Application of Mutant ap3 Allele-based Markers for the Selection of the Long-lasting Flower Phenotype (Misome-shō) in Evergreen Azalea Cultivars

Abstract

There is a long-lasting flower trait with a temporal color change, known as “misome-shō”, in Japanese evergreen azalea. This trait has been found in several wild Japanese evergreen azalea species, such as Rhododendron kaempferi ‘Nikkō-misome’, R. macrosepalum ‘Kochō-zoroi’, R. indicum ‘Chōjyu-hō’, and R. × hannoense ‘Amagi-beni-chōjyu’. The corollas of long-lasting flower cultivars undergo a conversion of normal corollas to sepaloid corollas due to loss of function of the MADS-box B class gene, APETALA3 (AP3) homolog. Also, the long-lasting flower trait was shown to be recessive to normal flowers and controlled by a single gene. To develop a DNA marker for selection of the long-lasting flower phenotype, we carried out a multiplex-PCR approach to detect the ap3 mutant allele related to long-lasting flower traits, and investigated the flower phenotypes in crossed progenies of 23 cross combinations and 245 individuals. The normal flower phenotype individuals in the crossed progenies were homozygous for the normal allele or heterozygous for the normal allele and the ap3 mutant allele. On the other hand, the long-lasting flower phenotype individuals in the crossed progenies were homozygous for the ap3 mutant allele of long-lasting cultivars. These results support the idea that the long-lasting flower trait is caused by a mutation in the AP3 homolog, and it has been newly clarified that any combination of these mutant alleles in long-lasting flower cultivars has the long-lasting flower phenotype. In conclusion, our data indicate that efficient selection of individuals with long-lasting flowers will be possible by using selection DNA markers linked to the long-lasting flower trait.

Introduction

Among the 50 Rhododendron species native to Japan (Kurashige et al., 1998), species within the subgenus Tsutsusi, section Tsutsusi, such as R. kaempferi, R. indicum, R. kiusianum, R. ripense, and R. macrosepalum, are important genetic resources for the evergreen azaleas used as ornamental shrubs or pot azaleas (Kobayashi, 2020). Each cultivar group, such as Edo Kirishima, Kurume, Ryūkyū, Hirado, and Satsuki, has been developed since the Edo era (1603–1867) by selection and crossing based on these wild species as they have high ornamental value, are easy to cultivate, and exhibit interspecific hybridization (Kobayashi et al., 2000; Kurashige and Kobayashi, 2008; Kobayashi, 2016). During the azalea boom in the middle of the 17^th century, cultivars with flower color variations and shape mutants, such as white flowers, double flowers, and hose-in-hose flowers, were selected and collected from wild species all over the country. Many of these cultivars were described in the first monograph on azaleas, “Kinshū - makura”, edited in 1692 (Ito and Creech, 1984; Kobayashi, 2016). This monograph provided the first description of ‘Misome-guruma’, a cultivar with small light-red flowers, long stamens, and a long blooming season, with the flowers turning greenish in June (Ito and Creech, 1984). The long-lasting flower trait, known as “misome-shō”, has been found in several wild species, such as R. kaempferi ‘Nikkō-misome’, R. macrosepalum ‘Kochō-zoroi’, R. indicum ‘Chōjyu-hō’, and R. × hannoense ‘Amagi-beni-chōjyu’.

In order to select breeding material for long-lasting flower cultivars, we have conducted various studies on floral morphology, inheritance, and a MADS-box B class gene related to floral morphology. In a previous morphological study, we indicated that the long-lasting flower trait is a floral homeotic mutant with a sepaloid corolla (Kobayashi et al., 2010; Gobara et al., 2017). In addition, the long-lasting flower trait was shown to be a recessive trait of the normal flower phenotype and is controlled by a single gene (Gobara et al., 2017). Therefore, it takes a long-term breeding cycle to develop long-lasting flowers with new flower traits, such as double flowers and hose-in-hose flowers.

