Lily, the king of flower bulbs, belongs to genus Lilium which comprises of around 100 species, 7 sections and ~ 10,000 documented cultivars (De Jong 1974; Matthews 2007). Lilies are high demanded cut flower in international flower trade due to its wide diversity of flower color, attractive flower shape, long multi-flowering stalk, and having long post-harvest shelf life (Lucidos et al. 2013). Lilium is well distributed in mostly Northern Hemisphere i.e. South-East Asia, North America, and Europe. Compared with other ornamentals, lily systematic breeding history is short and first Asiatic hybrid cultivar was developed in the 19th century (Jefferson-Brown and Howland 1995; McRae 1998). Classification of Lilium has been carried out during the last few decades on the basis of morphological parameters (Comber 1949), and then Jan de Graaff classified only hybrid lilies into different sections (Peterson 1960) but most recent classification is based on the genetic makeup (Miculka 2012). Advances in DNA techniques provide more accurate taxonomic information about Lilium species and cultivars (Horning et al. 2003). Each species in the genus Lilium possess great genetic diversity in its growth habit, flower color, form, shape, size, and as well as in persistence. This diversity in species of agronomic traits offers a substantial germplasm and opportunities for the development of hardy, and healthy varieties for variable climatic zones (McRae 1998; Anderson et al. 2010). In Lilium breeding, possible combining tendency of four genomes viz. Asiatic, Trumpet, Oriental, and Longiflorum have increased the prospect to accumulate assorted characteristics in one plant. Wild lily species and old cultivars being precious genetic material have successfully been used for production of new cultivars (Anderson et al. 2009). Plants obtained through inbreeding and backcrossing are generally very uniform in phenotype compared with when bred with wild and traditionally grown species that creates a challenging situation for taxonomists due to its heterogeneous inheritance.
The lily (Lilium longiflorum, Easter lily) genome size is 77.1 pg/2C nucleus which is one of the largest among all plants i.e. ~ 550x of Arabidopsis thaliana (0.14 pg) (Anderson et al. 2010). In Lilium, substantial size of genome may be due to the large amount of repetitive sequences metacentric and sub telocentric chromosomes (Siljak-Yakovlev et al. 2003). Most of the natural lily species are diploid (2n = 24), but few species are triploid (2n = 3x = 36) that are sterile in nature.
Tetraploids are also being reported in some wild lilies which are usually fertile. Spontaneous polyploidy in lilies have been reported which is assumed due to interference of the reproducing cells during spontaneous meiosis (Siljak-Yakovlev et al. 2003). It is significant to elucidate the cytogenetic composition, molecular genomic, and genetic consequence of lilies in order to carry more systematic breeding to develop hybrid cultivars with both commercial and horticultural traits.
Different Approaches in Lilium Breeding
In Lilium breeding classical breeding proved to be effective to create genetic variation and development of new cultivars with new flower colors, form, and shapes, resistance to insect/ pest and diseases, and as well as tolerance for unfavorable environmental conditions. In first decade of 20th century, crossing in different Lilium species: L. davidii, L. maculatum, L. tigrinum, and L. bulbiferum within section sinomartagon were carried out and different hybrids were developed. In the 1970s, extensive research on breeding and flowering physiology of the Easter lily cultivars ‘Ace’ and ‘Nellie White’ were conducted in USA. In the early 1980s, two cultivars ‘White American’ and ‘Snow Queen’ were released in Holland (Roh 2011). Until 1980, Asiatic lily cultivars were leading among different lily sections and were ranked as 2nd most important flowering bulb cut flower after tulip. Pixie series of Asiatic hybrids were bred in USA and were very popular among potted flowering plants due to its short stature (Roh 2011). Asiatic hybrid ‘Enchantment’ was bred in US by Jan de Graff which was one of the famous lily cut flower. After 1990, popularity of Asiatic cultivars started declining and Oriental lilies took the major market share in lily’s flower trade due to its fragrant flowers having conspicuous shape with early flowering. ‘Stargazer’, ‘Casa Blanca’, and ‘Acapulco’ were prominent Oriental cultivars at that time. Recently, more than 300 Oriental and orienpet hybrid cultivars like ‘Sorbonne’, ‘Crystal Light’, ‘Oberto’, ‘Siberia’, ‘Buckingham’, and ‘Conca d' Or’ are popular in market due to excellent flower color, its attractive shape and fragrance although their forcing period is a bit more compared with Asiatic hybrids. In the last decade, numerous Oriental hybrids having large size flowers in different with early flowering habits have been marketed. Improvement in the Asiatic, Oriental, and Longiflorum groups were by using classical breeding techniques. Hand pollination and selection for forcing quality were carried out under greenhouse conditions by breeders. But the conventional breeding methods hampered due to its heterozygosity and self -incompatibility among the species of the different Lilium sections (Asano 1985).
