isolated from young flower buds of “Common Violet” (Takii Seed Co., Kyoto, Japan) used as NT plants. The methods for total RNA isolation and cDNA synthesis were described previously (Niki et al., 2012). Briefly, partial cDNAs were amplified using degenerate primers based on highly conserved regions of amino acid sequences for CYC, RAD, and DIV genes. Sequences of the degenerate primers were as follows: forward 5'-AARAARGAYCGICAYAGYAARAT-3' and reverse 5'-TTYTCYTTIGTYCKYTCYCTIGC-3' or 5'-TCYCTIGCYCTIGCYCTIGC-3' for CYC, forward 5'-AAYARRGCITTYGARMGIGC-3' or 5'-AAIGAYACICCIGAYCRITGG-3' and reverse 5'-GYIYYRTARTTIGGRAWIGGIAC-3' for RAD, and forward 5'-TGGACIGCIGCIGAIAAYAA-3' or 5'-GTICCIGGIAARACIGTIKGIGA-3' or 5'-AARAARTAYGGIAARGGIGAYTGG-3' and reverse 5'-TTIACIGTIGTDATRTCRTG-3' or 5'-CCARTCICCYTTICCRTAYTTYTT-3' or 5'-AAICCRTCIAAICCRTGICC-3' for DIV. Then, 5' and 3' rapid amplification of cDNA ends (RACE) was performed. Full-length cDNAs for each gene were obtained using the following primers: forward 5'-AAGGGAAAGGCAACCTTCAAATAC-3' and reverse 5'-CAAGAATTAGGATAGAGTGGTGTAAATGC-3' for T. fournieri CYC1 (TfCYC1), forward 5'-GCACAGAGACCATTAAAGCTACG-3' and reverse 5'-CGAATGCAAGTGGAATATAACACTTC-3' for TfCYC2, forward 5'-GATTGGTAATATAATAGGGTTTAGATTGAG-3' and reverse 5'-GTTTCTGTAGAGTGTTTATCACCTGACC-3' for TfCYC3, forward 5'-TTCCATTAATAATACTCACTGGCTTTC-3' and reverse 5'-CGAAATAGTAGAATCATTTACAAACTCAAC-3' for TfRAD1, forward 5'-TACATCACTCACTATTAGCTTGCTTGC-3' and reverse 5'-CATACAACTTTGGCAACATTAAGTAAGAATC-3' for TfRAD2, forward 5'-ATGGAGATTTTGTCTCCCACAG-3' and reverse 5'-TCATGCATTAGGATAGTGATGAATC-3' for TfDIV1, forward 5'-TTACAACCCTTTGTGTGTCCTGA-3' and reverse 5'-GGCAACAATTATCTACATTTGATAATGC-3' for TfDIV2, forward 5'-CTAGAAAAGCCACTTTCATATTCGTTC-3' and reverse 5'-GAGTATCTTTTAAGATTTCTTACGATGC-3' for TfDIV3, forward 5'-ACATGCAACCAACCAACACC-3' and reverse 5'-CTCAAAGAACTGCAGATAAAGAACG-3' for TfDIV4, and forward 5'-CTCTTCCATATTAGCTCTGTTTTGCTC-3' and reverse 5'-TCTTCAAGAACTGCTGCAGAAC-3' for TfDIV5. The full-length nucleotide sequences of all cDNAs were then analyzed. Accession numbers of those genes are listed in the legend for Figure 3.

For phylogenetic analyses, deduced amino acid sequences of TfCYC, TfRAD, and TfDIV genes were compared with their orthologs using Vector NTI Advance 11 software (Invitrogen, Carlsbad, CA, USA).

