Chromosome characterization and physical mapping of 18S rDNA in Lilium longiflorum originated interspecific hybrids using combined genomic and fluorescent in situ hybridization

This study was aimed to differentiate parental genomes, examining intergenomic composition, and mapping mitotic metaphase chromosomes by localizing parental and 18S rDNA probes in seven interspecific hybrid progenies that originated from Lilium longiflorum. Since single method of in situ hybridization has been widely used in interspecific lily breeding, while in this study, flow cytometry was used in conjunction with genomic and fluorescent in situ hybridization to determine the genomic contribution of each parent to the interspecific progenies. A significant variation was observed in the DNA content, chromosome length, and 18S loci in F1 as compared to the female and male parents. L. longiflorum showed nearly two times higher DNA content than the male parents and L. longiflorum × Asiatic progenies, but eight times higher than L. longiflorum × L. hansonii. Genomic in situ hybridization results revealed that both female and male parents contributed an equal number of chromosomes to their interspecific F1 offspring. Fluorescent in situ hybridization mapping revealed that 18S rDNA had 8, 6 and 7 loci in L. longiflorum parents, i.e., White heaven, Bright tower, and White tower, respectively, whereas each Asiatic cultivar and L. hansonii used as male showed 8 and 12 loci respectively. Interspecific progenies showed 8 and 7 loci in LA, and 10–11 in LM hybrids. These cytogenetic results implied equal genetic and chromosomal contribution from both parents to their intergenomic progenies. Therefore, this combined cytogenetic method has the potential to be genome specific and time-saving approach in lily breeding that could determine the status of hybrids and their genomic origin while achieving physical mapping and detecting genes in different genomes.

. Lilium longiflorum Thunb. (2n = 2x = 24) is an important species involved in modern breeding programs, which has white, funnel-shaped flowers with a distinctive fragrance, low temperature tolerance, and a non-tunicate bulb but susceptible to the virus (Karlov et al. 1999;Lim et al. 2008). Asiatic lilies are widely used in breeding owing to their uprightfacing, early to late blooming, and vigorous and longliving flowers with a wide range of colors and their resistance against Fusarium oxysporum and viruses (Merhaut and Newman 2005;Lim et al. 2008). L. hansonii, belonging to the Martagon section, is a dwarf native species of the Korean Peninsula, growing under shady deciduous trees. It has 3-4 layers of whorled leaves and 8-10 pendulous flowers with yellow petals and numerous spots that have a long vase life (Lim 2014;Kumari et al. 2018). However, weak growth and virus susceptibility of L. hansonii is causing a decline in the number of cultivated plants every year. At present, there is no reported an interspecific lily hybrid originating from L. hansonii. LA (L. longiflorum × Asiatic) hybrids play an important role in commercial breeding as they introduce various flower colors, sizes, shapes, and distinct fragrances (Lim et al. 2008;Lucidos et al. 2013;Tang et al. 2020), and in some cases, resistance to Fusarium and other viruses (Liu et al. 2008). Therefore, interspecific hybridization among L. longiflorum, Asiatic, and L. hansonii may contribute to the creation of modern cultivars with desirable horticultural traits such as dwarfness and 6-8 medium-sized flowers with colorful petals, which could meet market demand.
In a majority of the crops, interspecific hybridization is one of the most essential methods for creating novel hybrids and improving desired features. In a taxonomic section, it is quite simple method to induce novel hybrids through normal pollination. Due to pre-and post-fertilization barriers, it is difficult to achieve interspecific hybrids. Moreover, poor pollen tube growth caused by stigmatic incompatibility is the first barriers, where incompatible pollen grains are unable to digest specific compounds in the stigma, depleting the stigma's own stores and resulting in poor growth. The second obstacle is seeds lacking endosperm cause the embryos to abort in the early stages of the development. Various techniques have been developed to overcome pre-fertilization barriers, such as cut-style pollination, grafted style method, intra-styllar pollination, mentor pollen, and in vitro pollination (Van Tuyl et al. 1997). Interspecific crossing between L. longiflorum and Asiatic hybrids most often result in sterile F 1 progenies due to failure of chromosome pairing during meiosis (Van Tuyl et al. 1997;Karlov et al. 1999). For efficient breeding, 2n gametes can be used, which play an important role in transferring desired parental genes to the F 1 and chromosome recombination during meiosis (Schwarzacher et al. 1992;Karlov et al. 1999).
Fluorescent in situ hybridization (FISH) is an advanced technique used to detect DNA sequences and allele copy numbers in different positions on chromosomes Jo et al. 2019;Islam et al. 2020). Analysis of nuclear organizing regions can be achievable using FISH technique (Fujiwara et al. 2007;Hwang et al. 2011). However, genomic in situ hybridization (GISH) is the most useful technique to differentiate whole parental genomes (Schwarzacher et al. 1992;Karlov et al. 1999;Chester et al. 2010;Van Laere et al. 2010;Silva and Souza 2013;Ramzan et al. 2017), alien chromosome segments, genome association in allopolyploids, and intergenomic recombination (Karlov et al. 1999;Nakazawa et al. 2011;Silva and Souza 2013;Ramzan et al. 2016). Both FISH and GISH are usually employed separately in Lilium breeding, which represents an expensive and time-consuming approach. Moreover, whole parental genomes and particular nucleic acid sequences cannot be both identified by using either FISH or GISH alone, and no combined GISH-FISH protocol has been published yet on lilies. Therefore, the purpose of this study was to introduce novel interspecific lily hybrids having parental attributes and to investigate status of hybrids, origin of genomes, chromosomal abnormalities, and chromosomal recombination in interspecific L. longiflorum × Asiatic, and L. longiflorum × L. hansonii hybrid progenies through combined GISH and FISH.