A marker-associated selection (MAS) (Xu and Crouch, 2008) system is important for rapid selection of the long-lasting flower phenotype, which is a recessive trait, during the juvenile period. In long-lasting cultivars, transcription of the AP3 homolog was reduced, and a long terminal repeat retrotransposon was independently inserted into exons 1, 2, and 7 or an unknown sequence in exon 1 of the gDNA of each cultivar. This insertion apparently abolished the normal mRNA sequence of the AP3 homolog (Cheon et al., 2018). Based on this insertion, DNA markers for the detection of the ap3 mutant allele were developed for each long-lasting cultivar. In a previous study (Cheon et al., 2018), normal flower phenotypes of the F₁ progenies between normal and long-lasting cultivars were heterozygous for the normal and ap3 mutant alleles. In addition, it was shown that long-lasting flowers in crossed progenies between normal and long-lasting cultivars were homozygous for the ap3 mutant allele in one cross combination.

In this study, to develop a rapid selection marker for the long-lasting flower phenotype in evergreen azaleas, we carried out a multiplex-PCR approach to detect the ap3 mutant allele related to the long-lasting flower trait in crossed progenies of various cross combinations using the DNA markers described by Cheon et al. (2018).

Materials and Methods

Plant materials

The normal flower azaleas R. kaempferi, R. macrosepalum, R. indicum ‘Ōsakazuki’, R. ripense, R. × pulchrum ‘Ōmurasaki’, and Kurume azalea ‘Wakakaede’, and the long-lasting flower cultivars R. kaempferi ‘Nikkō-misome’, R. macrosepalum ‘Kochō-zoroi’, R. indicum ‘Chōjyu-hō’, and R. × hannoense ‘Amagi-beni-chōjyu’ were used. A total of 69 F₁ progenies from 11 cross combinations between normal and long-lasting cultivars and a total of 25 F₁ progenies of six cross combinations between long-lasting cultivars were used (Table 1). In addition, a total of 151 crossed progenies of six cross combinations between the F₁ hybrids of normal and long-lasting cultivars were used (Table 2). These plants were obtained from the azalea resource collection at the Plant Breeding Laboratory of the Faculty of Life and Environmental Sciences, Shimane University.

Table 1

PCR amplification profiles for the ap3 mutant alleles and the flower phenotypes in F₁ progenies.

Table 2

PCR amplification profiles for the ap3 mutant alleles and the flower phenotypes in the crossed progenies between F₁ hybrids of normal and long-lasting cultivars.

DNA marker analysis

Genomic DNA was extracted from the leaves of each plant using the modified CTAB method (Kobayashi et al., 1998). The multiplex-PCR primers described by Cheon et al. (2018) that can distinguish each ap3 mutant allele in long-lasting cultivars are shown in Table 3 and Figure 1A. The Pr2 and Pr8 primers, which were newly designed from the AP3 homolog sequences of R. kaempferi ‘Nikkō-misome’ (rkap3-NM, DDBJ Acc. No. AB861605) and R. × hannoense ‘Amagi-beni-chōjyu’ (rhap3-AC, DDBJ Acc. No. AB861608), were used in this study. The three PCR primers Pr8, Pr9, and Pr10 were designed to amplify approximately 425 bp fragments in ‘Nikkō-misome’, and approximately 180 bp fragments in cultivars other than ‘Nikkō-misome’. The three PCR primers Pr1, Pr3, and Pr4 were designed to amplify approximately 315 bp fragments in ‘Kochō-zoroi’ (rmap3-KZ, DDBJ Acc. No. AB861606), and approximately 180 bp fragments in cultivars other than ‘Kochō-zoroi’. The three PCR primers Pr5, Pr6, and Pr7 were designed to amplify approximately 600 bp fragments in ‘Chōjyu-hō’ (riap3-CH, DDBJ Acc. No. AB861607), and approximately 220 bp fragments in cultivars other than ‘Chōjyu-hō’. The three PCR primers Pr1, Pr2, and Pr3 were designed to amplify approximately 500 bp fragments in ‘Amagi-beni-chōjyu’, and approximately 180 bp fragment in cultivars other than ‘Amagi-beni-chōjyu’. A multiplex-PCR was performed with the primer set in Table 3 to detect the ap3 mutant alleles of each long-lasting cultivar and the normal alleles of normal flowers as the insets in rmap3-KZ and rhap3-AC were located in exon 1. The normal alleles, rmap3-KZ, and rhap3-AC were detected using four primers, Pr1, Pr2, Pr3, and Pr4. The multiplex-PCR was performed in 10 μL reaction mixtures containing 10 ng DNA template, 1 × Ex-Taq buffer, 200 μM dNTPs, 0.25 U Ex-Taq (TaKaRa Bio Inc., Shiga, Japan), and 0.2 μM of each primer. The reaction conditions were as follows: denaturation at 94°C for 3 min; 35 cycles of denaturation at 94°C for 30 s, annealing at 61°C for 30 s, extension at 72°C for 30 s, and a final extension at 72°C for 2 min. The amplified PCR products were analyzed by electrophoresis in a 1.5% agarose gel (Nippon Genetics Co., Ltd., Tokyo, Japan) using 0.5× TBE buffer and run at 100 V for 30 min. When using the Pr2 and Pr9 primers, the annealing temperature was changed to 56°C. The crossed progenies between the F₁ hybrids of normal and long-lasting cultivars had their genotypes determined using the DNA markers before blooming. We investigated the flower phenotypes of the crossed progenies that bloomed from 2013 to 2020 and identified genotypes of the AP3 homolog through the multiplex-PCR.

