Abstract
Roses are one of the economically most important groups of ornamental plants. The internal transcribed spacers (ITS) of the nuclear ribosomal DNA and the chloroplast gene matK were used to investigate the genetic diversity and genetic relationships among Rosa germplasm including 39 wild species, 21 old garden roses, and 29 modern cultivars. Three dendrograms based on ITS and matK clustering data indicated that 1) 39 wild genotypes were consistent with their classification into botanical sections with only a few exceptions; 2) most of the wild genotypes were separated from rose cultivars. However, three sections, Synstylae, Chinenses, and Rosa, that contributed to the modern roses generally gathered together with almost all old garden and modern roses on the molecular level; and 3) the relationships between cultivated roses as inferred by ITS and matK sequences do not correlate with horticultural groups. Results demonstrated that both sequence techniques can contribute to clarifying the genetic relationships of rose accessions and germplasm conservation to enhance the ornamental and economic value of rose.
The genus Rosa comprises ≈200 species and 30,000 cultivated varieties, but only 10 to 15 species have contributed to the modern roses (Cairns, 2007; Chen, 2001). Compared with the abundant genetic diversity of the wild species, the genetic background of modern varieties is relative narrow (Liu and Liu, 2004; Matsumoto et al., 1998). The roses obtained before the breeding of ‘La France’ in 1867, the first Hybrid Tea rose, are generally called old garden roses (Cairns, 2003). The beauty of the old garden roses often lies in the heavy fragrance they can impart to the garden (Cairns, 2003). Nowadays, more efficient breeding strategies are needed as a result of the increasing demand for new modern roses. Many species and old garden roses have desirable traits such as superior disease resistance, higher fragrance levels, winter-hardiness, and the absence of prickles (Scariot et al., 2006; Tang, 2009). We can efficiently transfer these desirable traits from species and old garden roses into modern roses if the genetic determinants of the traits and their chromosomal location are known (Debener and Mattiesch, 1999).
The two nrDNA ITS regions (ITS1 and ITS2) among 18S, 5.8S, and 26S ribosomal RNA are suitable for studies of genetic relationships at the species and generic levels of divergence (Baldwin et al., 1995; Matsumoto et al., 2000). The matK gene in the chloroplast DNA has proved to be very useful for the analysis of large numbers of genotypes (Leus et al., 2004; Matsumoto et al., 1998).
Sequence data of Rosa nuclear, mitochondrial, and chloroplast DNA are more and more applied in the analysis of genetic relationships (Leus et al., 2004). Although some rose accessions had been researched by ITS, rbcL, matK, and other sequence data, many species and cultivars were not included in those previous analyses (Leus et al., 2004; Matsumoto et al., 1998). Moreover, there are no reports about the wild roses, old garden roses, and modern roses genome by ITS and matK sequences analyses.
In this study, 89 genotypes of genus Rosa, which included 39 wild roses (Table 1), 21 old garden roses (Table 2), and 29 modern roses (Table 3) were analyzed by ITS and matK sequences. The main aims of the study were 1) to analyze genetic relationships and distances among the three Rosa classes; and 2) to detect the role of wild roses in the history of rose breeding. The results of this study will clarify our understanding of which accessions were used in the rose breeding.
List of the wild roses sampled for internal transcribed spacer (ITS) and matK sequencing analyses.
List of the old Chinese garden roses sampled for internal transcribed spacer (ITS) and matK sequencing analyses.
List of the modern roses sampled for internal transcribed spacer (ITS) and matK sequencing analyses.
Materials and Methods
Plant materials.
A list of 39 wild roses, 21 old garden roses, and 29 modern roses were used in this study (Tables l to 3). They were observed to confirm their identity by morphological traits and literature descriptions (Cairns, 2007; Institutum Botanicum Chinenses Academiae Sinicae Edita, 1985; Institutum Botanicum Kunmingense Academiae Sinicae Edita, 2006; Ku and Robertson, 2003; Zhang and Zhu, 2006). Spiraea thunbergii was chosen as an outgroup and its GenBank sequence numbers were DQ897626 (ITS) and FM179934 (matK) that had been downloaded from GenBank. Wild roses were obtained in the wild in spring. Old garden roses and modern roses were collected at the Institute for Flower Breeding in Kunming.
