Platycodon grandiflorus (Jacq.) A. DC. is a temperature-sensitive perennial flowering plant (Yang et al. 2015), widely known as “bellflower” or “balloon flower,” that is a monotypic species of the family Campanulaceae (Hawke 2009). This species is native to East Asian countries (Wu et al. 2011). It has been used as ornamentals (Yan et al. 2012), foodstuffs (Lim et al. 2011) and folk medicine for various diseases (Xu et al. 1996). Studies also showed that this herbal plant contains phytochemicals that possess antioxidant, anti-inflammatory, anti-cancer and antimicrobial capabilities (Park et al. 2011) which are responsible for the plant’s numerous health benefits (Lim et al. 2011; Yan et al. 2012). These, in turn, have led researchers to enhance the quality of P. grandiflorus crop production and search for other potential health benefits through mitotic polyploidization (Kim et al. 2003; Dhooghe et al. 2011).

Kim et al. (2003) successfully reported the artificial polyploidization of P. grandiflorus. In their study, a tetraploid cultivar “Etteumbaek’ (2n = 4x = 36) was developed from its diploid cultivar ‘ Jangbaek’ (2n = 2x = 18) using colchicine. In nature, polyploidization is rare. It occurs in a slow and gradual process (Eng and Ho 2019). In fact, about 15% of angiosperm species were formed through this course (Wood et al. 2009). Along with its formation were the dynamic and stochastic changes in genomic rearrangements, genetic alteration, and expression (Paun et al. 2007). To better understand the chromosome and genetic stability of the tetraploid P. grandiflorus, molecular cytogenetic studies using fluorescence in situ hybridization (FISH) can be applied to support the limited cytogenetic and genomic studies of this plant species (Wu et al. 2012; Pellerin et al. 2017).

In many studies, FISH technique plays a vital role in clarifying genome structures, chromosome counts, physical locations of repeats and genes, and revealing genome rearrangements like chromosomal segment inversions, duplications and translocations (Koo et al. 2002; Pellerin et al. 2018). Useful cytogenetic markers like the highly conserved tandem repeats of 5S and 45S rDNAs provide suitable information about genome organization and chromosome identification (Heslop-Harrison 2000).

The 5S rDNA repeat unit consists of ∼120 bp coding region gene and a comparatively short intergenic spacer (IGS) with 100 to 1,000 bp long (Galián et al. 2012) can be found at more than one locus, either on the same or another chromosome (Wicke et al. 2011). Whereas the 45S rDNA repeat unit represents up to ∼10% of the plant genome. It forms the nucleolar organizer region (NOR) and consists of three coding regions (18S, 5.8S and 25S/26S/28S), internal transcribed spacers between genes (ITS1 and ITS2), and a long intergenic spacer (IGS) separating adjacent repeats (Pellerin et al. 2017; Waminal et al. 2018; Peniton et al. 2019; Zhou et al. 2019).

In this study, FISH karyotype analyses of Platycodon grandiflorus (Jacq.) A. DC. ‘Jangbaek’ and colchicine-induced tetraploid P. grandiflorus ‘Etteumbaek’ were carried out using 5S and 45S rDNA probes. The results will be useful in providing the chromosomal backbone for the comprehensive genome analysis and breeding programs for the Platycodon species.

Plant samples

Seeds of Platycodon grandiflorus (Jacq.) A. DC. ‘Jangbaek’ and colchicine-induced tetraploid P. grandiflorus ‘Etteumbaek’ were obtained from the Department of Herbal Crop Research, National Institute of Horticulture and Herbal Science, Rural Development Administration (RDA), Eumseong, Republic of Korea. The seeds were germinated in petridishes and moistened with distilled water until healthy roots emerged from the seeds. Root tips ∼2 cm long were harvested and pretreated with chilled 2 μM 8-hydroxyquinoline for 5 hours at 18℃. Finally, root tips were fixed in freshly prepared Carnoy’s solution and stored in 70% ethanol at 4℃.