Plant material
Three L. longiflorum cultivars (L-genome), three Asiatic (A), and two L. hansonii (M) were used for interspecific hybridization experiments to produce F 1 progenies. L. longiflorum cultivars "White heaven", Page 3 of 15 133 Vol.: (0123456789) "Bright tower", and "White tower" and Asiatic cultivars "Conception", "Gironde", and "Sky shim 2" were collected from a local commercial company. Whereas, highly genetically diversified L. hansonii was collected from various locations of Ulleungdo Island, Republic of Korea and cultivated in the greenhouse condition. All the bulbs were diploid (2n = 2x = 24) and were grown in a greenhouse at Kyungpook National University, Republic of Korea. The greenhouse temperature was maintained at 14-18 °C during the night and 20-25 °C during the daytime. The most healthy, ready to open L. longiflorum as female flower and newly opened mature male pollen were selected for the interspecific breeding. A detailed breeding scheme is shown in Table 1. As L. longiflorum, Asiatic, and Martagon lilies belong to different taxonomical sections, interspecific hybridization was performed through the CSM (Lim et al. 2005) to avoid any fertilization barrier as well as to protect the hybridized ovary from any undesired pollen entry. A brief diagram of CSM breeding in lily has been shown in Fig. 1.

Flow cytometry analysis
Genome sizes and ploidy levels were determined by flow cytometry. For the analysis, 0.5 cm 2 fresh leaf tissue was placed in a petri dish (SPL 10060) and chopped with 500 μL Nuclei Extraction Buffer (CyStain UV Precise P, Sysmex) using a sharp blade. The extraction buffer, containing exposed nuclei, was filtered through a 30 μm-nylon mesh filter into a 3 ml-tube and stained in a staining buffer (CyStain UV Precise P, Sysmex). After a short incubation of 30 s, nuclear suspensions were analyzed using a flow cytometer (Partec PA, Ploidy Analyzer, Germany). Mean DAPI fluorescence of the target samples was compared with an internal standard, Oriental-Trumpet "Yelloween" (Ramzan et al. 2016;Kwon et al. 2017). Genome size (2C) was calculated based on the ratio of relative DAPI fluorescence of a sample to that of the internal standard. The calculation procedure was as follows: DNA content of the standard × (mean of relative DAPI fluorescence of a sample/mean of relative DAPI fluorescence of the standard) (Hembree et al. 2020).

Chromosome observation
Young, actively growing root tips were collected from interspecific F 1 hybrids. To ensure the presence of metaphase chromosomes, root tips were pretreated with α-bromonapthalene for 4 h at 20 °C, followed by fixation in 3:1 (v/v) solution of 100% ethanol: glacial acetic acid for 24 h at room temperature (25 °C). The fixed root tips were stored in 70% ethanol at − 20 °C until use (Takahashi et al. 1997;Kwon et al. 2017). For slide preparation, root tips were rinsed 2-3 times in distilled water, incubated in a pectolytic enzyme mixture containing 0.2% (w/v) cellulase RS, 0.2% (w/v) pectolyase Y23, and 0.2% (w/v) cytohelicase in 10 mM citrate buffer (pH 4.3) at 37 °C for 70 min. Dissection and squashing of macerated root tips were made in a drop of 70% acetic acid that was subsequently air-dried. Finally, well-spread chromosomes were identified using a fluorescent microscope (BX61, Japan).

