Introduction

Chrysanthemum (Dendranthema × grandiflorum Ramat.) belongs to genus Chrysanthemum in Asteraceae (Compositae) and is native to China, Japan, and Korea (in alphabetic order). China is known as the first country cultivated chrysanthemum and a variety of wild chrysanthemums have been used for traditional medicine. During the 17th century, chrysanthemum was introduced to European countries. Western breeders generated new cultivars for cut flowers and potted plants and widely spread out to the world becoming one of the top ornamental plants over the world. India (11,000 ha), China (7,157 ha), and Japan (4,758 ha) have the largest production area and also the top three largest producers of chrysanthemums (AIPH 2020).

New cultivars of chrysanthemums have been created by conventional breeding techniques (cross-breeding due to self-incompatibility of ornamental chrysanthemums) or mutation breeding techniques using physical or chemical treatments following large-scale selection (Shibata 2008;Zhang et al. 2018). Radiations such as X-rays, gamma-rays, heavy-ion beams, high-energy photons, or high-energy electrons are irradiated to induce mutations in radiation breeding (Miler et al. 2021;Oladosua et al. 2016;Yamaguchi 2018). In addition, ethyl methanesulphonate (EMS) or colchicine treatment is used as a chemical mutagenesis to induce mutation of chrysanthemums (Latado et al. 2004;Purente et al. 2020).

In conventional breeding, back crossing to ideotype of chrysanthemums is not possible due to out-crossing nature of its mating system. In addition, it is difficult to characterize a function of a particular gene resulting from polyploidy trait of chromosomes (2n = 6x = 54 as hexaploidy), loss or gain of chromosomes, 9.36 Gbp of large genome size, and polygenic control of important traits (Garnatje et al. 2011;Shibata and Kawata 1986). Therefore, breeding of chrysanthemum is time consuming and laborious task to create a new single cultivar which several desirable traits are introduced. Furthermore, consumers preference tends to alter very fast driving breeders to develop new cultivars in floricultural industry every year.

Flower color and shape are the main breeding target in floricultural industry and a good shelf-life of cut flowers, uniformity in production, easy to handle in production and transportation system, resistant against pests and diseases (De Jong 2001). However, self-incompatibility of chrysanthemums hinders simultaneous introduction of many commercially important and attractive traits. In recent, advances in genetic engineering and sequencing technique facilitated development of new cultivars and a deeper understanding of physiological traits of chrysanthemums about flowering control, disease resistance or resistance against abiotic stresses (Shinoyama et al. 2006). In addition to genetic engineering, molecular breeding using DNA markers accelerates development of new chrysanthemum cultivars. Polymerase chain reaction (PCR) based marker development is the widely used technique in years and there are a variety of DNA markers such as; random Amplified polymorphic DNAs (RAPD), amplified fragment length polymorphism (AFLP), simple sequence repeat (SSR), Inter simple sequence repeats (ISSRs), or sequence characterized amplified region (SCAR) markers (Chatterjee et al. 2006;Feng et al. 2016;Samarina et al. 2021;Shirao et al. 2013;Wolff and Rijn 1993;Zhang et al. 2010;Zhang et al. 2013).

Next generation sequencing (NGS) technique has enabled to analyze whole genome sequence and transcriptome of several important ornamental plants such as carnation, chrysanthemum, rose, lily, tulip, Eustoma, orchids (Yagi 2015). Genomics approach, marker-assisted selection (MAS), and genome-wide association study (GWAS) are good methods to develop more accurate markers for important breeding traits (Sumitomo et al. 2021).

In chrysanthemums, whole genome sequence and transcriptomic data of C. boreale, C. nankingense, and C. morifolium is now available and will be used as reference genome sequences (Hirakawa et al. 2019;Lu et al. 2019;Song et al. 2018). Therefore, it is expected that wild Chrysanthemums will gain a great attention in breeding science and floricultural industry and also will be widely used to understand biological process of chrysanthemums.