Hybridization breeding within a section is comparatively easy compared with intersection crossing due to different fertilization barriers. Pre-fertilization and post-fertilization issues restrict the possibility of interspecific hybridization in lily. Poor pollen tube growth development which is due to stigma incompatibility causes the pre-fertilization hurdle (Asano 1980) whereas; aborted embryo or seed without endosperm are the main barriers of post fertilization (Asano and Myodo 1977). Cut style, intra stylar pollination, mentor pollen, ovule/ovary culture, and embryo rescue culture were successful methods to cope with these barriers. Now lily researchers focused their breeding objective to combine horticultural traits of Oriental (OO), Asiatic (AA), Trumpet (TT), and Longiorum (LL) lilies for commercial production. Therefore, resulted in the development of many new intersectional lily hybrids (Van Tuyl et al. 2002; Zhang et al. 2012). Many intersectional Lilium hybrid plants were developed through the grafted style method (GSM) and embryo rescue technique like L. longiflorum × Oriental hybrid, Oriental × Asiatic hybrid (Van Tuyl et al. 2002).
Many plant researchers (Asano and Myodo 1977; Asano 1980; Van Tuyl et al. 2002) have conducted interspecific hybridization successfully (Table 1). In the late 1990s, the first interspecific hybrids (LA) were developed through crossing between Longiflorum (L) and Asiatic (A). These LA hybrids had all color’s ranges of the Asiatic group, with fragrant, and elegant flower form of the Longiflorum (Grassotti and Gimelli 2011). These LA hybrids gained trumpet shaped flowers having long strong stalk with long vase life from the Longiflorum, whereas got upright calyx with bright warmer colors from Asiatic group. LA cultivars like ‘Menorca’, ‘Golden Era’, ‘Red Alert’, ‘Macallen’ are now days popular and high in demand in market (Fig 1). In year 2000, Longiflorum were crossed with Orientals to develop LO hybrids, which have a pleasant sweet fragrance with attractive flower shape (Grassotti and Gimelli 2011). Breeding through crosses between Oriental and Trumpet (OT) had brought revolution in flower colors in Lilium and its captured flower trade worldwide. Commercially important OT cultivars are ‘Yelloween’, ‘Baruta’, ‘Candy Club’ etc. (Fig 1) Dutch Lilium breeding firms are leading in the world and releasing ~ 100 new varieties every year. In 2009, Lilium × formolongi seed-propagated hybrids were developed that flowers without vernalization having frosttolerance and winter hardiness (Anderson et al. 2009). In 2010, first successful interspecific hybrids of L. longiflorum × L. brownii and L.x fomolongi × L. brownie were developed (Behzad and Kim 2011).