Chromosome spread preparation

Following the method used by Waminal et al. (2011) and Pellerin et al. (2017), the treated meristematic root tips were enzymatically digested with pectolytic enzyme solution [2% Cellulase R-10 (C224, Phytotechnology Laboratories) and 1% Pectolyase Y-23 (P8004.0001, Duchefa) in 100 μM citrate buffer] for 1.5 hours at 37℃. The root tips were transferred into a microtube containing aceto-ethanol (1:3, v/v) solution and vortexed for 30 seconds. The supernatant was discarded, then washed with distilled water. The protoplasts were resuspended in aceto-ethanol (9:1 v/v) and spread onto slides that had been precleaned with 70% ethanol and prewarmed in a humid chamber. The suspension was then pipetted onto the pre-warmed slides in a humid chamber, air-dried, fixed in 2% formaldehyde for 5 minutes (Vrána et al. 2012), and subjected to a series of ethanol treatments (70%, 90%, and 100%) for dehydration (Pellerin et al. 2018). Good metaphase spreads without overlapping chromosome arms were scanned with the DAPI-counterstained slides under the fluorescence microscope. High-quality metaphase chromosome spreads were then used for FISH procedure.

Probe preparation

The 5S rDNA with 120 bp long was attained through PCR labeling using a PCR DIG Probe Synthesis Kit (Roche, Germany) and genomic DNA as a template (Koo et al. 2002). The 45S rDNA with a 9-kb fragment of 18S-5.8S-25S rRNA (Waminal et al. 2011) was labeled with biotin-16-dUTP by nick translation (Roche, Germany).

Fluorescence in situ hybridization (FISH)

The procedures carefully followed those of Waminal et al. (2011) and Pellerin et al. (2018), with slight modification from Karlov et al. (1999). Briefly, the hybridization mixture comprised of 50% formamide, 10% dextran sulfate, 2× saline-sodium citrate buffer (SSC), 5 ng/μL salmon sperm DNA and 500 ng/μL of each DNA probe and was adjusted with nuclease-free water to a total volume of 40 μL per slide. The mixture was denatured at 90℃ for 10 minutes and 40 μL was mounted onto each slide. Chromosomes were denatured at 80℃ for 5 minutes and incubated overnight in a humidity chamber at 37℃. Afterward, the slides were stringently washed using 2× SSC at 20℃-25℃ for 10 minutes, 0.1× SSC at 42℃ for 25 minutes, and 2×SSC at RT for 5 minutes, followed by dehydration in an ethanol series of 70%, 90%, and 95% at room temperature. Slides were air-dried and counterstained with 1 μg/mL of 4ʹ, 6-diamidino-2-phenylindole (DAPI) in Vectashield (Vector Labs, H-1000, USA) and observed under an Olympus BX53 fluorescence microscope equipped with a Leica DFC365 FS CCD camera using an oil lens (×100 magnification). The captured images were sorted using Cytovision ver. 7.2 (Leica Microsystems, Germany). Adobe Photoshop CS6 was used for image enhancement and karyogram preparation. Chromosome length was measured using ImageJ, ver.1.51k (Schneider et al. 2012), and the statistical analyses were carried out using GraphPad Prism v7.00 (Prism; Motulsky 2017).

The FISH karyotype was analyzed with mitotic metaphase chromosome spread and paired according to chromosome length, centromeric position and ribosomal DNA signals. The chromosomes were arranged in decreasing order based on length.

FISH karyotype

We properly inspected metaphase chromosomes using seed germination. Fresh roots germinated from the seeds of the two Platycodon cultivars on the moisturized filter paper in the petridishes for 3-5 days yielded healthy meristematic root tips necessary for metaphase spread preparation. Figs. 1 and 2 show the FISH metaphase spreads and the karyograms of the diploid Platycodon cultivar ‘Jangbaek’ and colchicine-induced tetraploid Platycodon cultivar ‘Etteumbaek,’ respectively. The chromosome compositions of ‘Jangbaek’ and ‘Etteumbaek’ cultivars showed 2n = 2x = 18 with one pair of satellite chromosomes and 2n = 4x = 36 with two pairs of satellite chromosomes, respectively.