In this review, the recent progress in molecular biological studies and biotechnology of ornamental chrysanthemums is discussed in brief. In addition, potentiality and importance of wild chrysanthemums as bioresources for ornamental chrysanthemums by describing recent development in genomics approach. This review will be the first review which comprehensively introduces the recent research trends of chrysanthemums in terms of ornamental chrysanthemums, genetic engineering, and database of whole genome sequence and transcriptome obtained from wild chrysanthemums.

The Progress in Genetic Engineering of Chrysanthemums

In chrysanthemums, important traits tested to generate genetically engineered plants are flower color, plant architecture, flowering control, resistance to abiotic stress or biotic stresses such as; viruse, viroids, fungi, insect, etc. (Shinoyama et al. 2006;Shinoyama et al. 2020;Teng et al. 2021;Wei et al. 2017). Several methods for genetic transformation have been developed such as; particle bombardment, Agrobacterium mediated, electroporation, microinjection, in planta transformation, vacuum infiltration, etc. (Mackelprang and Lemaux 2020). Particle bombardment is one of the common genetic transformation techniques to physically introduce a target gene to plants. A gene of interest coated by gold or tungsten particles is directly delivered to plant organs such as; nucleus, chloroplasts, or mitochondria. This technique has been widely used to transform a variety of ornamental plants such as roses (Marchant et al. 1998). However, only a few studies have been reported in chrysanthemums (Teixeira Da Silva and Fukai 2002;Yepes et al. 1995;Yepes et al. 1999), but this method is one of the useful methods to characterize and localize a target gene in plant cells (Guan et al. 2021;Hosokawa et al. 2011). Agrobacterium-mediated gene transfer is the most common method widely used for genetic engineering of ornamental chrysanthemums (Noda et al. 2013;Renou et al. 1993;Shinoyama et al. 2020;Teng et al. 2021). Transformation efficiency of chrysanthemums using A. tumefaciens of chrysanthemums are largely dependent on the Agrobacterium strains such as; LBA4404, EHA101, EHA105, Ach5, and AGL0 (Shinoyama et al. 2012a). In addition, other DNA segments in T-DNA cassette such as promoters, terminators, or selection markers also affect the efficiency of genetic transformation resulting in complexity of optimizing a condition for genetic transformation. In order to overcome such conditions, a computational approach using machine learning was tried to optimize the condition of Agrobacterium mediated-genetic transformation for chrysanthemums (Hesami et al. 2020). Blue-colored chrysanthemums were generated using Agrobacterium-mediated gene transfer technique, since ornamental chrysanthemums are lack of flavonoid 3`,5`-hydroxlyase gene (F3`5`H) responsible for biosynthesis of delphinidin-based anthocyanins. Therefore, blue-colored chrysanthemums cannot be created using conventional breeding techniques. F3`5`H from several plants such as Campanula medium, Eustoma grandiflorum, Lobelia erinus, Clitoria ternatea, Viloa wittrockiana ‘Black Pansy’, Antirrhinum kelloggii, Pericallis x hybrida, Gentiana trifloral, and Verbena x hybrida were introduced to create blue-colored chrysanthemums (Brugliera et al. 2013;Noda et al. 2013).

During the last decade, advance in genome editing technique facilitated development of genetically engineered plants and deepen the understanding of particular genes in the functional genomics studies (Karkute et al. 2017;Shinoyama et al. 2020). Genome editing technique utilizes modified site-specific nucleases (SSNs) to generate mutants by deleting, inserting, or replacing a target DNA sequence. Zinc finger nucleases (ZFNs), transcription activator-like effector nucleases (TALENs), and clustered regularly interspaced short palindromic repeat-associated protein 9 (CRISPR/Cas9) are SSNs used in genome editing (Karkute et al. 2017). CRISPR/Cas9 system uses microbial CRISPRA/CAS9 in combination with single-guide RNA (sgRNA) that directs Cas9 to a complementary target DNA. Only small number of nucleotides (about twenty complementary nucleotides) obtained from target gene are enough to edit a gene of interest. Ornamental chrysanthemum has been regarded as being difficult for editing genome because chromosome is known as hexaploidy (2n = 6x = 54) and no genome information of ornamental chrysanthemum was available to design single-guide RNA (sgRNA) for genome editing. In recent, a couple of studies on genome editing of chrysanthemums were reported (Kishi-Kaboshi et al. 2017). Male and female sterile chrysanthemums were generated using TALEN technique to inhibit cross-compatibility by knocking out CmDMC1 which is involved in meiotic homologous recombination or by overexpressing Cm-ETR1/H69A, a mutated gene of ethylene receptor (Shinoyama et al. 2012b;Shinoyama et al. 2020). Six CmDMC1 genes were identified and all genes were knocked out simultaneously using TALEN- mediated genome editing technique. Genome editing of ornamental chrysanthemum using CRISPR/Cas9 was first reported by Kishi-Kaboshi et al. (2017). They created over expression line of chrysanthemum that expresses Chiridius poppei yellowish-green fluorescent protein (CpYGFP) gene and then application validity of CRISPR/Cas9 was assessed by editing CpYGFR gene in transgenic chrysanthemum. In addition, Shinoyama et al. (2015) improved insect and disease resistance by introducing a modified cry1Ab gene from Bacillus or sarcotoxin IA gene from Sarcophaga combining with or without 5’-untranslated regions of the alcohol dehydrogenase gene of Arabidopsis thaliana.