Table 3

Primers used for the multiplex-PCR analysis.

Fig. 1

Multiplex-PCR to detect the normal AP3 allele linked to the normal flower phenotype and the ap3 mutant allele linked to the long-lasting flower phenotype..A: Schematic diagram showing the primer regions that distinguish normal and ap3 mutant alleles (Cheon et al., 2018). E1–7 and lines indicate each exon and each intron, respectively. Horizontal arrows indicate regions and direction of primers. B: Results of multiplex-PCR amplification of the normal allele and rkap3-NM, riap3-CH, rmap3-KZ, and rhap3-AC. Numbers indicate (1) Rhododendron kaempferi, (2) R. macrosepalum, (3) R. indicum ‘Ōsakazuki’, (4) R. ripense, (5) R. × pulchrum ‘Ōmurasaki’ (6) Kurume azalea ‘Wakakaede’, (7) R. kaempferi ‘Nikkō-misome’, (8) R. macrosepalum ‘Kochō-zoroi’, (9) R. indicum ‘Chōjyu-hō’, and (10) R. × hannoense ‘Amagi-beni-chōjyu’. C: Results of multiplex-PCR amplifications in 99088 line (P1: R. × hannoense ‘Amagi-beni-chōjyu’ × P2: R. × pulchrum ‘Ōmurasaki’). D: Results of the multiplex-PCR amplifications in 06032 line (P3: R. kaempferi ‘Nikkō-misome’ × P4: R. indicum ‘Chōjyu-hō’). N and L indicate the normal flower phenotype and the long-lasting flower phenotype, respectively. M indicates a 100 bp ladder (Nippon Genetics).

Results

Detection of the ap3 mutant allele in long-lasting cultivars

The PCR products obtained using the three primers (Pr8, Pr9, and Pr10) amplified approximately 425 bp fragments in ‘Nikkō-misome’, and approximately 180 bp fragments in the other cultivars (Fig. 1B). The PCR products obtained using the primers Pr5, Pr6, and Pr7 amplified approximately 600 bp fragments in ‘Chōjyu-hō’, and approximately 220 bp fragments in cultivars other than ‘Chōjyu-hō’ (Fig. 1B). The PCR products obtained using the four primers Pr1, Pr2, Pr3, and Pr4 amplified approximately 315 bp fragments in ‘Kochō-zoroi’, 500 bp fragments in ‘Amagi-beni-chōjyu’, and 180 bp fragments in cultivars other than these (Fig. 1B).

Detection of the ap3 mutant allele in crossed progenies

All analyzed F₁ progenies between the normal and long-lasting cultivars had normal flower phenotypes and were heterozygous for the ap3 mutant allele and the normal allele (Table 1; Fig. 1C). On the other hand, all analyzed F₁ progenies between the long-lasting cultivars had long-lasting flower phenotypes and were homozygous for the ap3 mutant alleles (Table 1; Fig. 1D).

