- Email: [email protected]
Reproductive barriers in interspecific hybridizations among Chimonanthus praecox (L.) Link, C. salicifolius S. Y. Hu, and C. nitens Oliver from pollen–pistil interaction and hybrid embryo development
Scientia Horticulturae 177 (2014) 85–91
Contents lists available at ScienceDirect
Scientia Horticulturae journal homepage: www.elsevier.com/locate/scihorti
Reproductive barriers in interspecific hybridizations among Chimonanthus praecox (L.) Link, C. salicifolius S. Y. Hu, and C. nitens Oliver from pollen–pistil interaction and hybrid embryo development Wenpeng Wang a , Lihua Zhou a , Yaohui Huang a , Zhiyi Bao a , Hongbo Zhao a,b,∗ a b
Department of Ornamental Horticulture, School of Landscape Architecture, Zhejiang Agriculture and Forestry University, 311300 Lin’an Hangzhou, China Nurturing Station for State Key Laboratory of Subtropical Silviculture, 311300 Lin’an Hangzhou, China
a r t i c l e
i n f o
Article history: Received 28 April 2014 Received in revised form 24 July 2014 Accepted 28 July 2014 Available online 17 August 2014 Keywords: Chimonanthus praecox C. salicifolius C. nitens Interspecific hybridization Pollen–pistil interaction Hybrid embryo development
a b s t r a c t To investigate interspecific cross-compatibility among Chimonanthus praecox (L.) Link, C. salicifolius S. Y. Hu, and C. nitens Oliver, pollen–pistil interaction by fluorescence microscopy and hybrid embryo development by whole mount clearing technique were observed. The results indicated that for all reciprocal crosses among the three species, pollen grains of the male parents could adhere to and germinate in the stigmas of female parents. For any cross combination, pollen tube growth could reach the embryo sac, and double fertilization could be completed. These findings indicated that no obvious barriers to interspecific hybridization existed at the stages of pollen germination and pollen tube elongation until double fertilization. Although certain proportion (29.6% at 9 d and 36.3% at 15 d after pollination) of abortions (e.g., non-fertilization and retarded or non-existent endosperm development) in normally pollinated embryos of C. praecox at different stages were also observed, a greatly increased proportion (66.1% at 8 d and 76.7% at 15 d after pollination in C. praecox × C. nitens, and 60.0% at 8 d and 75.0% at 15 d after pollination in C. praecox × C. salicifolius) of hybrid embryos did not complete double fertilization or completed double fertilization but underwent abortion due to the retardation of endosperm development. Comparing to normally pollinated embryos of C. praecox, the retardation of endosperm development in hybridizations occurred earlier. Moreover, the occurrence of abortions among different cross combinations was basically consistent. Thus, the retardation of the endosperm development of hybrid embryos after double fertilization (after the globular embryo stage) is an important factor leading to the failure of distant hybridization between Chimonanthus species. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Calycanthaceae is a small and relatively primitive family that includes a total of three genera (Chimonanthus Lindley, Calycanthus L., and Sinocalycanthus Cheng et S.Y. Chang) and 7–9 species. This family has a disjunct distribution across East Asia and North America. In spite of a small number of species in the family, these species are highly diverse with respect to relevant traits, such as growth and ecological habits, flowering period, floral structure, odor, floral display, leaf structure, and leaf content, especially abundant diversity in the floral traits. Chimonanthus is endemic to China and the tertiary relict taxa, which include 6 species, mostly grow in mixed evergreen and deciduous broad-leaved forests and
∗ Corresponding author at: School of Landscape Architecture, Zhejiang Agriculture and Forestry University, 311300 Lin’an, Zhejiang, China. Tel.: +86 571 63748611. E-mail addresses: [email protected], [email protected] (H. Zhao). http://dx.doi.org/10.1016/j.scienta.2014.07.040 0304-4238/© 2014 Elsevier B.V. All rights reserved.
