Introduction
The genus Rosa belongs to the family Rosaceae. It comprises four subgenera including Hulthemia, Eurosa (Rosa), Platyrhodon, and Hesperrhodos. Subgenus Rosa is divided into 11 sections: Banksianae, Bracteatae, Caninae, Carolinae, Cinnamomeae, Chinensis, Gallicanae, Gymnocarpae, Laevigatae, Pimpinellifoliae, and Synstylae. The genus contains approximately 150 species and is genetically related to apple, pear, plum, cherry, and strawberry (Shulaev et al. 2008; Quest-Ritson and Quest-Ritson 2003). There are 12 wild rose species including Rosa multiflora, R . rugose, and R. wichuraiana etc., which are distributed in Korea (Kim 1996; Lee 1996). From the 19th century, when rose breeding began, more than 20,000 cultivars have arisen. Among the 150 wild rose species, only 8 to 11 wild species have used in the breeding of modern cultivated roses (Spethmann and Feuerhahn 2003; Zlesak 2006).
Roses are most economical horticultural crops in floricultural industry globally as cut flowers, pot and garden plants. Korea joined the International Union for the Protection of New Varieties of Plant (UPOV) from 2001. The UPOV has designated roses as a protected plant variety. Because domestic rose farms depend on foreign varieties, royalty issues have emerged. Since 1990, governmental institutions in Korea, such as National Institute of Horticultural and Herbal Science and Agricultural Research and Extension Services, have produced approximately 220 rose varieties (Ko et al. 2014; Lee et al. 2016).
Ploidy levels of rose range from diploid (2n = 2x = 14) to octoploid (2n = 8x = 56) with a basic chromosome number of n = 7 (Darlington and Wylie 1955). Recently, R. praelucens Byhouwer, an endemic species to Yunnan, China, was reported as a decaploid (2n = 10x = 70) (Jian et al. 2010a). Cytogenetic studies in rose have reported focusing on wild species to confirm the ploidy level by means of traditional staining method (Jian et al. 2010a, b; Liu and Li 1985; Ma and Chen 1991, 1992; Price et al. 1981). Although information of chromosome number in wild rose species has been reported, critical karyological analyses like identification characterization of each chromosome and molecular cytogenetic approaches have been less extensive (Akasaka et al. 2002, 2003; Fernández-Romero et al. 2001; Ma et al. 1997; Macphail and Kevan 2009). Commercial rose cultivars (Rosa × hybrid) are commonly either triploid (2n = 3x = 21) or tetraploid (2n = 4x = 28) (Rout et al. 1999). However, the exact ploidy level in commercial rose cultivars is rarely reported (Hwang et al. 2012). In particular, estimation of ploidy level of rose species or cultivar is basic information of breeding program. Ploidy can affect pollen viability and cross affinity (Hwang et al. 2010).
Fluorescence in situ hybridization (FISH) is a powerful tool to characterize individual chromosomes (Maluszynska and Heslop-Harrison 1993), explore genomic relationships among species (Kikuchi et al. 2008), localize repetitive sequences (Abd El-Twab and Kondo 2012; Hwang et al. 2015; Martınez et al. 2010), and provide information on genome organization (Agrawal et al. 2013). Ribosomal DNAs are organized into two multigene families, 5S and 45S rDNA, which are organized in abundant repeated sequences (hundred to thousand copies). These two types of ribosomal DNAs have been used as chromosome-specific markers for FISH karyotyping in many floricultural crops (Cuyacot et al. 2016; Hwang et al. 2015; Lee et al. 2014).
In the present study, we sought to confirm the number and positions of 5S and 45S rDNA sites in five wild rose species. Understanding the ploidy level and detailed chromosome information in rose species is important to improve breeding programs and further taxonomic studies.