DNA isolation and preparation
For GISH, non-shared parental genomic DNA was isolated from L. longiflorum "Woori tower" using the CTAB method with some modifications. DNA concentration was measured by NanoDrop 2000 (Thermo Fisher Scientific) using 1 µL per sample, and three measurements were performed for each sample. The isolated parental genomic DNA was used as a probe and was labeled with digoxigenin-11-dUTP using a standard nick translation kit according to the manufacturer's instructions (Roche Diagnostics GmbH, Germany). Labeled genomic DNA was used in a volume of 20 µL, including 1300 ng genomic DNA, 4 µL DIG-Nick translate, and extra distilled water. For FISH, a 18S rDNA probe was directly labeled with biotin-16-dUTP by nick translation in accordance with the manufacturer's protocol (Roche, Germany). The blocking DNA was obtained from herring sperm (Invitrogen, USA) that was fragmented into 100-300 bp fragments and subsequently autoclaved.

Combined genomic-fluorescence in situ hybridization
The combined GISH-FISH hybridization procedure can be divided into three parts: chromosome pretreatment, combined hybridization, and probe detection (Fig. 1). The selected slides were treated with 100 µg/ ml RNase for one hour at 37 °C, washed three times with 2 × SSC for 5 min each, and fixed for 10 min with 4% paraformaldehyde. Again, the slides were washed thrice with 2 × SSC for 5 min each time, dehydrated with a 70%, 90%, and 99.9% ethanol series for 3 min each, and air-dried. Hybridization was performed in a mixture containing 50% (v/v) deionized formamide, 20 × SSC buffer, 10% (w/v) sodium dextran sulfate (SDS), 50% dextran sulfate (DS), 5 ng/µl L. longiflorum "Woori tower" genomic DNA, 20 µg/ ml each of the 18S rDNA probes, and 30-45 ng/µl blocking DNA. DNA was denatured by heating of the hybridization mixture at 74 °C for 10 min, followed by incubation in an ice jar for 15 min. For each slide, 40 µl of hybridization solution was used, and the slides were denatured at 82 °C for 5 min, followed by an overnight incubation in a humidified chamber at 37 °C. After hybridization, the slides were washed with 2 × SSC buffer for 5 min at room temperature and subsequently washed with 0.1 × SSC buffer in a shaking jar at 42 °C for 30 min. Digoxigenin-labeled probes were detected with antidigoxigenin fluorescein (Roche Diagnostics GmbH, Germany) and amplified with fluorescein anti-sheep and fluorescein antirabbit (Vector Laboratories). Similarly, biotin-labeled probes were detected using streptavidin-labeled Cy3 and amplified with biotinylated goat anti-streptavidin (Amersham Biosciences, UK). Finally, the treated slides, containing chromosomes, were counterstained with 3 mg/ml 4, 6-diamidino-2-phenylindole (DAPI) in Vectashield mounting medium (Vector Laboratory Inc., USA). Different methods, such as nick translation, random-primed labeling, and polymerase chain reaction, were used to label the probe during marker labeling. Various methods, such as autoclaving, shearing the DNA with a tiny needle in a syringe, or sonication, are used to prepare blocking DNA. Chromosome slide preparation included selection of well-spread chromosomes prepared from a young root tip using an enzyme mixture at 37 °C. Slide pretreatment was comprised of enzymatic digestion of chromosomes in order to unmask DNA prior to hybridization. Hybridization involved attachment of both blocking and probe/genomic markers with chromosomes to identify the specific loci/origin of the genome of the respective chromosome. During detection, attachment of the designed antibody against the target marker along with blocking buffer was performed to detect the specific fluorochrome.