Introgression breeding and Polyploidization
Introgression is a significant basis for genetic variation in plant populations which is the foremost objective of interspecific hybridization. This will helpful in introduction of a limited number of qualities (gene flow) from the donor species to other species through repetitive backcrossing of interspecific hybrids with one of its parent species. In Lilium breeding, only F1 hybrid development is not sufficient as both desirable and undesirable characters of their parents can accumulate in a single variety (Zhou 2007). This makes further backcrossing essential in lily breeding but the limitation is that most of the 2n interspecific hybrids are sterile (Zhou 2007; Zhou et al. 2008). Hybrid sterility could be attributed to uneven chromosome segregation and as well as due to low chromosome pairing during meiosis (Asano 1982). To overcome this problem many plant researchers (Van Tuyl et al. 2002; Lim et al. 2004, Ramanna and Jacobsen 2003) suggested using unreduced or 2n gametes having somatic chromosome number. Polyploidization can solve this problem and efforts have been made to get fertility through chromosome doubling mitotically (mitotic polyploidization). Therefore, interspecific hybrid development followed by polyploidization had successfully employed to breed new cultivars (Barba-Gonzalez et al. 2005). To induce polyploidy, colchicine or oryzalin which can act as chromosome doubling and spindle inhibiting agents have effectively been applied in lily (Van Tuyl et al. 1992). It is reported that due to strict autosyndetic pairing in the progenies of intergenomic recombination, doubling in somatic chromosomes become limited (Lim et al. 2003). Aberration during meiosis can also helpful in intergenomic recombination in Lilium which has been reported as a preferred method for sexual polyploidization (Lim et al. 2003). Sexual polyploidization is therefore more valuable for lily breeding programs as intergenomic recombination can occur between parents (Barba-Gonzalez et al. 2004). The other reported significance of 2n gametes is that this recombination permits to achieve introgression (Lim et al. 2005). It is interesting that 2n gametes formation seemed to be consistent but their detection is a bit difficult and different approaches were testified for detection of 2n gametes formation such as flow cytometry (Van Tuyl et al. 1988), pollen size examination (Barba Gonzalez et al. 2005), and progeny analysis (Bingham and McCoy 1979). Induction of 2n gametes is quite possible through genetic selection, low and high temperature exposure (Lim et al. 2005), by applying caffeine to young flower buds and through N2O application (Barba Gonzalez et al. 2006).
Generally, allotriploid hybrids are sterile and cannot be used in breeding but Zhou (2007) reported that triploid LA can be crossed with diploid Asiatic cultivars. Zhou et al. (2011) gave a new hypothesis that in Lilium breeding during triploid x diploid/ tetraploid crosses, five similar genomes of endosperm are necessary for its development. This presented hypothesis can better describe the success or failure of 3 × X 2 ×/4 × crosses in Lilium. Lim and his coworkers (2001) reported that allotetraploids of interspecific hybrid F1 encountered complications for the homoeologous recombination among parent species. They stated that reduction in homoeologous recombination is because of preferential pairing of homologous chromosomes during meiosis (metaphase I). Breeding of diploid and allotetraploid plants resulted into the development of allo-triploid plants. Tetraploid plants can contribute to produce triploid progenies which are more vigorous, and healthier than diploids (Lim et al. 2001). Zhou et al. 2011 reported survival of aneuploid embryo resulted after crossing euploid endosperm of triploid × diploid/tetraploid and considered them good source for lilies cultivar breeding. But, aneuploidy lily cultivars sometime exhibit considerable phenotypic variations which lead to opportunity of odd-teraploids to be used as maternal parent in lily breeding programs (Zhou et al. 2013).
Even though in both approaches chromosome number can be doubled however their comparative efciency is different in terms of heterozygosity, epistasis, genetic variability, and tendency to transfer specic genetic characters (Barba-Gonzalez et al. 2004; Beuselinck et al. 2003). It was submitted that the sexual polyploidy by 2n gametes could create more genetic variations compared with polyploids produced through colchicine application (Beuselinck et al. 2003). 2n-gamete formation had open new avenues for lily breeding programs to expand the genetic basis of commercially cultivated hybrid varieties and that will defiantly increase the breeding potential (Van Tuyl 1990; Younis et al. 2014).