evergreen broad-leaved forests in warm temperate and subtropical regions (Zhang et al., 1998; Li and Li, 2000). The plants in this genus are widely distributed and are most abundant in the Yangtze River basin. Chimonanthus praecox (L.) Link is a traditionally ornamental fragrant plant that flowers in winter (from December to February) in China (Zhang et al., 1998; Kozomara et al., 2008). The flowers have an elegant aroma and flavor, with abundant variations in floral traits. C. nitens Oliver is a fragrant evergreen shrub that has leathery leaves with smooth surfaces, white tepals with a flowering period from September to November. C. salicifolius S. Y. Hu is a semi-evergreen shrub with narrow, willow-like leaves that have a rough upper surface and a lower surface covered in white powder. This plant’s flowers are similar to the flowers of C. nitens and also have a flowering period from September to November. These three species exhibit rich variations in growth habits, floral traits, flowering period, and stress resistance. In particular, the leaves of C. nitens and C. salicifolius are rich in alkaloids and flavonoids
86
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
(Xu et al., 2006; Ouyang et al., 2010; Shu et al., 2010), which not only possess extremely high medicinal value but also confer good insect resistance. The use of distant hybridization to introduce heterologous genes and thereby create novel germplasm with advantages from both parents to expand genetic pools is currently a primary method (Eeckhau et al., 2007). With respect to the distant hybridization in Calycanthaceae, Lasseigne et al. (2001), first reported obtaining intergeneric hybrids of S. chinensis × Calycanthus floridus, and Yao et al. (2007), have also reported obtaining intergeneric hybrids of these two species (S. chinensis × C. floridus). We have utilized conventional hybridization approaches to obtain C. floridus var. oblongifolius × S. chinensis hybrids, which first flowered in 2013 (unpublished work). Although these intergeneric hybrids have been obtained, the observed seed sets in hybridizations have been extremely low, indicating the presence of certain reproductive barriers in the intergeneric hybridization. Therefore, Wang et al. (2013), examined the underlying mechanisms of barriers to the hybridization of S. chinensis var. oblongifolius and C. floridus and found that reciprocal crosses involved different incompatibility mechanisms. When S. chinensis was the female parent, the main barrier to successful hybridization was the abnormal disintegration of the early hybrid embryo after fertilization. When C. floridus var. oblongifolius was the female parent, the main barriers to hybridization were due to the combination effects of high proportion of developmental abnormalities of female pistils and the abortion of early hybrid embryos. No prior studies have reported on the cross-compatibility of interspecific hybridizations within Chimonanthus. Interspecific hybridization may represent a method of successfully transferring or aggregating desirable traits to create new varieties with high ornamental value and strong stress resistance. However, in numerous tests of interspecific hybridization, conventional hybridization experiments involving the aforementioned these three species could produce the enlargement of fruit during the early post-pollination stages but could not successfully obtain hybrid seeds. Therefore, in this paper, the reproductive barrier of interspecific hybridization within Chimonanthus is investigated from the perspectives of pollen–pistil interaction (pollen germination, pollen tube elongation, and fertilization), and hybrid embryo development after pollination. Using this approach, the present study clarified the underlying mechanisms of reproductive barriers to hybridization, thereby providing a reference for future breeding efforts involving the distant hybridization of Chimonanthus. 2. Materials and methods 2.1. Materials The materials of C. praecox ‘CHg08’ (Fig. 1a), C. nitens (Fig. 1b), and C. salicifolius (Fig. 1c) were from the germplasm resource of Zhejiang Agriculture and Forestry University. C. praecox ‘CHg08’ is an early-blooming germplasm whose anthesis is from the end of October to the beginning of December. C. nitens and C. salicifolius are introduced from the field respectively in Lishui, Zhejiang and Qiyunshan, Anhui. These plants had been cultivated for above 8 years and were capable of normal flowering. Hybridization pollinations were conducted during the October–December flowering periods in 2010–2012. 2.2. Cross-pollination and sample fixation Reciprocal crosses was conducted for each pairwise combination of the three aforementioned Chimonanthus species according