DNA isolation, PCR amplification, and sequencing.
Fresh leaf tissue of each accession was used to extract genomic DNA using the CTAB method as described by Doyle and Doyle (1987). For amplification of the entire ITS1-5.8S-ITS2 region or matK region, polymerase chain reaction (PCR) was performed in a 25-μL reaction mixture consisting of 15 mm Tris-HCI (pH 8.0), 75 mm KCl, 1.5 mm MgCl2, 200 μM dNTP, 1.5 U Taq-DNA polymerase, 100 ng template DNA, 1.0 μM of ITS5 (5′-GGAAGTAAAAGTCGTAACAAGG-3′), and ITS4 (5′-TCCTCCGCTTATTGATATGC-3′) primers (White et al., 1990) or trnK-3914F (5′-GGGGTTGCTAACTCAACGG-3′) and tranK-2R (5′-AACTAGTCGGATGGAGTAG-3′) primers (Matsumoto et al., 1998). Amplification conditions for ITS were an initial 5 min at 95 °C, followed by 30 cycles of 45 s at 95 °C, 1 min at 55 °C, 45 s at 72 °C, and 10 min at 72 °C for final extension. Amplification cycles for matK consisted of an initial step of 5 min at 95 °C followed by 25 cycles of 1 min at 94 °C, 2 min at 48 °C, 3 min at 72 °C, and 15 min at 72 °C for final extension. Amplifications were performed in a Biometra (UNO-Thermoblock) thermocycler. The amplified products were purified with the QIAquick PCR purification kit (CN 28104; QIAGEN) and then they were sent to the Invitrogen Company (Shanghai, China) for sequencing. Moreover, the unresolved sequences were cloned with the TA cloning kit (Invitrogen Co., Ltd.). The sequencing primers for ITS were the same ones used for amplification. The sequencing primers matK-WF (5′-CTTTGCATTTATTACGGCTC-3′) and matK-SR (5′-GATTGGTTACGGGAGAAAAAG-3′) were used for matK (Matsumoto et al., 1998).
Data analysis.
The sequences were aligned with the computer program Clustal X (Thompson et al., 1997) and checked manually in BioEdit (Hall, 1999). Insertions or deletions were included and gaps were treated as missing data in the analysis. The maximum parsimony method in the computer software PAUP Version 4.0 (Swofford, 2000) was performed by the HEURISTIC search option to find the most parsimonious trees. In the analysis, character state changes were weighted equally. For bootstrap analyses, 1000 replications were conducted for each branch (Felsenstein, 1985).
Results
ITS phylogeny.
In 89 accessions, the length of ITS 1 (from 242 to 259 bp) was longer than that of ITS 2 (from 197 to 217 bp). The sizes of ITS1 and ITS2 were within the range of those reported previously for other angiosperms and are similar to those reported for Rosa (Baldwin et al., 1995; Campbell et al., 1995; Matsumoto et al., 2000). The 5.8S coding region was from 163 to 165 bp long in all accessions examined.
The phylogenetic analyses with a total of 90 accessions, including one outgroup taxa and 89 ingroup taxa, were conducted. A total of 642 manually aligned nucleotides were used for phylogenetic analyses. We found 367 constant characters (55.0%), 187 variable parsimony uninformative characters (28.0%), and 129 parsimony informative characters (19.3%) between the ingroup and outgroup. The heuristic search resulted in more than 10,000 trees with a tree length (TL) of 667, a consistency index (CI) 0.8152, and a retention index (RI) of 0.7135. The strict consensus tree with bootstrap values was constructed from the sequence data (Fig. 1). Several different alignment parameters (i.e., gap opening and extension penalties) using the Clustal X program (Thompson et al., 1997) resulted in finding the same major lineages.