Genomic in situ hybridization (GISH)
Genomic in situ hybridization (GISH) is considered as an efficient method for detection of chromosomal recombination as it provides a clear difference between genomes. Through this technique we can visualize DNA as well as part of chromosome, by introducing labeled DNA probe to DNA of the target chromosomes. Allopolyploids of lily are considered ideal to employ genomic in situ hybridization (GISH) due to its well-differentiated large genome size. Different researchers have used this technique in Lilium for the analysis of genomic composition (Barba-Gonzalez et al. 2005), to identify the site of genomic recombination (Zhou et al. 2008), to understand the mechanism of unreduced 2n gametes formation (Chung et al. 2013). F1 hybrids developed through crossing of L. longiflorum and L. rubellum were verified and identified by using GISH (Lim et al. 2000). This sophisticated method was also used successfully for tracing different events in recombination of backcross-1 and backcross-2 progeny (Karlov et al. 1999; Lim et al. 2000; Chung et al. 2013). In lilies, intergenomic recombination allopolyploids are developed because of crossing over during meiosis; whereas in multivalents and bivalents pairing of non-homologous chromosome possibly helps in the development of gametes with actual chromosome rearrangements (Xie et al. 2013). Lim et al. (2000) reported this technique efficient in detection and determination hybridization quality when total genomic DNA probe to target chromosomes during somatic prophase and metaphase. After developing protocols for GISH it can be used for diverse and constant “painting” along paternal chromosomes and as well as low level of probe crosshybridization to maternal chromosomes in lilies and other ornamentals. In 2007, Zhou and his coworker conducted GISH analysis for 19 cultivars which were developed after from backcrossing between F1 hybrids (♀ Longiflorum × Asiatic) and Asiatic (♂) and reported that 17 triploid (2n = 3x = 36), and two aneuploidy (2n = 3x + 1 = 37). They further found that triploids were developed from the functional eggs (2n) which were denoted by the (♀) F1 hybrids due to rst division restitution that happened during meiosis in megasporogenesis (Zhou et al. 2007). It is also clear from (Zhou et al. 2008) finding that variations produced by n, 2n or aneuploid gametes can effectively be multiplied and as well as can be preserved as potential parents. Few triploid lily cultivars as female parent can be crossed with appropriate male parent to get aneuploidy progeny (Zhou et al. 2011; Zhou et al. 2012) and also, Zhang et al. (2012) considered allotriploid breeding as a future trend for new cultivar’s development in lilies as most of the allotriploid cultivars developed through 2n gametes produced by F1 distant hybrids. It is important to note that variations produced by different gametes can perform a key role in Lilium breeding, as lilies can be multiplied through scaling and micro propagation.
Fluorescence in situ hybridization (FISH)
Fluorescence in situ hybridization (FISH) is a consistent method applied in plant molecular cytogenetics for detection of changes in chromosomes and to recognize and focus the presence or absence of specific DNA sequences (Jiang and Gill 2006). This technique coupled with digital imaging systems (DIS) proved to be an efficient technology to generate physical maps of different plant species. This technique was successfully implemented by different ornamental plant breeders in Lilium breeding (Lim et al. 2001; Hwang et al. 2011; Wang et al. 2012). Lim et al. 2001 carried out karyotype investigation of L. longiflorum and L. rubellum through banding of chromosomes and FISH. They reported that during metaphase in L. longiflorum, rDNA probes of 5S and 45S indicated overlapping signals at proximal points of short arms of chromosome number 4 and 7. They also observed signal of 5S rDNA on secondary constriction of chromosome number 3, and one 45S rDNA signal adjoining to the 5S rDNA signal on the subdistal portion of the long arm of chromosome 3. Wang et al., 2012 conducted karyotypes study based on FISH analysis and chromosome arm lengths by using probes of 45S and 5S rDNA sequences on four China native lily species viz. Lilium brownii, L. leucanthum, L. regale, and L. duchartrei. Four pairs of homologous chromosomes having 6 signals of 45s rDNA and two 5s rDNA in L. regale, seven pairs of homologous chromosomes having twelve 45s rDNA and four 5s rDNA loci in L. duchartrei. Advancement in FISH methodology permits higher resolution with precise positioning of specific probes compared with traditional methods (Schwarzacher and Heslop- Harrison 2000). FISH technique was also used by Hwang et al. (2011) to create more precise karyotype investigations on 2x and 3x L. tigrinum chromosomes by using ribosomal DNA. Comparative difference between 2x and 3x was revealed when 45S rDNA signals in diploid were noticed; both 6 pair (12 loci) and 5 pair (10 loci) were detected while triploid exhibited only 5 pairs (15 loci). This technique proved useful in detection of interspecific hybrids and chromo-somes numbers in backcross plants in Lilium as well as different other plant species (Inceer et al. 2002; Leitch and Heslop- Harrison 1992; Jiang; Xie 2010). Furthermore, the karyotypes study proved indispensable in mapping of genes present on chromosome arms and it has the ability to locate these chromosome fragments in generations that bred after frequent backcrossing.