Wild Chrysanthemums: Important Bioresources for Chrysanthemum Studies

A pure line of C. boreale with self-compatibility was generated and it will accelerate not only the understanding of biological process at molecular levels in the genus Chrysanthemum but also the generation of new genetically engineered chrysanthemums (Nakano et al. 2019). Recently, it was found studies on biological process in C. boreale is increasing (Fig. 1). Publications were surveyed and total 71 (16 publications from C. seticuspe, a synonym of C. boreale) were identified from the citation database using scientific name (Web of Science, as of August 5th, 2021). The most remarkable aspect in this survey is 49 out of 71 (69%) articles were published during the last ten years implying C. boreale is gaining more attention in the field of plant science, pharmacology, medicinal chemistry, etc.

Second, the pure line (Gojo-0) of C. seticuspe with self-compatibility has been generated and it is expected that Gojo-0 will be a useful model platform to accelerate genetic studies at molecular levels in chrysanthemums (Nakano et al. 2019). In addition to this, the whole-genome sequence database of C. seticuspe was open to public a couple years ago (Hirakawa et al. 2019). Ornamental Chrysanthemum is hexaploidy and self-incompatible while the chromosome of C. seticuspe is diploid with simpler chromosomal structure (2n = 2x = 18) and some strains (Gojo-0, XMRS10) are self-compatible. A homologous line of C. seticuspe (XMRS10) was generated from self-compatible mutant line (AEV2) by reducing heterogeneity through multi generations by selection and whole genome sequence of XMRS10 was analyzed (Hirakawa et al. 2019). Assembled sequences (2.72Gb) of XMRAS10 covered 89.0% of C. boreale genome (3.06Gb) and 71,057 genes which are expected to encode proteins were annotated. Six varieties of ornamental chrysanthemums (fall color, fendian, jinba, nannongxunzhang, youxiang, zaoyihong) were used to evaluate the quality of genome of C. seticuspe. As a result, 954,706 of single nucleotide polymorphism (SNPs) were found and 294,601 SNPs were present in all six varieties. In addition, large-scale variation at a gene level was suggested due to a large number of SNPs which were missing according to different varieties (Hirakawa et al. 2019). Furthermore, whole genome sequence of chloroplast of C. boreale were investigated with 298 SNPs and 106 insertions/deletions (indels) were identified from around 151,000 bp according to strains (Tyagi et al. 2020).