Figure 1. Dual-color FISH detection of 5S (green fluorescence) and 45S (red fluorescence) rDNA probes on somatic metaphase chromosomes in the diploid cultivar ‘Jangbaek’ (A) and the colchicine-induced tetraploid cultivar ‘Etteumbaek’ (B). The arrows indicate the co-localization of 5S and 45S rDNA signals (yellow) in the nucleolar organizing region. Bars; 5 μm.

Figure 2. FISH karyotypic analysis of mitotic metaphase complements of the diploid cultivar ‘Jangbaek’ (A) and the colchicine-induced tetraploid ‘Etteumbaek’ (B). The chromosomes were arranged based on length, type, and 5S (green) and 45S rDNA (red) signals. The arrows indicate satellite chromosomes and co-localization of 5S and 45S rDNA signals (yellow) are shown in the nucleolar organizing region. Bars; 5 μm.

The results of karyotypic analysis based on chromosome lengths, type, and rDNA signals are shown in Table 1 and 2. The total chromosome length of ‘Jangbaek’ was in the range of 2.34 ± 0.13-3.99 ± 0.197 μm; its short arm, 1.15 ± 0.08-1.74 ± 0.11 μm and long arm, 1.19 ± 0.08-2.25 ± 0.10 μm. In case of the tetraploid cultivar ‘Etteumbaek’, those were estimated as 2.37 ± 0.08-4.20 ± 0.16 μm, 1.12 ± 0.07-1.52 ± 0.11 μm, and 1.25 ± 0.08-2.68 ± 0.13 μm, respectively. Considering the chromosomal length variation according to the mitotic stage, the chromosomal length difference was not considered as significant between the two cultivars.

Table 1 . FISH karyotype analysis of the diploid Platycodon grandiflorus (Jacq.) A. DC. ‘Jangbaek’ showing chromosomal lengths, types and number of FISH signals of rDNA observed.

Chr. no.	Chromosome length (μm)			Arm ratio(L/S)	Chr. type	FISH signal no. (n)
	Short arm (S)	Long arm (L)	Total (S+L)			5S	45S
1	1.74 ± 0.11	2.25 ± 0.10	3.99 ± 0.197	1.30 ± 0.06y	m	-	-
2	1.50 ± 0.09	1.99 ± 0.10	3.50 ± 0.17	1.33 ± 0.07	m	-	-
3	1.54 ± 0.07	1.93 ± 0.12	3.46 ± 0.15	1.25 ± 0.08	m	2	-
4	1.39 ± 0.09	1.76 ± 0.07	3.15 ± 0.07	1.27 ± 0.12	m	-	-
5	1.28 ± 0.16	1.72 ± 0.15	2.99 ± 0.10	1.37 ± 0.28	m	-	-
6*	1.06 ± 0.06	1.84 ± 0.03	2.90 ± 0.07	1.75 ± 0.09	sm	2	2
7	1.20 ± 0.08	1.44 ± 0.12	2.64 ± 0.17	1.20 ± 0.09	m	-	-
8	1.09 ± 0.05	1.37 ± 0.02	2.46 ± 0.05	1.25 ± 0.08	m	-	-
9	1.15 ± 0.08	1.19 ± 0.08	2.34 ± 0.13	1.04 ± 0.09	m	-	-
Total: 16m + 2sm						4	2

m, metacentric; sm, submetacentric; *, satellite chromosome; ^y, mean ± standard deviation.

Table 2 . FISH karyotype analysis of the colchicine-induced tetraploid Platycodon grandiflorus ‘Etteumbaek’ showing chromosomal lengths, types and number FISH signals of rDNA observed.