Ploidy and DNA content analysis
Ploidy levels and genome size of female L. longiflorum cultivars, male Asiatic cultivars, L. hansonii and their F 1 progenies were assessed by flow cytometry and estimated by the comparison to the control "Yelloween" (Table 2). Genome size of F 1 progenies varied from 51.6 to 59.9 pg, which was similar to the male parents, i.e., 48.9-55.4 pg, but lower than that of the female plants, i.e., 68.9-70.1 pg (Fig. 2).  Fig. 3. Biotin-labeled 18S rDNA loci (red fluorescence) are marked by yellow arrows (Fig. 3), and karyotype ideograms and FISH summary are shown in Figs. 4, 5 and Table 3. Each of the seven interspecific hybrids and their respective female parents were found to be diploid (2n = 2x = 24). The number of 18S rDNA loci varied among parents and L. longiflorum-originated progenies. The female parents i.e., "White heaven," and "Bright tower", and "White tower," showed eight, six, and seven 18S rDNA loci, respectively. Whereas, male Asiatic lilies and L. hansonii showed eight and twelve 18S loci each, respectively. Out of eight 18S rDNA loci in L. longiflorum "White heaven," four were equally found on the long arm and the centromeric region. Similarly, "Bright tower" (two loci on the long arm and four on the centromeric region) and "White tower" (three loci on the long arm and four on the centromeric region) showed variation in terms of 18S rDNA loci distribution. In male parents, four loci on both long arm and centromeric region (Conception, Sky Shim 2), five loci on long arm and three on centromere (Gironde), eight loci on long arm and four on centromere (111114), and nine loci on long arm and three on centromere (121038-1) were observed as shown in Table 3 and Fig. 3 (Fig. 6).

Genomic composition of interspecific progenies
A minimum of 18 individual progenies of each cross combination were examined using GISH, as shown in Table 4, and representative images are illustrated in Fig. 7. According to the GISH results, all the seven interspecific progenies showed true hybridity and possessed three different genomes, namely, Longiflorum (L), Asiatic (A), and Hansonii (M), which have been clearly distinguished on the respective chromosomes. An equal number of 24 chromosomes in each individual, representing seven interspecific hybrids, was observed without any break points and recombination (Table 4). Since L. longiflorum was used as a female genomic DNA probe, 12 chromosomes of Longiflorum (113521, 132025, 143024, 128115-7, 128130-5, and 128133-8) and 12 chromosomes of Asiatic (113521, 132025, and 143024), and Hansonii (128115-7, 128130-5, and 128133-8) were characterized ( Fig. 7 and Table 4). However, due to the higher genetic diversity of L. hansonii, the number of 18S rDNA loci increased in F 1 hybrid progenies. In LL × AA-crossed progenies (113521, 132025), among the detected 18 rDNA loci, four were detected on the long arm (one locus from a female and three loci from a male) and four loci on the short arm (two loci from a female and two from a male). The remaining LL × AA hybrid (143024) showed seven 18S loci, including four on the long arm (one locus from a female and three loci from a male) and three loci on the short arm (two loci from a male and one locus from a male).
The number of 18S rDNA loci was higher in LL × MM progenies (128115-7, 128130-5, and 128133-8). Out of the 11 detected loci (128115-7), 7 were located on the long arm (1 locus from a female and 6 loci from a male) and 4 on the short arm (2 loci equally from a female and male). 128130-5 and 128133-8 progenies showed 10 and 11 18S rDNA loci, including 6 and 7 found on the long arm and 4 loci each on the short arm (2 loci each from female and male), respectively. Among the six interspecific progenies, all 18S rDNA loci were detected on the chromosome pairs 3, 4, and 7, except for 132025 (LL × AA) hybrids (Fig. 8).
Karyotype analysis of the female, male parents and interspecific progenies Diploid L. longiflorum, "White heaven", "Bright tower", and "White tower" as female parents, "Conception", "Gironde", "Sky Shim 2", L. hansonii (111114, 121038-1) used as male parents showed 24 chromosomes each. The longest and shortest chromosome of female parental lilies ranged from 17.19 to18.83 µm and 9.07-11.18 µm (Table 5), while it was 10.71-17.23 µm and 8.71-10.53 µm in male parents, respectively. Total length of all chromosomes in female and male parents were 259.90-281.94 µm and 211.54-245.89 µm, respectively. However, among the six interspecific hybrid progenies, number of chromosomes and their karyotype varied. Moreover, number of metacentric, sub-metacentric, and telocentric chromosomes also varied among them. The highest number of submetacentric chromosomes was scored among the female, male parents and their F 1 progenies. Among F 1 progenies, the longest and the shortest arm ranged from 15.27 to 19.33 µm and 7.59-11.10 µm, respectively, and total length of all chromosomes was 229.20-271.58 µm.