to Wang et al. (2013), and a normal pollination (natural pollination) of C. praecox ‘CHg08’ using the conspecific pollen from other strains was as a control. The following specific operating procedures were employed. On a clear day, well-developed flower buds on a female parent were selected. Emasculation was performed on the relaxed flower buds by entirely removing the upcoming inner and outer tepals as well as the stamens (without damaging the stigma). Anthers just before dehiscence from the male parents were collected and dehydrated under incandescent lamp for 2 h, and then pollen were picked up and used to pollinate the stigmas of female parents. After pollination, flowers were promptly bagged and labeled. Pistils were collected at 1, 2, 6, 12, 24, 36, 48, 72, and 96 h after pollination, fixed with FAA (formalin: acetic acid: 50% alcohol, 1:1:8) in a 5 mL centrifuge tube for more than 24 h, and observed by fluorescence microscopy. In addition, to observe hybrid embryo development, for 3–10 d after pollination, sampling was performed once per day; for 10–20 d after pollination, sampling was performed at 11 d, 13 d, 15 d, 17 d and 20 d; beyond 20 d after pollination, sampling was performed every 5 d until shedding occurred. 2.3. Fluorescence microscopy A slightly modified version of the fluorescence microscopy method described by Zhou et al. (2006), was utilized in this study. The developing fruits were gently cut with a razor blade, and the whole pistils including stigmas and ovaries were separated from fruits and collected with tweezers. Each sample was softened and bleached with 2 mol L−1 NaOH for 2 h and then soaked in 0.1% aniline blue staining solution for 10 h (overnight). Samples were mounted in glycerol, and sections were prepared. The adhesion to and germination of the pollen on the stigma and pollen tube elongation was observed under the fluorescence microscope (Leica DM4000). 2.4. The observation of embryo development using the whole mount clearing technique The whole mount clearing technique allows for the rapid observation of internal organization and structural development within certain materials that are small in size, difficult to prepare by section, or inefficient to observe by section (Jana et al., 2006; Hao and Qiang, 2007). The following specific procedures were employed in this process. (1) Samples (developing fruits) were fixed for more than 24 h was gently cut with a razor blade, and the pistils were separated with tweezers and placed into a centrifuge tube; (2) the pistils were successively treated with a 75–85–95–100% I–100% II ethanol gradient for 2 h per ethanol solution, (3) were treated with an ethanol and methyl salicylate mixture (1:1 by volume) for 3 h, (4) and were cleared three times with methyl salicylate for 12 h for each of the first two clearing treatments and for 24 h for the final clearing treatment. (5) Then samples were placed on a concave glass slide and observed and photographed under the differential interference microscope (Nikon Eclipse 80i). 3. Results 3.1. Pollen germination and pollen tube elongation of hybridization between C. praecox and C. salicifolius In C. praecox × C. salicifolius, pollen grains adhered to the stigma by 6 h after pollination (Fig. 2a); pollen germination had begun at 12 h after pollination (Fig. 2b); the pollen tube was continuing to grow downward at 24 h after pollination (Fig. 2c); and the pollen tube had reached the embryo sac at 48 h after pollination (Fig. 2d). When C. salicifolius was the female parent, observations again
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
87
Fig. 1. Flower morphology of C. praecox (a), C. nitens (b) and C. salicifolius (c).
indicated that the pollen tube reached the embryo sac by 48 h after pollination (Fig. 2e).
and the pollen tube had approached the embryo sac by no later than 48 h after pollination (Fig. 4f).
3.2. Pollen germination and pollen tube elongation of hybridization between C. praecox and C. nitens
3.4. Normal embryo development and abortion in C. praecox after natural pollination
In C. praecox × C. nitens, pollen germination had begun at 12 h after pollination (Fig. 3a); the pollen tube had grown to reach the middle of the style by 24 h after pollination (Fig. 3b); the pollen tubes in certain samples had reached the embryo sac by 48 h after pollination (Fig. 3c); and in most observations of styles, pollen tubes had grown to reach the embryo sac by 72 h after pollination (Fig. 3d). When C. nitens was the female parent, observations again indicated that the pollen tube had reached the embryo sac by 72 h after pollination (Fig. 3e).