Chr. no.	Chromosome length (μm)			Arm ratio(L/S)	Chr. type	FISH signal no. (n)
	Short arm (S)	Long arm (L)	Total (S+L)			5S	45S
1	1.52 ± 0.11	2.68 ± 0.13	4.20 ± 0.16	1.77 ± 0.17y	sm	1	-
2	1.32 ± 0.06	2.28 ± 0.12	3.61 ± 0.12	1.73 ± 0.11	sm	-	-
3	1.52 ± 0.15	2.00 ± 0.17	3.52 ± 0.21	1.33 ± 0.18	m	2	-
4	1.43 ± 0.15	1.68 ± 0.11	3.11 ± 0.21	1.19 ± 0.12	m	-	-
5	1.22 ± 0.10	1.55 ± 0.08	2.77 ± 0.11	1.29 ± 0.15	m	1	-
6*	0.81 ± 0.11	1.68 ± 0.17	2.49 ± 0.19	2.12 ± 0.39	sm	4	4
7	1.18 ± 0.08	1.27 ± 0.15	2.45 ± 0.16	1.08 ± 0.15	m	-	-
8	1.10 ± 0.11	1.31 ± 0.10	2.41 ± 0.12	1.22 ± 0.18	m	-	-
9	1.12 ± 0.07	1.25 ± 0.08	2.37 ± 0.08	1.13 ± 0.13	m	-	-
Total: 24m + 12sm						8	4

sm, submetacentric; m, metacentric; *, satellite chromosome; ^y, mean ± standard deviation.

In chromosomal type based on the arm ratio as described by (Levan et al. 1964), the chromosome composition of the diploid cultivar ‘Jangbaek’ exhibited 16 metacentrics and 2 submetacentrics including one pair of satellite chromosome 6, with the karyotypic formula of 2n = 2x = 18 = 16m + 2sm (2 satellites). Those of tetraploid cultivar, 24 metacentrics and 12 submetacentrics including four satellite chromosomes and the karyotypic formula of 2n = 4x = 36 = 24m + 12sm (4 satellites) were revealed. The tetraploid ‘Etteumbaek’ showed the doubled chromosome composition of its donor diploid ‘Jangbaek’ cultivar.

Localization of rDNA probes

The somatic metaphase chromosome spreads of both cultivars displayed clear 5S and 45S rDNA tandem repeats signals (Fig. 1). In the diploid cultivar ‘Jangbaek,’ a pair of 5S rDNA signals were detected and noticed in the interstitial or pericentromeric region of chromosome 3 and the colocalization of 5S and 45S rDNA signals on chromosome 6 (Fig. 2A). Whereas in the tetraploid cultivar ‘Etteumbaek,’ only a pair of intense 5S rDNA signal was detected on chromosome 2 and dispersed minor signals on the single chromosome of chromosomes 1 and 5. The location of 5S rDNA varied in some regions of the chromosomes. For instance, the vague signal of 5S rDNA in one of the chromosomes 1 was found to be in the subtelomeric region and the two chromosomes in chromosome 3 were found in the interstitial or pericentromeric region alike with the signal in one chromosome of chromosome 5 (Fig. 2B). Both cultivars showed colocalization of 5S and 45S rDNA in the NOR regions of chromosomes 6 (Fig. 2). The co-localization of 5S and 45S rDNA probes displayed a yellow signal on the chromosomes.

Polyploid plants were established in agriculture and horticulture as they usually possess desirable agronomic traits over their diploid counterparts (Boo et al. 2013; Zhang et al. 2015). Since polyploidization occurs gradually in nature, the artificial occurrence of polyploid cultivar was achieved with the aid of some known anti-mitotic agents like colchicine (Dhooghe et al. 2011; Eng and Ho 2019). The successful production of the colchicine-induced tetraploid cultivar ‘Etteumbaek’ of P. grandiflorum significantly contributes to the increasing demands of this valuable plant (Kim et al. 2003; Wu et al. 2011; Kwon et al. 2016). This tetraploid plant shows no significant differences in growth characteristics compared to the diploid cultivar, except on leaf morphology (Kim et al. 2003; Wu et al. 2011). The increased leaf length and width as well as stomatal size observed on this cultivar correlate significantly to its photosynthetic rate as well as to the increased fresh weight. These warrant medicinal plant researchers to conduct further analyses for its functional compounds (Wu et al. 2011; Kwon et al. 2016). Several studies revealed that tetraploid cultivar showed an increase in cytotoxicity, antimicrobial, antioxidant enzyme activity (Boo et al. 2013; Boo et al. 2015), biomass production and other effective medicinal compounds compared to the diploid cultivar (Kim et al. 2003; Wu et al. 2011).