The foremost value of molecular approaches over conventional approaches in plant breeding is depicting of true picture of plant characteristics despite of phenotypic of homoplasious characteristics and environmental effects (Selkoe and Toonen 2006; Van Tienderen et al. 2002). The application of different molecular markers in Lilium is mainly focused on lily classification, species identification, genetic diversity estimation, genetic relationship evaluation, genetic maps construction, and marker assisted breeding. But the application of genetic marker in lilies is limited compared with other plant species, due to the large genome size of Lilium (Anderson et al. 2011). Random amplified polymorphic DNA (RAPD) technique has been considered as a rapid approach to investigate genetic diversity within and among populations at the molecular level (Fritsch and Rieseberg 1996; Riaz et al. 2011; Williams et al. 1990). In 2009, Huang and his coworkers conducted RAPD analysis of 199 Lilium brownii samples which were collected from Guangdong, China. They reported this technique efficient for estimation of genetic diversity and for studying genetic structure of the L. brownii natural populations. Interspecific hybrids verification by using SCAR and PCR-RFLP markers was also carried out in lily oriental hybrid ‘Lombardia’ and L. longiflorum ‘White Heaven’ (Nesi et al. 2011). Abe et al. (2002) carried out RAPD and ISSR for genetic study of flower traits like anthocyanin pigmentation in hybrid cultivars of Asiatic lilies. Among different molecular marker approaches, the simple sequence repeat (SSR) has added advantages compared with other molecular techniques like, high reproducibility, high polymorphism and more capacity for transferability (Park et al. 2009). Even though lily is a commercially important flower crop, application of SSRs markers were not carried out extensively except of 6 SSRs in L. philadelphiicum (Horning et al. 2003). 165 EST-SSRs markers were developed by (Lee et al. 2011), in which 17 from Lilium regale, 23 belongs to Lilium formosanum whereas, 92 and 33 from Lilium longiflorum and Lilium hybrids, respectively. They identified 26 EST-SSRs markers for the amplification of lilies and these could be important for marker-assisted selection in lily interspecific hybridization. These identified EST-SSRs markers will offer substantial molecular means for molecular-assisted breeding, genetic diversity studies, germplasm management, and as well as fingerprinting in lilies (Lee et al. 2011). Inter simple sequence repeats (ISSR) method reveals more polymorphism because in these sequences, variability exists in repeats (Levinson and Gutman 1987). Due to its variable SSR-based primers and its ability for higher-stringency DNA amplifications this is considered more reproducible fingerprinting than that of RAPD (Anderson et al. 2009). This fingerprinting method is reported as a more reliable tool for studying genetic differences in closely related plants having little polymorphism (Hale et al. 2005). In many studies, ISSRs have successfully been instigated in genetic and fingerprinting analysis of Lilium [e.g., Asiatic Lilium (Abe et al. 2002; Yamagishi et al. 2002), Lilium longiflorum hybrids (Wang et al. 2009), interspecific (Lilium formolongi × Lilium martagon hybrids (Anderson et al. 2009), Easter lily ‘Nellie White’ (Anderson et al. 2010)]. Amplified fragment length polymorphism (AFLP) is a consistent and reliable technique valuable for genetic as well as clonal integrity analysis. But this technique was reported as inappropriate for those individuals that large genome size like lilies (Zlesak et al. 2007).