The whole clearing technique was applied to observe the embryo development of C. praecox under natural pollination (xenogamy with conspecific pollen). The results indicated that eggs and central cells were observed at 6 d after pollination (Fig. 5a), and that double fertilization in embryo sac was observed at 9 d after pollination (Fig. 5b). At this time, the egg was located at the end of the micropyle, next to two synergids; the central cell was in the middle, with three antipodal cells above the central cell. Globular embryos were observed at 20 d after pollination (Fig. 5c); at this time, endosperm cells had formed. The longitudinal division of the central endosperm cells had formed two strips of endosperm cells. Eventually, endosperm cells filled the entire long, narrow embryo sac cavity (Fig. 5e); during this process, nucellar tissue was gradually digested and absorbed. Heart-shaped embryos appeared at 25 d after pollination (Fig. 5d), and mature embryos were observed at approximately 35 d after pollination (Fig. 5f). Under natural pollination for C. praecox, a certain percentage of embryo abortions were observed at various time points after pollination (Table 1). At 9 d after pollination, double fertilization was not observed in approximately 29.6% of the sampled materials (Fig. 6a). At 15 d after pollination, 36.3% of fertilized eggs in the embryo sac
3.3. Pollen germination and pollen tube elongation of hybridization between C. salicifolius and C. nitens When C. salicifolius was the female parent, pollen had begun to germinate by 12 h after pollination (Fig. 4a); pollen tube growth was continuing at 48 h after pollination (Fig. 4b); and the pollen tube had grown sufficiently to approach the embryo sac by 72 h after pollination (Fig. 4c). When C. nitens was the female parent, the similar results was observed; in particular, a pollen tube was clearly visible at 24 h (Fig. 4d) and 36 h (Fig. 4e) after pollination,
Fig. 2. Growth of pollen tubes in styles after pollination between C. praecox and C. salicifolius. (a)–(d) Pollen germination and pollen tube growth at 6, 12, 24 and 48 h, respectively, after pollination for the C. praecox × C. salicifolius cross. e Pollen tube growth at 48 h after pollination for the C. salicifolius × C. praecox cross. Bar = 100 m.
88
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
Fig. 3. Growth of pollen tubes in styles after pollination between C. praecox and C. nitens. (a)–(d) Pollen germination and pollen tube growth at 12, 24, 48 and 72 h, respectively, after pollination for the C. praecox × C. nitens cross. e Pollen tube growth at 72 h after pollination for the C. nitens × C. praecox cross. Bar = 100 m.
Fig. 4. Growth of pollen tubes in styles after pollination between C. salicifolius and C. nitens. (a)–(c) Pollen germination and pollen tube growth at 12, 48, and 72 h, respectively, after pollination for the C. salicifolius × C. nitens cross. (d)–(f) Pollen tube growth at 24, 36, and 48 h, respectively, after pollination for the C. nitens × C. salicifolius cross. Bar = 100 m.
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
89
Fig. 5. Embryogenesis of C. praecox at different stages after natural pollination. (a) Unfertilized embryo sac (at 6 d after pollination). (b) Double fertilization (at 9 d after pollination, with synergids and antipodal cells visible within the embryo sac). (c) Globular embryo (at 20 d after pollination). (d) Heart-shaped embryo (at 25 d after pollination). (e) The formation of endosperm cells in a long and narrow shape (at 25 d after pollination). (f) Mature embryo (at 35 d after pollination). cc central cell, ec egg cell, ant antipodal cell, sn sperm nucleus, syn synergid. Bar = 50 m. Table 1 The number of normal and aborted embryos in C. praecox at different stages after natural pollination. Days after pollination
No. of embryos
No. of normal embryos
Percentage of normal embryo (%)
No. of aborted embryos
Percentage of aborted embryo (%)
5 9 15 20 25
43 27 33 25 36
36 18 14 8 8
83.7 66.7 42.4 32.0 22.2
4 8 12 12 18
9.0 29.6 36.3 48.0 50.0
cavity stopped developing without undergoing divisions (Fig. 6b). At 25 d after pollination, approximately 50.0% of fertilized eggs in the embryo sac had divided and enlarged, but fertilized polar nuclei had not yet split (Fig. 6c). These findings suggest that at this time, future nutrients required for embryo development could not be supplied leading to eventual embryo disintegration (Fig. 6d).
25 d, the globular embryos had gradually degraded, and fertilized polar nuclei had not developed (Fig. 7d). At this stage, the setting fruits turned yellow and then fell off, and the inner ovaries gradually withered. Embryo development for C. praecox × C. salicifolius hybrids was essentially the same as C. praecox × C. nitens hybrids (Table 3), and the abortions also included mainly no fertilization (Fig. 7e), and endosperm development retardation (Fig. 7f).