Discussion
Interspecific L. longiflorun × L. hansonii and L. longiflorum × Asiatic lilies have higher economic potentiality due wide phenotypic colour, fragrance and genetic diversities as compared to the existing hybrids. It has been reported to have a 2C value approximately 72 pg in L. longiflorum (Karlov et al. 1999), that found to vary genetically between the lily cultivars used in this study such as 70.1 pg (Bright tower; L), 55.4 pg (Conception; A), 49.2 pg (L. hansonii; M), 56.1 pg (LA), 51.6 pg (LM). As the male and female parents used in this study were: is diploid (2n = 2x = 24), it should be the interspecific progeny obtained diploid, and this was confirmed by both flow cytometry and observations. Similarly, triploid lily was obtained by crossing tetraploid and diploid lilies (Gao et al. 2014). In this study, cut style breeding method was used although interspecific hybridization is a challenging task in the genus Lilium breeding due to both pre and post-fertilization barriers, which can be overcome by using CSM and GSM crossing (Van Tuyl et al. 1997;Barba-Gonzalez et al. 2005). However, using the CSM method, a relatively low success rate of interspecific hybridization was achieved in both LA and LM, yet it was higher in comparison to the OT hybrids (Cao et al. 2019).
Cytogenetic variations are common phenomena in interspecific LO, LA and LM hybrids. According to Zhou et al. (2008a, b), only a few individuals out of a large number of interspecific hybrid progenies, including LA, can produce functional n gametes. Therefore, either high pressure N 2 O treatment or polyploidization agent such as colchicine, oryzalin treatment is needed to induce fertile gamate (Ramanna 1992;Lim et al. 2001;Barba-Gonzalez et al. 2005;Zhou et al. 2015). Genetic recombination and homologous crossing-over are important features   2, 3, 4, 7, 8,11, 12 L. hansinii (121038-1) 12 9 3 Ch# 1, 2, 3, 4, 5, 6, 7, 10, 11 Page 11 of 15 133 Vol.: (0123456789) of interspecific breeding, especially for LA and OA progenies (Lim et al. 2001;Van Tuyl et al. 2002;Sun et al. 2014). In this study, both female and male parents provided equal portions of genetic material to their interspecific progenies without any notable homologous recombination. L. longiflorum × Asiatic hybrid progenies usually show quantitative differences between genotypes that result in chromosome pairing failure; however, in this study, almost normal pairing was observed. Using advanced cytogenetic   22  2x  12  12  0  True  132025  LA  23  2x  12  12  0  True  143024  LA  19  2x  12  12  0  True  128115-7  LM  32  2x  12  12  0  True  128130-5  LM  18  2x  12  12  0  True  128133-8  LM  19  2x  12  12  0  True tools like genomic in situ hybridization, chromosome recombination can be easily detected and commonly found in BC 1 progenies, whereas in a very recent study, nearly 3% intergenomic recombination has been identified in F 1 L. longiflorum × L. hansonii progenies which is under further investigation to find out the recombination origin through meiotic stages analysis following Lim et al. (2003) and the result will be reported in the next article (Table 6). FISH mapping of 18S rDNA in L. longiflorum had 6-8 loci, usually on the third, fourth and eleventh pairs Chromosomes (Table 4), similar to (Lim et al. 2001) but different FISH mapping of 45S has been observed in L. lancifolium (Zhou et al. 2008b). Also, in Asiatic lilies, the 45S rDNA loci are stable on the first, fourth and seventh pair of chromosomes (Cao and Al. 2019). However, L. hansonii was found to be a very genetically diverse species on Ulleung-do Island, Korea (Robledo et al. 2009), and not of L. hansonii origin interspecific hybrids have not yet been reported. As a result, LM offspring inherited more 18S rDNA locus of the male parent L. hansonii, causing their numbers to increase. GISH and FISH are the most advanced cytogenetic techniques used for the evaluation of hybridity status, chromosomal abnormalities, homologous crossing and FDR mechanisms and identification of the origin of F 1 offspring (Robledo et al. 2009;Younis et al. 2015;Kwon et al. 2017;Ramzan et al. 2017). Hence the development of combined GISH-FISH analysis would be a new addition to cytogenetic studies on lily breeding using parental DNA and 18S rDNA to detect specific loci on respective chromosome, that will help to analyze chromosomal attributes and determine the karyomorphology of the F 1 interspecific offspring and their female parents.

Conclusions
The obtained interspecific hybrid progenies had equal genetic characters inherited from the female and male parents, according to the results of chromosome characterization and FISH mapping. In addition, this study illustrates the potential of combined GISH-FISH analysis for providing information about hybridity status, chromosomal abnormalities, and genomic origin of L. longiflorum-derived interspecific hybrid progenies and can be useful as a timesaving and low-cost cytogenetic method in breeding programs.