Quantitative real-time polymerase chain reaction analysis

Total RNA was isolated separately from sepals, dorsal, lateral and ventral petals, stamens, and pistils of flower buds in late Stage 6, and from limbs and tubes of dorsal, lateral and ventral petals of flower buds in the late corolla development stage (Stage 7). cDNA was synthesized using a PrimeScript RT Master Mix (Takara Bio, Shiga, Japan). Quantitative real-time PCR (qPCR) was performed using gene-specific primers for TfCYC, TfRAD, and TfDIV genes and the actin gene (TfACT3; AB330989), which was used as an internal control. The primer sequences and the lengths of PCR products used for the qPCR reactions were as follows: forward 5'-CATCACAGTCCAACCTTTGCTC-3' and reverse 5'-CACTAATTGAACACCACTGAATCG-3' for TfCYC1 (136 bp), forward 5'-GAGGGAGGACACAGATTATTTCC-3' and reverse 5'-AGGACAGCATAAGTTATAACTGCG-3' for TfCYC2 (117 bp), forward 5'-ATCAGGAAGGGGCAAAGG-3' and reverse 5'-ACAACACAAACTTACTGCAGCA-3' for TfCYC3 (119 bp), forward 5'-GGTGGCAACAAGAGAGACGAG-3' and reverse 5'-CAAGACACACAAAGCATCTTCTTCAC-3' for TfRAD1 (116 bp), forward 5'-GAGGTGATTGTAACTCCAACAGGAAG-3' and reverse 5'-CTTGATAGCCTGATGGATGATTCAC-3' for TfRAD2 (102 bp), forward 5'-TGCATCTGCTGCCTCTTACG-3' and reverse 5'-GCATTAGGATAGTGATGAATCCCTG-3' for TfDIV1 (111 bp), forward 5'-ACTGTAGATCTCAACGATACCAAGGTTGAG-3' and reverse 5'-CTGATATTGATAGACGCCATTAATCAATCAC-3' for TfDIV2 (255 bp), forward 5'-GGATTATCGCAGTGTGAAAGATGG-3' and reverse 5'-AAGCATGAGATGCTTATGCTGTGT-3' for TfDIV3 (162 bp), forward 5'-CTGTGTTAAGGTGATTGATTAAGGAGC-3' and reverse 5'-GTCGGACATTAAGTATTATGAAACTCCTCT-3' for TfDIV4 (124 bp), forward 5'-TGCAGATTAGCATGCGTGTAGAT-3' and reverse 5'-GAAATTCCTCAGAGCGCAGC-3' for TfDIV5 (80 bp), and forward 5'-TGCAGTAAAGTGTATTGTGGAAG-3' and reverse 5'-GGAACTATCTGGGTAGGATC-3' for TfACT3 (145 bp). Expression of the genes was quantified using SYBR Premix Ex Taq II and qPCR machine (Thermal Cycler DICE Real Time System; Takara). PCR reactions were performed with an initial denaturation step of 30 s at 95°C, followed by 50 cycles of a two-step reaction for 5 s at 95°C and 30 s at 60°C. Fluorescence was measured at the end of each extension phase. The raw data were analyzed with Thermal Cycler DICE Real Time System Software version 5.11 (Takara). The plasmids harboring full-length cDNA of each gene or a partial cDNA fragment of TfACT3 were used to obtain the standard curves. The ratio of the expression of each gene to that of TfACT3 was calculated. Expression analyses were performed independently in triplicate.

Results and Discussion

Morphological characterization of “Begonia”

Normal-type (NT) flowers of torenia had five petals: two dorsal, two lateral, and one ventral (Fig. 1). The two dorsal petals were fused and pigmented in pale violet, whereas limbs of lateral and ventral petals were violet owing to anthocyanin accumulation (Nishijima et al., 2013). The ventral petal had a yellow nectar guide. These characters clearly show that the normal torenia flower is zygomorphic.

Fig. 1

Flowers of normal-type (NT) “Common Violet” and mutant-type (MT) “Begonia”.

The ventral petal of mutant-type (MT) flowers changed to pale violet and the nectar guide disappeared, although dorsal and lateral petals showed no marked changes (Fig. 1). The ventral petal was folded inside along the proximal–distal axis like the dorsal petals. These observations indicate that the ventral petal of “Begonia” was converted into dorsal petal, resulting in radially symmetric, or more specifically, dorsoventrally symmetric flowers.

The developmental processes of NT and MT flowers were compared. No clear morphological difference was observed between NT and MT flower buds until early Stage 6 (Fig. 2). At late Stage 6, the ventral petal was placed inside the lateral petals, whereas dorsal petals were placed outside the lateral petals in NT flowers. In contrast, in “Begonia”, the ventral petal was placed outside the lateral petals, like the dorsal ones. The two dorsal petals were fused and folded inside along the proximal–distal axis at early Stage 5 in both NT and MT flowers, and this folding also appeared in the ventral petal of the MT flowers at Stage 7. In the NT flowers, limbs of lateral and ventral petals were pigmented in violet, owing to anthocyanin accumulation, and a nectar guide appeared on ventral petal at Stage 7; however, in “Begonia”, the ventral petal was not pigmented in violet, and the nectar guide did not appear. These results also support the idea that ventral petal is converted into dorsal petal in “Begonia”. However, no evidence that the dorsalized ventral petal of the MT flowers consisted of two fused petals as dorsal petals was obtained.