3.5. Hybrid embryo development in C. praecox × C. nitens and C. praecox × C. salicifolius
4. Discussion
In C. praecox × C. nitens, double fertilization occurred within the embryo sacs by 8 d after pollination (Fig. 7a), and cases with normal double fertilization accounted for 30.4% of the observed samples (Table 2). At 15 d after pollination, normal globular embryos were observed in only 20.5% of embryo sacs (Fig. 7b and Table 2); however, no endosperm development was detected (Fig. 7c). By
Reproductive barriers in plants can mostly be divided into prefertilization barriers and post-fertilization barriers (Arnold, 1998). Pre-fertilization barriers result from interactions between stigmas and pollens, a lack of pollen germination, or the premature stoppage of pollen tube growth after germination, preventing the pollen tubes from reaching the embryo sac to complete fertilization; the
Fig. 6. The embryo abortion of natural pollination in C. praecox. (a) No double fertilization (at 9 d after pollination). (b) The fertilized egg did not develop (at 15 d after pollination). (c) The fertilized egg has developed, but fertilized polar nuclei did not develop (at 25 d after pollination). (d) Embryo disintegration (at 30 d after pollination). eg egg, pn polar nuclei. Bar = 50 m.
90
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
Fig. 7. Embryogenesis in C. praecox × C. nitens ((a)–(d)) and C. praecox × C. salicifolius after pollination ((e) and (f)). (a) Double fertilization (at 8 d after pollination). (b) Globular embryo (at 15 d after pollination). (c) Lack of endosperm development (at 15 d after pollination). (d) Globular embryo had disintegrated and endosperm did not develop (at 25 d after pollination, with arrows indicating fertilized polar nuclei). (e) Unfertilized embryo sac (at 10 d after pollination). (f) Retardation of endosperm development (at 15 d after pollination). cc central cell, ec egg cell, sn sperm nucleus, pn polar nuclei. Bar = 50 m. Table 2 The number of normal and aborted embryos at different development stages of hybrid embryo in C. praecox × C. nitens. Days after pollination
No. of embryos
No. of normal embryos
Percentage of normal embryo (%)
No. of aborted embryos
Percentage of aborted embryo (%)
6 8 13 15 25
56 56 49 73 55
12 17 9 15 0
21.4 30.4 18.4 20.5 0.0
32 37 37 56 55
57.1 66.1 75.5 76.7 100.0
primary post-fertilization barrier is deficiencies in the viability of hybrid embryos (Campbell et al., 2003; Martin et al., 2005). Prefertilization barriers can be overcome by improving pollination techniques, whereas post-fertilization barriers can be managed through the in vitro culture of hybrid embryos, which can improve the success rate of hybridization efforts (Van Tuyl et al., 1991). Pollen viability and stigma receptivity are the first considerations for successful hybridization. Relevant research on the pollen viability and stigma receptivity of C. praecox and C. nitens has been conducted (Zhou et al., 2003, 2006), and the results of these prior studies indicated that these two species exhibited high pollen viability and stigma receptivity for approximately 5 d and that pollen viability and stigma receptivity issues did not significantly impact hybrid seed setting. The considerations of whether pollen can germinate on the stigma and whether the germinated pollen tube can grow normally in the style to reach the embryo sac also have important effects on the success of hybridization. In the hybridization of tobacco (Nicotiana), Christopher et al. (2008), found that the pollen germination rate and the length of pollen tubes were positively correlated with the length of the male parent’s pistils. When a variety with short pistils was used as the male parent and a variety with long pistils was used as the female parent, the length of pollen tube growth would be insufficient to reach the embryo sac; thus, fertilization could not be completed (Christopher et al., 2008). In interspecific hybridization between Kunzea pomifera and K. ericoides (Myrtaceae), the main reason for hybridization failure was that the tip of the pollen tube swelled and ceased growing (Page et al., 2010). In Solanaceae, one type of interspecific
reproductive barrier (IRB), called unilateral incompatibility/incongruity (UI), was present (Mutschler and Liedl, 1994). Solanum pennellii rejected pollen from domesticated tomato, whereas domesticated tomato accepted pollen from S. pennellii; pollen rejection occurs in the upper part of the style in S. pennellii when crossed with S. lycopersicum pollen (Liedl et al., 1996). For interspecific hybridization involving the three examined Chimonanthus species, for both halves of all reciprocal crosses, pollens of male parents could adhere to and germinate on the stigmas of female parents, and pollen tubes were formed that reached the embryo sacs in 48 to 72 h after pollination (Figs. 2–4), and double fertilization occurred within the embryo sac by 8 d after pollination (Fig. 7a). The reproductive process was very similar with the intraspecific pollination in C. praecox (Zhou et al., 2006). These findings indicated that there were no pre-fertilization barriers to interspecific hybridization among the examined Chimonanthus species. The pistil and ovule development of the female parents are closely correlated with the seed-setting rates of hybrids. In particular, a high proportion of pistil dysplasia will greatly reduce the hybrid seed-setting rates. Under natural conditions, C. praecox has a winter flowering period. To ensure reproductive success by attracting pollinators and appropriately adapting to the harsh environment, it produces a large number of flowers, thereby enhancing their probabilities of successful reproduction; therefore, to properly regulate reproductive allocation, abortion will gradually occur in a large number of ovaries after flowering, and normal development will ultimately only occur in a certain percentage of ovaries.