In this present study, FISH karyotype between two P. grandiflorus cultivars, diploid ‘Jangbaek’ and its colchicine-induced tetraploid ‘Etteumbaek’ were compared and analyzed. The somatic chromosome composition of the diploid ‘Jangbaek’ (2n = 2x = 18 ) and tetraploid ‘Etteumbaek’ (2n = 4x = 36) agreed with the previously published reports (Kim et al. 2003; Wu et al. 2011; Yang et al. 2015; Pellerin et al. 2017). However, a respective karyotype formula of 2n = 2x = 18 = 16m + 2sm (2 satellites) and 2n = 4x = 36 = 24m + 12sm (4 satellites) in diploid and colchicine-induced cultivar had slight differences on the centromeric position of the chromosomes as compared to other studies (Yang et al. 2015; Pellerin et al. 2017). Yang et al. (2015) observed that there are 24 metacentric and 12 submetacentric chromosomes in the tetraploid cultivar. Although the number of metaand submetacentric chromosomes was similar to our present result, the position of these chromosomes was different in the karyogram. Their study also showed that there are 12 metacentric and 6 submetacentric chromosomes in the diploid cultivar whereas Pellerin et al. (2015) reported the karyotype formula of the other seven diploid cultivars having 14 metacentric, 2 submetacentric and 2 subtelocentric chromosomes. These differences could be due to genetic recombination and genome rearrangement during cultivar development (Koo et al. 2002; Yonemori et al. 2008) or during the pretreatment of antimitotic agents before chromosome spread preparation wherein arrested chromosomes were at a different metaphase stage. Somehow, pretreatment affects chromosome length and condensation causing a shift of karyotype characteristics (Mártonfiová 2013).

To investigate chromosome characterization, cytogenetic markers such as ribosomal DNA (rDNA) tandem repeats could provide information about genome organization and allow the identification of homologous chromosome pairs (Heslop-Harrison 2000). In this study, we used the highly conserved coding region of both the non-NOR-forming 5S rDNA and the nucleolar organizing region (NOR) of 45S rDNA to characterize chromosome composition (Pellerin et al. 2017). Both rDNA signals were detected on different chromosomes of the two cultivars. The 5S rDNA signals were located at the interstitial or pericentromeric regions of the chromosomes which is the common pattern in most plant species (Castilho and Heslop-Harrison 1995), whereas the 45S rDNA signals were seen on the NORs or the secondary constriction of the satellite chromosomes, which is common in identifying chromosomes (Galián et al. 2012). The co-localization of 5S and 45S rDNA signals detected in the same chromosome of the diploid and tetraploid cultivars in this study is rarely observed in some plant species (El-Twab and Kondo 2012). However, this colocalization revealed the complete doubling of chromosome occurred during colchicine treatment (Bergeron and Drouin 2008). Although 45S rDNA signals indicate that there was a doubling of chromosomes, FISH result on 5S rDNA probe showed some vague distribution of signal on different chromosomes of the tetraploid cultivars. This signal may be due to repeat redistribution or chromosome rearrangement (Garcia et al. 2009; Wicke et al. 2011). Nevertheless, the fact that the complete set of 5S rDNA signals was not located in the same chromosome (2n = 4x) implies the idea of sequence recombination, genomic instability and chromosomal aberrations of this tetraploid cultivar that possibly occurred during the culture process (Rogers and Bendich 1987; Paun et al. 2007). This problem has to be cleared in future investigations using additional repeats and species-specific markers to further elucidate the aberrations initially observed.

In conclusion, FISH karyotype analysis using 5S and 45S rDNA probes revealed the occurrence of the doubling of chromosomes in the tetraploid cultivars of P. grandiflorus induced by colchicine treatment. There should be more cytogenetic markers such as minor and major repeats, and transposable elements to clearly investigate the chromosomal and genomic diversity of this species. However, the result of this study will provide a chromosomal backbone for the comprehensive genome analysis, which is useful for designing programs to breed and improve this species.