Materials and methods
Plant materials and chromosome preparation
R. helenae, R. mulliganii, R. multiflora, R. rubus, and R. solieana were kindly supplied by Chonnam National University, Gwangju, Korea. All were maintained in the greenhouse at Kyungpook National University. Young root tips (2 - 3 cm in length) were collected and pretreated in 2 mM 8-hydroxyquinoline solution for 4 hr at 20oC, then fixed with Carnoy’s solution (glacial acetic acid : absolutealcohol = 1 : 3, v/v) for at least 2 hr. The fixed root tips digested with the enzymatic solution consisted of 0.3% cellulase RS (Yakult Pharmaceuticals, Japan), 0.3% pectolyase Y23 (Yakult Pharmaceuticals, Japan), and 0.3% cytohelicase (Sigma, Germany) in 150 mM citric buffer for 1 hr at 37oC. Treated root tips were placed on slide glass and squashed with 60% acetic acid.
Probe labeling and FISH
The 45S rDNA containing 9 kb of rDNA genes from wheat (Gerlach and Dyer 1980) was labeled with biotin-16-dUTP (Roche, Germany). The 5S rDNA was amplified from rose genomic DNA using the forward primer 5’-GAG AGT AGT ACT AGG ATG GGT GAC C-3’ and reverse primer 5’-CTC TCG CCC AAG AAC GCT TAA CTG C-3’ (Akasaka et al. 2002) and labeled with Digoxygenin-11-dUTP (Roche, Germany). FISH was performed as described by Lim et al. (2001). Briefly, hybridization mixture was composed of 50% deinonizd formamide, 10% dextran sulphate, 2x SSC, 0.25% sodium dodecyl sulphate, 2 ng·μL-1 of each probe, and 100 ng·μL-1 of block DNA. The hybridization mixture and slide were denatured at 70oC for 10 min and 80oC for 5 min, respectively. After hybridization at 37oC for 10 min overnight, the slides were washed and dried. Digoxigenin-11-dUTP or biotin-16-dUTP were detected with anti-digoxigenin-FITC or streptavidin Cy3, respectively. Chromosomes were counterstained with 2 ng·μL-1 of 4’,6’-diamidino-2-phenylindole (DAPI) and mounted with Vetashield (Vecta Laboratories Inc., USA).
Observation and karyotyping
Somatic metaphase chromosomes were observed using a BX53 fluorecence microscope (Olympus, Japan) equipped with a CoolSNAPTM CCD camera (Photometrics, USA). Each short arm and long arm length were measured using GenusTM version 3.1 program (Applied Imaging, USA) and Adobe Photoshop CC. According to the criteria proposed by Levan et al. (1964), chromosomes were discriminated one of four types which are metacentric (m), submetacentric (sm), subtelocentric (st), or telocentric (t). Chromosomes were arranged according to their decreasing total chromosome length.
Results and Discussion
On the basis of chromosome morphological characters, such as chromosome length and chromosome type, an overview of morphological data is shown in Table 1. Among the seven pairs of chromosomes, chromosome #7 possessed nucleolar organizing regions (NORs). This chromosome was the shortest in all species.
Chromosome characteristics
The total chromosome length varied from 1.29 ± 0.01 to 2.05 ± 0.04 μm with total length of 12.20 μm in R. helenae, 3.32 ± 0.01 to 6.82 ± 0.71 μm with total length of 35.56 μm in R. mulliganii, 1.58 ± 0.03 to 2.24 ± 0.04 μm with total length of 12.94 μm in R. multiflora, 2.05 ± 0.14 to 3.46 ± 0.05 μm with total length of 19.81 μm in R. rubus, and 1.62 ± 0.28 to 2.46 ± 0.01 μm with total length of 14.83 μm in R. solieana (Table 1). The karyotype formula was 5m + 2sm in R. helenae, R. mulliganii, R. solieana; 6m + 1sm in R. rubus; and 7m in R. multiflora. Previous studies reported that R. multiflora (Chen et al. 2003) and R. rubus (Jian et al. 2013) consist of 6m + 1sm and R. mulliganii consists of 7m (Ma et al. 1997). Until now, no karyo-morphological information has been available for R. helena and R. solieana (Jian et al. 2013). The cytological studies in many plants have provided meaningful information concerning chromosome number in a cell, chromosome length, karyotypic formula, and presence of satellite. The chromosome length has been standardized by several measurement methods, such as short arm or long arm chromosome length, total chromosome / complement length (the sum of all chromosome length in one nucleus), and relative length of individual chromosomes (chromosome length of the individual chromosome / total chromosome length x 100) (Mártonfiová 2013). Most of the rose chromosome length is characterized by relative length (Akasaka et al. 2002, 2003), but few reports have calculated rose chromosome length using fine