Table 3 The number of normal and aborted embryos at different development stages of hybrid embryo in C. praecox × C. salicifolius. Days after pollination
No. of embryos
No. of normal embryos
Percentage of normal embryo (%)
No. of aborted embryos
Percentage of aborted embryo (%)
5 8 10 15 25
42 40 54 48 64
14 11 10 8 0
33.3 27.5 18.5 16.7 0.0
16 24 38 36 64
38.1 60.0 70.4 75.0 100.0
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
Under natural pollination conditions, high proportions of disorders and/or lack of coordination in embryo and endosperm development occur at different stages after pollination. These issues are the main causes of embryo abortion and contribute to the fact that the natural seed-setting rate for C. praecox is only 9.85% (Zhou et al., 2006). This study found that a certain percentage of abortions occurred after pollination among un-hybridized C. praecox embryos (Table 1 and Fig. 6) but that among hybrid embryos created by interspecific cross-pollination, there were greatly increased proportions of abortions and developmental disorders in various developmental stages (Tables 2 and 3 and Fig. 7). Sun et al. (2009), investigated reproductive disorders associated with the hybridization of Chrysanthemum morifolium ‘Aoyuntianshi’ and C. japonense and found that only 10% of embryos could develop to the cotyledon stage but that a large number of embryos underwent abnormal degradation during development; this phenomenon is the main reason for the low seed-setting rate in the hybridization. The results in the Trifolium genus revealed that in the hybridization of T. sarosiense and T. pratense, hybrid embryos underwent degradation at 7 d during the course of development into a globular embryo; in addition, hybrid fruit could not be obtained because the expansion of endothelial cells caused the shrinkage of embryo sac cavities (Jana et al., 2006). In this study, hybridizations involving C. praecox as the female parent and C. salicifolius or C. nitens as the male parent revealed that after pollination, the development of the embryos and the endosperms was uncoordinated and endosperm development was retarded; as a result, embryo degradation occurred in the globular stage, which was the main cause of hybridization failure. 5. Conclusions In summary, during interspecific hybridization in Chimonanthus, there were no reproductive barriers during the stage of pollen–pistil interactions; instead, reproductive barriers primarily occurred after fertilization. In the hybrids, the parent genomes interacted and underwent reorganization at different stages of development; as a result, there was a greatly increased occurrence of disorders and lack of coordination in the development of embryos and endosperms at various developmental stages. This phenomenon was the main reason for the lack of seed setting by interspecific hybrids. The artificial culture in vitro of immature embryos at appropriate stages could effectively overcome embryo abortion- or dysplasia-induced hybrid seed problems related to the lack of an embryo or to seed germination difficulties; in this case, embryo age, culture media, and culture methods have important impacts on the success of hybrid embryo culture (Van Tuyl et al., 1991; Van Creij et al., 1999, 2000; Ahn et al., 2003). Cheng et al. (2010) employed ovary rescue to create interspecific hybrids between Chrysanthemum morifolium ‘rm20-12’ and its wild relative C. nankingense, whose cold tolerance was significantly superior to that of the parents. Therefore, in future interspecific hybridizations among Chimonanthus species, an individual with a high seedsetting rate should be selected as the female parent; in addition, a combination of embryo rescue should be utilized before the abortion of the hybrid (before and after the globular embryo stage) to overcome the hybridization barriers and thereby obtain hybrid offspring. Acknowledgements This work was supported by the Program for the National Natural Science Foundation of China (Grant No. 31101571 and 31170656) and Zhejiang Provincial Natural Science Foundation of China (Grant No. Y3100332 and Y3110357), and Zhejiang Provincial