Phylogenetic tree of 89 accessions based on internal transcribed spacer (ITS) data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Phylogenetic tree of 89 accessions based on internal transcribed spacer (ITS) data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Phylogenetic tree of 89 accessions based on internal transcribed spacer (ITS) data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Eighty-nine accessions were clustered into seven major groups (Fig. 1). All the members in the section Pimpinellifoliae were included in Group A. Group B contained almost all the old garden roses, modern roses, and all the accessions of the sections Synstylae, Chinenses, Rosa, and Microphyllae (except for R. praelucens). Group C constituted only one species, belonging to the section Laevigatae. Group D included the rest of five cultivars. Group E consisted of all the accessions of the section Cinnamomeae plus R. praelucens, which belongs to the section Microphyllae. Group F contained one accession of the section Bracteatae. All three genotypes of section Banksianae were included in Group G.
matK phylogeny.
In all accessions, the length of matK sequence ranged from 1401 to 1619 bp. A total of 1636 aligned nucleotides were used for phylogenetic analyses. There are 1362 constant characters (83.2%), 229 variable but parsimony uninformative characters (14.0%), and 45 parsimony informative characters between the ingroup and outgroup (2.8%) in the data matrix. The default heuristic search found more than 50,000 trees with a TL of 327, a CI of 0.8991, and a RI of 0.8472. A single most parsimonious tree with bootstrap values was constructed from the data (Fig. 2). Phylogenetic analyses using character weighting and character state weighting found the same strict consensus trees (not shown) as the unweighted default search strict consensus tree (Fig. 2). Therefore, our discussion is based on this strict consensus tree.
Phylogenetic tree of 89 accessions based on matK data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Phylogenetic tree of 89 accessions based on matK data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Phylogenetic tree of 89 accessions based on matK data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
They were divided into six major groups (Fig. 2). Group A contained all the accessions of the section Pimpinellifoliae. Group B included two accessions of the section Banksianae and two species belonging to Microphyllae. Group C constituted one species of the section Bracteatae. All the species of the section Cinnamomeae and R. praelucens were included in Group D. This cluster result is the same with the previously described ITS result. Group E contained only one species, belonging to Laevigatae. Group F consisted of all the cultivars from old garden roses and modern roses and all the accessions of sections Synstylae, Chinenses, and Rosa.
Combined ITS and matK phylogeny.
There are several clades with weak bootstrap supports (less than 50%) in these two trees, but also several incongruences between ITS and matK phylogenic. So we decided to combine the two data sets to improve resolution. The same heuristic search with 90 taxa sequenced with both ITS and the matK gene found more than 50,000 equally parsimonious trees with a TL of 1064, a CI of 0.8743, and a RI of 0.7728. The strict consensus tree is shown in Figure 3.
Phylogenetic tree of 89 accessions based on combined internal transcribed spacer (ITS) and matK data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Phylogenetic tree of 89 accessions based on combined internal transcribed spacer (ITS) and matK data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Phylogenetic tree of 89 accessions based on combined internal transcribed spacer (ITS) and matK data using maximum parsimony method. The bootstrap confidence values (%) are indicated on the branches.
Citation: HortScience horts 48, 12; 10.21273/HORTSCI.48.12.1445
Eighty-nine accessions were clustered into six major groups (Fig. 3). All the section Pimpinellifoliae accessions were included in Group A. All the species of the Cinnamomeae and R. praelucens from the Microphyllae were included in Group B. Cluster C constituted all the cultivars of old garden roses and modern roses and almost all the accessions of sections Synstylae, Chinenses, Rosa, and Microphyllae. Two small Groups D and E were set apart and were constituted by R. laevigata and R. bracteata, belonging to sections Laevigatae and Bracteatae, respectively. All three species of section Banksianae were included in Group F.
Discussion
Central to the issue of whether nrDNA and matK are good phylogenetic data are the consideration of their mode and rate of evolution. Empirical studies on gymnosperm showed the existence of serious problems using ITS and matK to reconstruct known species relationships because of their faster evolutionary rates (Karvonen, 1995; Wang and Hong, 1997). On the other hand, studies on most angiosperm, especially on closely related populations, were more successful (Ohsako and Ohnishi, 2000; Stanford et al., 2000; Xiang et al., 1998).