Fig. 2

Development of normal-type (NT) and mutant-type (MT) flower buds. Floral stages are defined as described in Nishijima and Shima (2006). Scale bars, 100 μm. D, dorsal petal; L, lateral petal; V, ventral petal.

Stability and genetics of “Begonia” trait

Trait reversion, in the form of flowers with nectar guide on the violet-pigmented ventral petal, was sometimes observed in “Begonia”. Of 40 S4 plants, eight (20%) bore reverted flowers 1 month after the first flower opened. These plants had one to three (8%–23%, 14% on average) reverted flowers among nine to 17 (12.7 on average) opened flowers. Reverted flowers accounted for 2.9% of total flowers among the 40 plants tested. The trait reversion occurred irregularly; flowers more apical than the reverted flowers in the same inflorescence were often MT.

All S1 plants, not only those obtained from MT flowers but also those obtained from reverted flowers, were MT (Table 1). These results, together with the irregular trait reversion described above, suggest that the “Begonia” mutation is genetically stable, and that trait reversion may be attributed to inconsistent phenotypic expression.

All F1 plants obtained from crosses between NT and MT plants were NT irrespective of the cross direction. NT and MT plants segregated in an approximately 3:1 ratio in the F2 generation (Table 1), suggesting that the “Begonia” mutation is governed by a single recessive gene. It should be noted that a plant bearing only NT flowers was defined as an NT plant, whereas a plant bearing only MT flowers or both MT and reverted flowers was defined as an MT plant.

Structure and expression analyses of floral symmetry associated genes

We isolated full-length cDNAs of three CYC (TfCYC1, 2, and 3), two RAD (TfRAD1 and 2), and five DIV (TfDIV1, 2, 3, 4, and 5). Phylogenetic analysis of the deduced amino acid sequences of these genes and those of the orthologs in other plants showed that the torenia genes were highly similar to the orthologs of A. majus (Fig. 3a, b; Barg et al., 2005; Baxter et al., 2007; Machemer et al., 2011; Martín-Trillo and Cubas, 2010). The TCP domain for DNA binding activity, R domain involved in the CYC/TB1 family of TCP genes, and ECE motif were conserved in TfCYC1, 2, and 3 (Fig. 4a; Howarth and Donoghue, 2005). Both the MYBI and II domains, which are suggested to function in protein–protein interaction and DNA binding activity, respectively, were conserved in TfDIV1, 2, 3, 4, and 5 (Fig. 4b). In TfRAD1 and 2, structures characteristic of RAD were also conserved; thus, only the MYBI domain and the CREB motif for DNA binding activity at the C terminus were detected (Fig. 4b; Barg et al., 2005).

Fig. 3

Phylogenetic analysis of floral symmetry genes in torenia and other plant species. Phylogenetic tree of TCP proteins belonging to the CYC/TB1 family (a) and DIV and RAD proteins (b). Accession numbers are as follows: A. majus proteins, CYC, Y16313; DICH, AF199465; DIV, AY077453; RAD, AY954971; Arabidopsis thaliana proteins, AtDIV1, At5g58990; AtDIV2, At5g04760; AtDIV3, At2g38090; AtDIV4, At5g01200; AtDIV5, At3g11280; AtDIV6, At5g05790; AtRL1, At4g39250; AtRL2, At2g21650; AtTCP1, At1g67260; AtTCP12, At1g68800; AtTCP18, At3g18550; maize protein, ZmTB1, AAK60255; rice protein, OsTB1, AB088343; tomato proteins, SlFSM1, AJ583670; SlMYBI, AJ243339; torenia proteins, TfCYC1, LC102728; TfCYC2, LC102287; TfCYC3, LC102288; TfDIV1, LC090631; TfDIV2, LC090632; TfDIV3, LC102289; TfDIV4, LC102290; TfDIV5, LC102291; TfRAD1, LC090634; TfRAD2, LC090633. The phylogenetic trees were constructed using the neighbor-joining method (Saitou and Nei, 1987). Values in parentheses indicate P-distance.

Fig. 4

Comparison of amino acid alignment of floral symmetry genes in torenia and Antirrhinum majus. Underlining indicates TCP and R domains of CYC (a) and MYBI and II domains of RAD and DIV proteins (b). Conserved amino acids in each domain are shown in the lower column. Asterisks indicate conserved amino acids in TCP domain of TCP/TB1 family proteins (Martín-Trillo and Cubas, 2010). Residues of the ECE motif (a) and CREB motif (b) are boxed.