Positional cloning was tried by comparing ALB1 (a locus which is known to be involved in chloroplast and chlorophyll development) between Gojo-0 and an albino mutant obtained from the first generation of AEV2 by selfing and deletion of one nucleotide in exon 2 was found in albino mutant implying that Gojo-0 can be a representative model platform for chrysanthemums and will provide important information for genomics studies and breeding of chrysanthemums (Nakano et al. 2021). In addition, C. seticuspe has been used as a model plant in functional genomics to understand flowering control of chrysanthemums. The response of plants on the day length was first reported from tobacco in 1920 and plants were categorized into three groups depending on the response to the day light; long day plants (LDPs), short-day plants (SDPs), and day-neutral plants (DNPs) (Garner and Allard 1920). Florigen, which is considered hormone like molecule, has been known to be responsible for flowering control in plants and flowering locus T (FT) protein is known as a component of florigen (Chailakhyan 1936;Tsuji 2017). Including FT protein, the other proteins, TERMINAL FLOWER 1 (TFL1)-like proteins, are involved in floral initiation in response to photoperiodism. Two proteins are antagonistically involved with each other because FT protein acts as a signal to promote flowering (florigen) while TFL1-like proteins inhibit flowering (antiflorigen). CsFTL1, CsFTL2, and CsFTL3 are FT-like genes identified in C. seticuspe. In addition, four TFL1/BFT-like genes (CsAFT, CsTFL1, and two candidate genes) were also identified (Higuchi et al. 2013;Higuchi et al. 2015;Hirakawa et al. 2019;Oda et al. 2012). Oda et al. (2012) generated overexpression line of CsFTL3 gene (FLOWERING LOCUS T-like gene) which stimulates flowering even in long day condition implying the potentiality of developing photoperiodism-independent chrysanthemum cultivars in the future.

In C. indicum, total 313 articles found to use C. indicum from the citation database which is about four times more than those of C. boreale. 218 articles were published during the last ten years implying C. indicum is also gaining a great attention more and more (Fig. 1). However, the field of studies are unequally distributed to pharmacology/pharmacy, medicinal chemistry, food science technology, integrative complementary medicine, multidisciplinary chemistry, or biochemistry molecular biology rather than horticulture, plant Sciences, and agronomy. This is because C. indicum is well known wild chrysanthemum which has been used as a conventional medicine in China for a long time and C. indicum was used to analyze functional compounds and to evaluate biological activities on disease, symptoms, or fungi (Shao et al. 2020). 2,575 simple sequence repeat (SSR) marker were identified from C. indicum (Han et al. 2018). A precise genomics sequence has not been reported from C. indicum yet but genomics research using NGS technique will further facilitate development of DNA markers from C. indicum to improve biological activities

Genomic Database of Wild Chrysanthemums

During the last decade, by virtue of a great advance in sequencing technique of nucleotides, so called next generation sequencing (NGS), whole genome sequences of numerous plants were identified including ornamental chrysanthemums, C. boreale, C. indicum and other Chrysanthemums (Fu et al. 2021;Hirakawa et al. 2019;Nakano et al. 2019;Sasaki et al. 2017;Song et al. 2018;Song et al. 2020). Whole genome sequence of C. seticuspe is provided from Mum GARDEN database (Mum Genome And Resource Database Entry, http://mum-garden.kazusa.or.jp/) (Table 1). Hirakawa et al. (2019) used Hiseq 2000 and Miseq techniques using a homogeneous line (XMRS10) obtained from C. boreale mutant (AEV2) which has self-compatibility. The target gene can be found using a name of target gene using keyword search and also Mum GARDEN offers BLAST (Basic Local Alignment Search Tool) using two genome database (CSE_r2.0 genome and CSE_r1.0 genome) and four type of gene database (CSE_r1.1, CSE_r1.1_maker, CSE_r1.0, CSE_r1.0_maker). Whole genome sequence of XMRS10 and Gojo-0 is also provided from the portal site for plant genome and marker, Plant GARDEN (Genome And Resource Database Entry, https://plantgarden.jp/ja/index).

In addition to Mum GARDEN and Plant GARDEN, Chrysanthemum Genome database (http://www.amwayabrc. com/index.html) comprehensively provides whole genome sequence of C. nankingense and transcriptome database obtained from C. morifolium (Song et al. 2018) (Table 1). Both genome sequence and annotated data of C. nankingense can be obtained as a fasta format for further analysis and transcriptome data from different organs root, stem, leaf, bud, tubular and tongue flowers are also available.

Compositae Genome Project (CGP) had been launched a few years ago but this project was stopped for certain reasons. The CGP had provided expressed sequence tag (EST) databases from several Asteraceae plants including lettuce and sunflower. Unfortunately, the database has not been working for a long time and only the main page is remaining (https://compgenomics.ucdavis.edu/archive/) (Table 1). However, EST sequence assembles of some plants can be obtained by directly accessing the Michelmore Lab at UC Davis (https://cgpdb.ucdavis.edu/asteraceae_assembly/) and it is expected that the database will be open in the near future again.