Biotechnology / genetic transformation approaches
Recent developments in Lilium breeding is the introduction of genetic transformation for an effective integration of disease resistance in lilies. Different plant breeders working on Lilium heightened the advantages of transformation technique for the incorporation of well identified genes directly into elite varieties, retaining their noble quality traits and adding desired characteristics (Azadi et al. 2010a; Cohen 2011; Hoshi et al. 2004; Li et al. 2008; Mercuri et al. 2003; Ogaki et al. 2008; Thao et al. 2008). Transformation mediated through Agrobacterium and biolistic has effectively been practiced with lily (Hoshi et al. 2004), even-though it was challenging to transform lily as drastic decrease in pH produced due to some unknown substances secreted through Lilium calli inhibited the growth of Agrobacterium. With addition of MES (2-morpholinoethanesulfonic acid) in the culture media has resolved the problem associated with low pH in the co-cultivation medium and transgenic plants were successfully developed (Ogaki et al. 2008). By exploiting this method, transgenic Lilium × formolongi plants having multiple genes of the carotenoid biosynthesis pathway was identified. It facilitated in the production of astaxanthin that is responsible for the orange color in the Longiflorum hybrid’s flowers (Azadi et al. 2010b). Novel transformation system was developed by complete exclusion of KNO3, KH2PO4, CaCl2 and as well as of NH4NO3 from culture media which increase the transformation efficiency in lily (Azadi et al. 2010a). By employing this method, transgenic plants having genes resistant against cucumber mosaic virus (CMV) were successfully achieved in one of the Oriental hybrid cultivar, ‘Acapulco’ (Azadi et al. 2011). Cohen and Meredith (1992) first revealed that Agrobacterium C58 had the ability to infect and transformation in lilies but this infection was strain-dependent which was confirmed by Langeveld et al. (1995). Micro projectile bombardment mediated transformation was used successfully in lilies to produce transgenic plantlets (Watad et al. 1998) and attempt leads to successful transformation in lily plants (Cohen 2011; Irifune et al. 2003; Kamo and Han 2008; Lipsky et al. 2002). Abnormal flowers with abnormal pollen and without viable seeds in transgenic L. longiflorum were observed. The possibility to pool products gained through classical breeding and genetic engineering could be feasible in L. longiflorum. Developing plants having normal flower shape with viable seeds following genetic engineering approach can be a substantial success in producing hybrids (Roh 2011). During last decade attempts have been made to utilize the genetic engineering tools to develop new lily cultivars but not yet succeeded (Cohen 2011). For a consistent, reproducible reliable method, efficient gene-delivery system is required that helps in the selection of a sufficient number of stable transformants.
Future prospects in lily breeding
To flourish the Lilium flower industry, development of novel cultivars with better ornamental, agronomic as well as commercial traits are required, based on total understandings of mechanisms and procedures involved during breeding, specific developmental processes, and as well as application of new biotechnological tools. Transformation is not intended technique to substitute conventional breeding methods, but rather to complement them therefore; it has potential to become another method in the development of new lily varieties with some novel characteristics. This may comprise, the introduction of new colors, flower’s forms, disease and insect/pest resistance, enhanced abiotic, and biotic stress hardiness either through genetic engineering or by classical breeding method. As the flowers are mainly used for aesthetic beautification and are not consumed for food therefore there will be less community opposition to transgenic flowering plants as they do not create any risk to human health, and safety to the environment. It can be concluded that classical breeding approaches should greatly be improved through collaborative approaches incorporating molecular markers applications, transgenic genes transfer, and as well as by using functional and structural genomics. Prospective advan-tages of linking genomic tools with conventional breeding have been a basis of widespread interest and that will help to accomplish the desired synergy among different lily groups for its improvement.