In this study, the independent and combined clustering results for the matK and ITS sequences of wild genotypes were substantially consistent with their classification into botanical sections with only a few exceptions. The morphological similarity between sections Synstylae and Chinenses (Ku and Robertson, 2003; Matsumoto et al., 1998) was confirmed on the molecular level by this study (Figs. 1 to 3). R. praelucens from the section Microphyllae has many morphological similarities with the section Cinnamomeae (Institutum Botanicum Chinenses Academiae Sinicae Edita, 1985), and here it was also clustered into one group with other accessions of Cinnamomeae (Figs. 1 to 3). The relationships between cultivated roses as inferred by ITS and matK sequences do not correlate with the classical rose classification system (Cairns, 2003). Moreover, the chromosome ploidy level of Rosa does not correlate with the genetic relationships in our cluster study.
The narrow genetic background of cultivars makes the gene pool used in breeding programs very small; therefore, it is not surprising to find a clear distinction between the group of rose cultivars and the group of species on the molecular level with the exception of three sections. The three sections Synstylae, Chinenses, and Rosa out of nine original sections that contributed to the modern roses were found to gather in one group with almost all ancient and modern cultivars (Figs. 1 to 3). This result was consistent with the viewpoints of others (Chen, 2001; Koopman et al., 2008; Liu and Liu, 2004; Matsumoto et al., 1998; Scariot et al., 2006; Tang, 2009). The distance between cultivars and species of other six sections can be explained by the fact that these sections contributed very little to the breeding of modern rose varieties in the last two centuries (Chen, 2001; Koopman et al., 2008; Krüssmann, 1981).
Moreover, many species from nine sections and old garden roses have many good characteristics, e.g., disease resistance, cold resistance, fragrance, and lack of pricks (Liu and Liu, 2004; Qiu et al., 2009). Information about genetic relationships could be used in breeding programs to introduce these horticultural important characteristics into cultivated roses from wild or ancient germplasm starting with accessions from the most closely related taxons.
Literature Cited
Baldwin, B.G., Sanderson, M.J., Porter, J.M., Wojciechowski, M.F., Campbell, C.S. & Donoghue, M.J. 1995 The ITS region of nuclear ribosomal DNA: A valuable source of evidence on angiosperm phylogeny Ann. Mo. Bot. Gard. 82 247 277
Cairns, T. 2003 Horticultural classification schemes, p. 117–124. In: Roberts, A., Debener, T. and Gudin, S. (eds.). Encyclopedia of rose sciences. Elsevier Science, Oxford, UK
Cairns, T. 2007 Modern roses XII. Academic Press, New York, NY
Campbell, C.S., Donoghue, M.J., Baldwin, B.G. & Wojciechowski, M.F. 1995 Phylogenetic relationships in Maloideae (Rosaceae): Evidence from sequences of the internal transcribed spacers of nuclear ribosomal DNA and its congruence with morphology Amer. J. Bot. 82 902 918
Chen, J.Y. 2001 Taxonomy of flower cultivars of China. China Forestry Publishing House, Beijing, China [in Chinese]
Debener, T. & Mattiesch, L. 1999 Construction of a genetic linkage map for roses using RAPD and AFLP markers Theor. Appl. Genet. 99 891 899
Doyle, J.J. & Doyle, J.L. 1987 A rapid isolation procedure for small quantities of fresh leaf tissue Photochemical Bull 19 11 15
Felsenstein, J. 1985 Confidence limits on phylogenies: An approach using the bootstrap Evolution 39 783 791
Hall, T.A. 1999 BioEdit: A user friendly biological sequence alignment editor and analysis program for Windows 95/98/NT Nucleic Acids Symposium Series 41 95 98