Genetic information of three wild Chrysanthemums has been registered National Center for Biotechnology Information (NCBI) database and can be found using Taxonomy browser (https://www.ncbi.nlm.nih.gov/Taxonomy/Browser/wwwta x.cgi, accessed on August 9th, 2021) (Federhen 2012) (Table 2). Three genome information of C. boreale were registered to BioProject (Accession PRJNA453211, PRJNA413827, PRJNA392380). PRJNA4534211 is registered as RefSeq genome data and the data provide two genomic DNA sequences obtained from mitochondrion (NC_039757.1) and chloroplast (NC_037388.1), respectively. In addition, 119 protein sequences have been registered to NCBI database. The project data type of other two genome sequences, PRJNA413827 and PRJNA392380, is raw sequence reads but full genome sequences in the database is not available. Twelve projects of genome sequences of C. indicum have been registered to BioProject so far. The types of data registered are three raw sequence reads (PRJNA687838, PRJNA648854, PRJNA335710), one assembled genome sequence (PRJNA421412), seven transcriptomes (PRJNA361213, PRJNA266706, PRJNA245057, PRJNA244464, PRJNA239175, PRJNA227262, PRJNA227261), one RefSeq genome sequence (PRJNA190114), respectively.

In addition, the size of genome can be easily identified from Genome size in Asteraceae database (GSAD, https:// www.asteraceaegenomesize.com/). The database offers genome sizes of 1,555 species in Asteraceae and 82 genome size of Chrysanthem species has been registered so far (Garnatje et al. 2011).

Conclusion

Ornamental chrysanthemums have hexaploidy (2n = 6x = 54) and autoploidy structure of chromosome and many agronomically important traits are controlled by number of genes. It was difficult to understand biological process controlled by a particular gene at molecular level because of polygenic control of many important agronomical traits. In addition, it was not possible to generate a model plant for chrysanthemum due to self-incompatibility and hybridization by cross-breeding was the main discipline for breeding of chrysanthemums. Physical and chemical mutagenesis have been also used to develop new cultivars. Flower color, flowering control, plant architecture, resistance to virus, fungi, and insect are main traits in both of conventional breeding and genetic engineering. During the last two decades, a variety of novel traits have been introduced using genetic engineering. Genetic engineering enables to develop brand new cultivars by introducing a particular gene that chrysanthemum does not have. Introduction of F3`,5`H gene into chrysanthemum is a representative example because blue-colored chrysanthemum is the first commercialized GM chrysanthemum in the world.

Advances in sequencing technique of nucleotide accelerated the understanding biological processes in number of ornamental plants by analyzing whole genome sequence and transcriptome datasets (Yagi 2015). NGS technique has contributed to develop large-scale DNA markers to select new cultivars with novel traits from agronomically important crops such as rice, maize, soybean, tomato, etc. (Torkamaneh et al. 2018). In chrysanthemum, self-compatible wild chrysanthemum (XMRS10) was generated from a mutant strain of C. seticuspe and whole genome sequence has been open a couple years ago (XMRS10) and a pure line of C. seticuspe (Gojo-0) was also generated and it safely proceeded to tenth generations by selfing (Nakano et al. 2021). In addition, positional cloning of a particular gene using a model platform Gojo-0 is now identified and seeds of Gojo-0 are now being provided as a model plant for research purpose through National BioResource Project: Chrysanthemum (NBRP: Chrysanthemum, https://shigen.nig.ac.jp/chrysanthemum/).

As discussed above, many important resources are now easily accessible such as; whole genome sequence information, transcriptome data, a model platform, database of nucleotide datasets, genetic engineering methods, NGS techniques, marker assisted selection, etc. Comprehensive studies in combination with such important resources will shed light on the chrysanthemum studies and breeding of ornamental chrysanthemums at a molecular level. In addition, contents in this review, which provide the reader with comprehensive information will pave the way for the successful research on ornamental chrysanthemums.