metaphase stage (Ma et al. 1997; Hwang et al. 2012). Ma et al. (1997) reported that handling of rose chromosome is difficult, because of the small chromosome, low seed germination, and weak development of roots. In addition, because of difficulty of accumulation of exact metaphase stages in woody plant cell, it is difficult to analyze the sufficient spread using karyotype. Prior reports used prometaphase stage for the chromosome studies in rose species (Akasaka et al. 2003; Jian et al. 2013; Ma et al. 1991, 1992). Using a well-condensed chromosome is key for the exact measurement of chromosome length and karyotypic formula. According to Ma et al. (1997), uniform size of mean chromosome length in genus rose ranges from 1.9 (shortest) to 2.7 μm (longest). Hwang et al. (2012) also reported that chromosome length in seven cultivars ranged from 1.5 (shortest) to 2.2 μm (longest). Our results showed that, except for R. mullganii, the mean chromosome length ranged from 1.6 to 2.6 μm, which is within the range of previous studies (Table 1 and Fig. 1). As shown in the Table 1, the chromosome sizes of R. mullganii were longer than other species because the prometaphase stage was used.
FISH analysis
Because of the similar size of the chromosome length and similar morphologies, it is difficult to identify individual and match each homologous. 5S and 45S rDNA are valuable markers to discriminate each homologous chromosome. Fig. 1 shows the FISH image of metaphase chromosome in the five wild rose species. One pair of nucleolar organizing region (NOR) bearing chromosome were measured as a chromosome #7, which was the shortest among the seven pairs in five rose species. Dual-color FISH analysis with 5S (green fluorescence) and 45S (red fluorescence) rDNAs revealed that one pair of signal co-localized on the chromosome #7 in all five roses species (Figs. 1, 2, and 3). The 5S rDNA located at proximal region of the long arm of chromosome #7. One pair of 45S rDNA signal was observed at NOR located at the terminal region of the short arm of chromosome #7. Previous studies reported that a pair of 45S rDNA detected in R. chinensis, R. odorata, R. laeviata, R. roxburghii, R. multiflora, R. rugosa, R. moschata, R. gigantea, R. sempervirens (Feranández-Romero et al. 2001; Ma et al. 1997) but they could not detect simultaneously 5S rDNA loci. More than a single locus of 45S rDNA per chromosome in diploid or polyploid rose species has been reported (Akasaka et al. 2003; Kirov et al. 2016). Presently, a number of 45S rDNA loci in R. multiflora corresponded with the results of Feranández-Romero et al. (2001), but differed from the findings of Akasaka et al. (2003). The latter authors reported that the location of 5S rDNA was separate from the 45S rDNA loci. More than a single locus of 5S rDNA was detected by Kirov et al. (2016), who mentioned being unable to clearly detect 5S rDNAs on both metaphase and pachytene stages. They did not explain their result.
There are several possible explanations for the differences between our results and others. Rosa originated in East Asia and migrated to North America (Flory 1950). During this migration, chromosome rearrangement including translocation, duplication, or deletion occurred. These possibilities may have been produced distinct karyotype within diploid subgenus. As well, because of development of experimental technique and computational analysis, resolution and reliability are more sophisticated than before. Many factors influence FISH resolution. They include the condition of chromosome preparation, probes, hybridization stringency, and researcher’s skill. For example, non-specific signals are detected under low hybridization stringency. Since, for the accuracy and reliability of the results, we mainly used high stringency condition, non-specific binding probes or minor signals may have been removed. In addition, small sized chromosome hampered the construction of fine karyotype using FISH technique (Kirov et al. 2015).
We constructed fine karyotypes and characterized chromosomal morphologies of five wild rose species using FISH. Understanding of chromosomal data will contribute to improve breeding efficiency and can also help in phylogenetic and evolutionary studies. In addition, localization of 5S and 45S rDNA will facilitate physical mapping of its genome, which can be used as a basic data for further genome research.