In the crossed progenies between the F₁ hybrids of normal and long-lasting cultivars, normal and long-lasting flower phenotypes exhibited an approximately 3:1 split (Table 2; Fig. 3A). In all combinations of the crossed progenies between the F₁ hybrids of normal and long-lasting cultivars, normal alleles were detected in normal flower phenotypes, and homozygous ap3 mutant alleles were detected in the long-lasting flower phenotypes (Table 2; Fig. 2). The genotype segregation pattern in the crossed progenies between the F₁ hybrids of the normal and long-lasting cultivars, identified by the multiplex-PCR, corresponded to the expected ratio (Table 4).

Fig. 2

Results of the multiplex-PCR amplifications in the 13006 line (04093 × 07097). N and L indicate the normal flower phenotype and the long-lasting flower phenotype, respectively. PC indicates the positive control (Pr8 + Pr9 + Pr10: ‘Nikkō-misome’; Pr5 + Pr6 + Pr7: ‘Chōjyu-hō’). M indicates a 100 bp ladder (Nippon Genetics).

Fig. 3

The relationships between flower phenotypes and genotypes in crossed progenies of long-lasting cultivars derived from the four different species. A: The inheritance model of the long-lasting flower trait. Parentheses show the segregation ratio. B: The inheritance model of the flower phenotypes based on the AP3 genotype combinations.

Table 4

Segregation data for the ap3 mutant alleles in the crossed progenies between F₁ hybrids of normal and long-lasting cultivars.

Discussion

The vase life of flowers is one of the most important characteristics required by consumers (Imanishi et al., 1992). Normal flowering azalea plants finish flowering in about two weeks. In contrast, long-lasting flower cultivars keep their corollas for more than 100 days, with the flowers then turning greenish (Kobayashi, 2013). This is similar to the “autumn-like color” of hydrangeas, in which flowering continues with changes in the flower color (Yoshida et al., 2008).

The corollas of long-lasting flower cultivars involves the conversion of normal corollas to sepaloid corollas due to loss of function of the AP3 homolog (Cheon et al., 2018). In addition, the long-lasting flower trait was shown to be recessive to normal flowers and controlled by a single gene (Gobara et al., 2017). Similar mutations of sepaloid petals have been found in other plants, and Singh et al. (2014) suggested that the development of sepaloid petals in Papaver somniferum was caused by the reduced expression of PapsAP3-1; this trait is a recessive trait controlled by a single gene. In herbaceous peony (Paeonia lactiflora), a spontaneous corolla mutant under natural conditions was closely associated with selective expression alterations of duplicated AP3 and PI genes; down-regulation of duplicated B class genes may result in a reduction in the AP3-PI heterodimer in sepaloid corollas (Gong et al., 2017). Additionally, Lange et al. (2013) suggested that sepaloid petal development in Eschscholzia californica was caused by a lack of function of the MADS-box B class gene, EScaGLO. DNA markers related to flower shape mutation have been developed for double-flowered cultivars caused by mutations of the MADS-box C class gene AGAMOUS in gentians (Gentiana scabra) (Tasaki et al., 2017) and Matthiola incana (Nakatsuka and Koishi, 2018). However, there are no reports of any DNA markers related to an improvement in the ornamental value and vase life of flowers caused by B class gene mutations such as the long-lasting flower trait in azalea.

The long-lasting flower trait in azalea is commercially important, and we have been conducting studies on breeding materials (Kobayashi, 2020). In a previous study, DNA markers to detect the ap3 mutant allele were developed in long-lasting cultivars (Cheon et al., 2018). In this study, we used a multiplex-PCR approach to detect the ap3 mutant allele related to the long-lasting flower trait in normal flowers and long-lasting flower cultivars, and crossed progenies of 23 cross-combinations and 245 individuals using the DNA markers described by Cheon et al. (2018). As a result of the multiplex-PCR, we found normal flowers were homozygous for the AP3 gene, and each long-lasting cultivar had a homozygous ap3 mutant allele (Fig. 1B). In all cross combinations, individuals with the normal flower phenotype were homozygous for the normal allele or heterozygous for the normal allele and the ap3 mutant allele (Tables 1 and 2). On the other hand, the F₁ progenies between long-lasting cultivars and long-lasting flower phenotype individuals in the cross progenies between F₁ hybrids of normal and long-lasting flower cultivars were homozygous for the ap3 mutant allele derived from different species of the long-lasting cultivars (Tables 1 and 2).