Institutum Botanicum Chinenses Academiae Sinicae Edita 1985 Flora of China, vol. 37. Science Press, Beijing, China [in Chinese]
Institutum Botanicum Kunmingense Academiae Sinicae Edita 2006 Flora of Yunnanica, vol. 12. Science Press, Beijing, China [in Chinese]
Jian, H.Y., Zhang, T., Wang, Q.G., Li, S.B., Zhang, H. & Tang, K.X. 2013 Karyological diversity of wild Rosa in Yunnan, southwestern China Genet. Resources Crop Evol. 60 115 127
Karvonen, P. 1995 Genetic variation and structure of ribosomal DNA (rDNA) in Scots pine and Norway spruce Acta University Ovluensis Series A Scientiae Rerum Naturalium 10 1 68
Koopman, W.J.M., Wissemann, V., Cock, K.D., Huylenbroeck, J.V., Riek, J.D., Sabatino, G.J.H., Visser, D., Vosman, B., Ritz, C.M., Maes, B., Werlemark, G., Nybom, H., Debener, T., Linde, M. & Smulders, M.J.M. 2008 AFLP markers as a tool to reconstruct complex relationships: A case study in Rosa (Rosaceae) Amer. J. Bot. 95 353 366
Krüssmann, G. 1981 The complete book of roses. Timber Press, Portland, OR
Ku, T.C. & Robertson, K.R. 2003 Rosa (Rosaceae), p. 339–381. In: Wu, Z.Y. and P.H. Raven (eds.). Flora of China, vol. 9. Science Press, Beijing, China
Leus, L., Jeanneteau, F., Van, H.J., Van, B.E. & De, R.J. 2004 Molecular evaluation of a collection of rose species and cultivars by AFLP, ITS, rbcL and matK Acta Hort. 651 141 147
Liu, Y.G. & Liu, Q.L. 2004 Evaluation and exploitation of genetic resources in roses Journal of Plant Genetic Resources 5 87 90 [in Chinese with English summary]
Matsumoto, S., Kouchi, M. & Fukui, H. 2000 Phylogenetic analyses of the subgenus Eurosa using the ITS nrDNA sequence Acta Hort. 521 193 202
Matsumoto, S., Kouchi, M., Yabuki, J., Kusunoki, M., Ueda, Y. & Fukui, H. 1998 Phylogenetic analyses of the genus Rosa using the matK sequence: Molecular evidence for the narrow genetic background of modern roses Sci. Hort. 77 73 82
Ohsako, T. & Ohnishi, O. 2000 Intra and interspecific phylogeny of wild Fagopyrum (Polygonaceae) species based on nucleotide sequences of noncoding regions in chloroplast DNA Amer. J. Bot. 87 573 582
Qiu, X.Q., Zhang, H., Li, S.F., Jian, H.Y. & Tang, K.X. 2009 The relative relationships analysis of rose germplasm in Yunnan based on SSR Acta Botanica Boreali-occidentalia Sinica 29 1764 1771
Scariot, V., Akkak, A. & Botta, R. 2006 Characterization and genetic relationships of wild species and old garden roses based on microsatellite analysis Amer. Soc. Hort. Sci. 131 66 73
Stanford, A.M., Harden, R. & Parks, C.R. 2000 Phylogeny and biogeography of Jugland (Juglandaceae) based on matK and ITS sequence data Amer. J. Bot. 87 872 882
Swofford, D.L. 2000 PAUP*4.0b10. Phylogenetic analysis using parsimony (*and other methods), Version 4. Sinauer Associates, Sunderland, MA
Tang, K.X. 2009 Study on germplasm resources of Rosa L. in Yunnan. PhD diss., Yunnan University, Kunming, China [in Chinese with English summary]
Thompson, J.D., Gibson, T.J., Plewniak, F., Jeanmougin, F. & Higgins, D.G. 1997 The clustal_X windows interface: Flexible strategies for multiple sequence alignment aided by quality analysis tools Nucleic Acids Res. 25 4876 4882
Wang, W.Q. & Hong, D.Y. 1997 Progress in molecular systematics of plants in recent five years Acta Phytoecol. Sin. 35 465 480
White, T.J., Bruns, T., Lee, S. & Tayler, J. 1990 Amplification and direct sequencing of fungal ribosomal RNA genes for phylogenetics, p. 315–322. In: PCR protocols: A guide to methods and applications. Academic Press, San Diego, CA
Xiang, Q.Y., Crawford, D.J. & Wolfe, A.D. 1998 Origin and biogeography of Aesculus L. (Hippocastanaceae): A molecular phylogenetic perspective Evolution 52 988 997
Zhang, Z.S. & Zhu, X.Z. 2006 Chinese roses. China Forestry Publishing House, Beijing, China [in Chinese]