In NT flowers, TfCYC1, 2, and 3 were highly expressed in dorsal petals, whereas the expression was much lower in lateral petals and lowest in the ventral petal at Stage 6 (Fig. 5). The expression levels of genes in dorsal and lateral petals of “Begonia” were approximately the same as in NT flowers; however, those genes were expressed in the ventral petal of the mutant at approximately the same level as those in dorsal ones (Fig. 5). No difference between NT and MT flowers in expression profiles was observed in the sepals, stamens, or pistils (Fig. 5). At Stage 7, the expression of TfCYC genes was analyzed in the limb and tube separately, and the expression of TfCYC1 and 2 in dorsal petals of MT flowers was lower in both the limb and tube than those in NT flowers. Expression profiles of TfCYC 1, 2, and 3 along the dorsoventral axis were largely the same at Stage 7 as in Stage 6. In brief, higher expression of these genes was observed in dorsal petals, and the expression was reduced in lateral petals in both NT and MT flowers. It was lowest in the ventral petal of NT flowers, but high in the ventral petal of MT flowers, with approximately the same level in dorsal petals (Fig. 5). It should be noted that the expression of TfCYC1 and 3 in tubes was higher than those in limbs (Fig. 5).

Fig. 5

Expression of TfCYC genes in floral organs of torenia. Flower buds of normal-type (NT) and mutant-type (MT) were collected at Stages 6 and 7. Floral stages were defined as described in Figure 2. The expression levels are shown as a value relative to that of TfACT3, which was used as an internal standard. D, dorsal petal; L, lateral petal; Pi, pistil; Se, sepal; Sm, stamen; V, ventral petal. Vertical bars indicate SE (n = 3).

Similarly to TfCYC1, 2, and 3, TfRAD1 was expressed highly in dorsal petals, and reduced greatly in lateral and ventral petals in NT flowers at Stage 6, and the expression in the ventral petal of “Begonia” was as high as in dorsal petals. Expression levels of these genes were slightly lower in all petals of MT flowers than in those of NT flowers (Fig. 6). The same expression profile of TfRAD1 in NT and MT flowers was observed also at Stage 7 (Fig. 6). In contrast, TfRAD2 showed stamen-specific expression in both NT and MT flowers at Stage 6. At Stage 7, TfRAD2 was mainly expressed in tubes of all petals of both NT and MT flowers, although the expression level was much lower than that in the stamens at Stage 6 (Fig. 6). Thus, TfRAD2 exhibited no expression profile associated with dorsoventral position.

Fig. 6

Expression of TfRAD genes in floral organs of torenia. Flower buds used were the same as in Figure 5. The expression levels are shown as a value relative to that of TfACT3. D, dorsal petal; L, lateral petal; MT, mutant-type; NT, normal-type; Pi, pistil; Se, sepal; Sm, stamen; V, ventral petal. Vertical bars indicate SE (n = 3).

Expression of TfDIV2, 3, 4, and 5 was marked in lateral and ventral petals, whereas expression levels in dorsal petals were lower in NT flowers at Stage 6 (Fig. 7). In “Begonia”, expression of these genes in the ventral petal was reduced to a level comparable to that in dorsal petals, while expression of TfDIV2, 3, and 4 in all petals was lower than in NT flowers. At Stage 7, the expression profiles of TfDIV2, 3, 4, and 5 in limbs was largely the same as at Stage 6; that is, TfDIV2, 3, 4, and 5 were expressed less in dorsal than in lateral and ventral petals in NT flowers, and the expression in the ventral petal was reduced to the same level as that in dorsal petals in MT flowers (Fig. 7). In tubes, there was no difference of TfDIV3, 4, and 5 expression in both NT and MT flowers, as well as TfDIV2 expression in MT flowers, among dorsal, lateral, and ventral petals, though TfDIV2 was expressed less in dorsal than in lateral and ventral petal tubes in NT flowers. There was no difference in TfDIV1 expression between NT and MT petals in either Stages 6 or 7 (Fig. 7).

Fig. 7

Expression of TfDIV genes in floral organs of torenia. Flower buds used were the same as in Figure 5. The expression levels are shown as a value relative to that of TfACT3. D, dorsal petal; L, lateral petal; MT, mutant-type; NT, normal-type; Pi, pistil; Se, sepal; Sm, stamen; V, ventral petal. Vertical bars indicate SE (n = 3).