The genotype that was detected before blooming by using the DNA markers was linked to the flower phenotype in all crossed progenies. The reliability of the DNA markers for long-lasting flower trait selection was confirmed in this study. Figure 3 shows the relationship between flower phenotypes and genotypes in crossed progenies of long-lasting cultivars derived from the four different species clarified in this report. It was revealed that all of the ap3 mutant alleles in each long-lasting flower cultivar were recessive to normal alleles (Fig. 3B). There was no dominant/recessive relationship between the ap3 mutant alleles of each long-lasting flower cultivar derived from the four different species, and we confirmed that any combination of these mutant alleles had a long-lasting flower phenotype (Fig. 3B). These results strengthen the concept that the long-lasting flower trait is caused by a mutation in the AP3 homolog. These results also suggest the possibility of breeding new azalea cultivars with the characteristics of multiple species by interspecific crossing using long-lasting cultivars, rather than simply through varietal crossing.

We have been conducting various studies on the morphological mutations found in evergreen azalea cultivars such as hose-in-hose, double-flowers, narrow-petals, and narrow-leaf mutations, to apply them in our breeding program (Kobayashi, 2016). In previous studies, we reported the development of DNA markers to select hose-in-hose cultivars (Cheon et al., 2017a), and the genetic analysis of the double-flowered trait (Cheon et al., 2017b) and narrow-petal and -leaf mutations (Tasaki et al., 2019). We are selecting the long-lasting flower phenotype during the juvenile period using these MAS systems. Among the selected individuals, candidates for new long-lasting flower cultivars have been developed.

In conclusion, interspecific crossing using the long-lasting flower trait that is present across species can contribute to the development of new long-lasting cultivars with high ornamental value. In addition, it was shown that an efficient selection process for long-lasting individuals is possible by using selection DNA markers linked to the long-lasting flower trait.

Acknowledgements

The authors thank the faculty of Life and Environmental Sciences of Shimane University for financial support in publishing this report.

Literature Cited

•  Cheon,  K. S.,  A.  Nakatsuka,  Y.  Gobara and  N.  Kobayashi. 2017a. Mutant RoPI-1 allele-based marker development for selection of the hose-in-hose flower phenotype in Rhododendron obtusum cultivars. Euphytica 213: 3. DOI: 10.1007/s10681-016-1808-x.
•  Cheon,  K. S.,  A.  Nakatsuka,  K.  Tasaki and  N.  Kobayashi. 2017b. Floral morphology and MADS gene expression in double-flowered Japanese evergreen azalea. Hort. J. 86: 269–276.
•  Cheon,  K. S.,  A.  Nakatsuka,  K.  Tasaki and  N.  Kobayashi. 2018. Long-lasting corolla cultivars in Japanese azaleas: a mutant AP3/DEF homolog identified in traditional azalea cultivars from more than 300 years ago. Front. Plant Sci. 8: 2239. DOI: 10.3389/fpls.2017.02239.
•  Gobara,  Y.,  A.  Nakatsuka,  K. S.  Cheon and  N.  Kobayashi. 2017. Floral morphology and inheritance of the long-lasting flower trait (misome-sho) in Japanese evergreen azalea cultivars. Hort. Res. (Japan) 16: 383–390 (In Japanese with English abstract).
•  Gong,  P.,  X.  Ao,  G.  Liu,  F.  Cheng and  C.  He. 2017. Duplication and whorl-specific down-regulation of the obligate AP3–PI heterodimer genes explain the origin of Paeonia lactiflora plants with spontaneous corolla mutation. Plant Cell Physiol. 58: 411–425.
•  Imanishi,  H.,  F.  Yonezawa and  H.  Imanishi. 1992. Psychological research on the attitude of florist customers towards flowers. J. Japan. Soc. Hort. Sci. 60: 981–987 (In Japanese with English abstract).
• Ito, I. and J. L. Creech. 1984. A Brocade Pillow: Azaleas of Old Japan. John Weatherhill, Inc., New York and Tokyo.
•  Kobayashi,  N. 2013. Evaluation and application of evergreen azalea resources of Japan. Acta Hortic. 990: 213–219.
• Kobayashi, N. 2016. Tsutsuji. p. 151–180. In: M. Shibata (ed.). Hana no hinshu kairyou no nihonshi (In Japanese). Bunkyō, Tokyo.
•  Kobayashi,  N. 2020. Japanese azalea germplasm; the living horticulture properties. Acta Hortic. 1291: 163–168.
•  Kobayashi,  N.,  T.  Handa,  Y.  Yoshimura,  Y.  Tsumura,  K.  Arisumi and  K.  Takayanagi. 2000. Evidence for introgressive hybridization based on chloroplast DNA polymorphisms and morphological variation in wild evergreen azalea populations of the Kirishima mountains, Japan. Edinb. J. Bot. 57: 209–219.
•  Kobayashi,  N.,  T.  Horikoshi,  H.  Katsuyama,  T.  Handa and  K.  Takayanagi. 1998. A simple and efficient DNA extraction method for plants, especially woody plants. Plant Tissue Cult. Biotechnol. 4: 76–79.
•  Kobayashi,  N.,  M.  Ishihara,  M.  Ohtani,  K. S.  Cheon,  D.  Mizuta,  K.  Tasaki and  A.  Nakatsuka. 2010. Evaluation and application of the long-lasting flower trait (misome-sho) of azalea cultivars. Acta Hortic. 855: 165–168.
•  Kurashige,  Y. and  N.  Kobayashi. 2008. Evergreen azalea cultivars and breeding trends in the Taisho era inferred by discovered “azalea research notes” of Kanagawa agricultural research station. Hort. Res. (Japan) 7: 323–328 (In Japanese with English abstract).
•  Kurashige,  Y.,  M.  Mine,  N.  Kobayashi,  T.  Handa,  K.  Takayanagi and  T.  Yukawa. 1998. Investigation of sectional relationships in the genus Rhododendron (Ericaceae) based on matK sequences. J. Jap. Bot. 73: 143–154.
•  Lange,  M.,  S.  Orashakova,  S.  Lange,  R.  Melzer,  G.  Theißen,  D. R.  Smyth and  A.  Becker. 2013. The seirena B class floral homeotic mutant of California Poppy (Eschscholzia californica) reveals a function of the enigmatic PI motif in the formation of specific multimeric MADS domain protein complexes. Plant Cell 25: 438–453.
•  Nakatsuka,  T. and  K.  Koishi. 2018. Molecular characterization of a double-flower mutation in Matthiola incana. Plant Sci. 268: 39–46.
•  Singh,  S. K.,  A. K.  Shukla,  O. P.  Dhawan and  A. K.  Shasany. 2014. Recessive loci Pps-1 and OM differentially regulate PISTILLATA-1 and APETALA3-1 expression for sepal and petal development in Papaver somniferum. PloS One 9: e101272. DOI: 10.1371/journal.pone.0101272.
•  Tasaki,  K.,  A.  Higuchi,  K.  Fujita,  A.  Watanabe,  N.  Sasaki,  K.  Fujiwara,  H.  Abe,  Z.  Naito,  R.  Takahashi,  T.  Hikage and  M.  Nishihara. 2017. Development of molecular markers for breeding of double flowers in Japanese gentian. Mol. Breed. 37: 33. DOI: 10.1007/s11032-017-0633-9.
•  Tasaki,  K.,  A.  Nakatsuka,  K. S.  Cheon and  N.  Kobayashi. 2019. Genetic demonstration of the involvement of WUSCHEL-related homeobox (WOX) genes in narrow-petal and narrow-leaf mutations in traditional Japanese azalea cultivars. Euphytica 215: 5. DOI: 10.1007/s10681-018-2323-z.
•  Xu,  Y. and  J. H.  Crouch. 2008. Marker-assisted selection in plant breeding: from publications to practice. Crop Sci. 48: 391–407.
•  Yoshida,  K.,  D.  Ito,  Y.  Shinkai and  T.  Kondo. 2008. Change of color and components in sepals of chameleon hydrangea during maturation and senescence. Phytochemistry 69: 3159–3165.