These results clearly suggest that high expression of TfCYC1, 2, and 3 and TfRAD1 is essential to dorsal petal identity (Figs. 1, 2, 5, and 6). Thus, high expression of these genes in the ventral petal may have converted ventral to dorsal petal in “Begonia” (Fig. 8a). In our results, no difference between NT and MT flowers in morphology or pigmentation pattern in lateral petals was observed (Figs. 1 and 2), probably because the expression of TfCYC1, 2, and 3 and TfRAD1 was not changed in lateral petals of NT and MT flowers (Figs. 5 and 6), causing no change in identity of lateral petals. However, in “Begonia”, expression of TfDIV2, 3, 4, and 5 in the ventral petal was reduced to the same level in dorsal petals, whereas the genes were expressed more highly in ventral than in dorsal petals in NT flowers (Fig. 7). These results suggest that high expression of TfDIV2, 3, 4, and 5 influences ventral petal identity (Fig. 8a).

Fig. 8

Hypothetical relationship between expression level of floral symmetry genes and petal identity (a) and regulatory network of those genes (b) in normal-type (NT) and mutant-type (MT) torenia flower buds. For dorsal petal identity, TfDIV function may be inhibited based on protein–protein interaction with TfRAD1 induced by TfCYC. D, dorsal petal; L, lateral petal; V, ventral petal.

RAD is induced by CYC/DICH (Corley et al., 2005; Costa et al., 2005; Luo et al., 1999), and DIV function is counteracted by RAD in A. majus (Cui et al., 2010; Galego and Almeida, 2002). In our results, TfCYC1, 2, and 3 and TfRAD1 showed the same expression profiles (Figs. 5 and 6), suggesting that TfCYC induces TfRAD1 also in torenia (Fig. 8b). In addition, expression of TfDIV2, 3, 4, and 5 may be downregulated by TfRAD1 in torenia (Fig. 8b). Among the five TfDIV proteins, the amino acid sequence of TfDIV1 showed the highest similarity to DIV in A. majus (Fig. 3b); however, expression of TfDIV1 was not different among petals of either NT or MT flowers (Fig. 7). Although DIV is an important player in ventral petal identity (Galego and Almeida, 2002; Rose et al., 1999), it is expressed in both ventral and dorsal petals (Almeida et al., 1997; Galego and Almeida, 2002). Recently, it has been shown that RAD negatively regulates DIV function via protein-protein interaction (Raimundo et al., 2013). DIV must form a heterodimer with another MYB protein, DIV-and-RAD-interacting-factor (DRIF), to function as a transcription factor in establishment of ventral petal identity. RAD binds to DRIF in competition with the DIV-DRIF interaction, promoting translocation of DRIF to cytoplasm and resulting in downregulation of DIV function. The same interaction mechanism for MYB proteins has been reported for RAD and DIV orthologs in tomato during fruit development (Machemer et al., 2011). In our results, differences between dorsal and ventral petals in TfDIV expression were less clear than those for TfCYC1, 2, and 3 and TfRAD1 in NT flowers (Figs. 5, 6, and 7). Thus, the above regulatory mechanisms based on protein-protein interaction between TfRAD1 and TfDIV may function also in torenia (Fig. 8b).

Future prospects

The roles of the floral symmetry genes CYC, RAD, and DIV and their regulatory network involved in dorsoventral flower asymmetry in A. majus have been identified; however, the regulator causing high CYC expression in dorsal petals remains unknown. This regulator may be mutated in “Begonia”, given that the expression profiles of TfCYC1, 2, and 3 were changed simultaneously in the mutant, ruling out mutations in individual TfCYC. Given that “Begonia” was isolated from selfed progeny of “Flecked” in which the transposable element Ttf1 was activated, the altered phenotype observed in “Begonia” may be caused by Ttf1. We have not yet identified the responsible gene by a transposon display method based on Ttf1 sequence (data not shown), suggesting that the phenotype of “Begonia” is due to the footprint left by Ttf1 excision from the responsible gene.

As described in the Introduction, torenia has little variation in flower morphology; no commercial cultivars bear double- and large-flowers or altered flower symmetries. Torenia has been cultivated since the end of the 19th century in Japan (Morise, 2001); however, its market is small, probably because of the low variation in flower appearance. Development of new flower morphology would attract consumers and increase the plant’s commercial value. In “Begonia”, conversion of the ventral into dorsal petal changed the zygomorphic flower into a dorsoventrally symmetric one, bringing the appearance of the flower closer to that of a begonia flower. Thus, “Begonia” could serve as a useful breeding material to develop cultivars bearing flowers with new symmetrical structure.

Acknowledgements

We thank Mrs. Toshie Iida, Mrs. Setsuko Kimura, and Mrs. Tomoko Kurobe for their technical assistance.

Literature Cited
 
© 2016 The Japanese Society for Horticultural Science (JSHS), All rights reserved.
feedback
Top