91
Major Program of New cultivar breeding (Grant No. 2012C12909-9 and 2012C12909-19). References Ahn, M.S., Lee, K.J., Jeong, J.S., 2003. Investigation on pollination methods for the intergeneric hybrid between Hemerocallis and Lilium. Acta Hortic. 620, 305–310. Arnold, M.L., 1998. Natural Hybridization and Evolution. Oxford University Press, Oxford. Campbell, D.R., Alarcon, R., Wu, C.A., 2003. Reproductive isolation and hybrid pollen disadvantage in Ipomopsis. J. Evol. Biol. 16, 536–540. Cheng, X., Chen, S., Chen, F., Fang, W., Deng, Y., She, L., 2010. Interspecific hybrids between Dendranthema morifolium (Ramat.) Kitamura and D. nankingense (Nakai) Tzvel. achieved using ovary rescue and their cold tolerance characteristics. Euphytica 172, 101–108. Christopher, B.L., Lawrence, E.P., Bruce, A.M., Timothy, P.H., 2008. Post-pollination hybridization barriers in Nicotiana section Alatae. Sex. Plant Reprod. 21, 183–195. Eeckhau, T.T., Keyser, E.D., Huylenbroeck, J.V., Riek, J.D., Bock-staele, E.V., 2007. Application of embryo rescue after interspecific crosses in the genus Rhododendron. Plant Cell, Tissue Organ Cult. 89, 29–35. Jana, R., Barbara, J., Hana, J., 2006. Identification of barriers to interspecific crosses in the genus Trifolium. Euphytica 151, 39–48. Hao, J., Qiang, S., 2007. The examples and analysis of whole mount clearing techniques in plant biology. Chin. Bull. Bot. 24, 490–497. ´ Lj., Vinterhalter, D., 2008. In vitro propagaKozomara, B., Vinterhalter, B., Radojevic, tion of Chimonanthus praecox (L.), a winter flowering ornamental shrub. In Vitro Cell. Dev. Biol. Plant 44, 142–147. Lasseigne, F.T., Fantz, P.R., Raulston, J.C., Straley, G.B., 2001. Sinocalycalycanthus raulstonii (Calycanthaceae): a new intergeneric hybrid between Sinocalycanthus chinensis and Calycanthus floridus. HortScience 36, 765–767. Li, Y., Li, B., 2000. The origin, evolution and distribution in Calycanthaceae. Guihaia 20, 295–300. Liedl, B.E., McCormick, S., Mutschler, M.A., 1996. Unilateral incongruity in crosses involving Lycopersicon pennellii and L. esculentum is distinct from selfincompatibility in expression, timing and location. Sex. Plant Reprod. 9, 299–308. Martin, N.H., Bouck, A.C., Arnold, M.L., 2005. Loci affecting long-term hybrid survivorship in Louisiana irises: implications for reproductive isolation and introgression. Evolution 59, 2116–2124. Mutschler, M.A., Liedl, B.E., 1994. Interspecific crossing barriers in Lycopersicon and their relationship to self-incompatibility. In: Williams, G., Clarke, A.E., Knox, B.R. (Eds.), Genetic Control of Self-incompatibility and Reproductive Development in Flowering Plants. Klewer, Netherlands, pp. 164–188. Ouyang, t., Mai, X., Xia, H., Chen, D., Wei, J., Tong, W., 2010. Analysis of chemical constituents of volatile oil from Chimonathus salicifolius S. Y. Hu by GC–MS. J. Nanchang Univ. (Nat. Sci.) 34, 78–81. Page, T., Moore, G.M., Will, J., Halloran, G.M., 2010. Breeding behaviour of Kunzea pomifera (Myrtaceae): self-incompatibility, intraspecific and interspecific crosscompatibility. Sex. Plant Reprod. 23, 239–253. Sun, C., Chen, F., Fang, W., Liu, Z., Ma, J., Teng, N., She, L., 2009. Cellular mechanism of reproductive barrier during cross breeding between Dendranthema grandiflorum ‘Aoyuntianshi’ and D. japonense. Sci. Agric. Sin. 42, 2085–2091. Shu, R., Li, S., Hu, H., Zhang, P., 2010. Studies on chemical constituents of Chimonanthus nitens Oliver. Chin. Pharm. J. 145, 1134–1135. Van Tuyl, J.M., Van Dien, M.P., Van Creij, M.G.M., Van Kleinwee, T.C.M., Franken, J., Bino, R.J., 1991. Application of in vitro pollination, ovary culture, ovule culture and embryo rescue for overcoming incongruity barriers in interspecific Lilium crosses. Plant Sci. 74, 115–126. Van Creij, M.G.M., Kerckhoffs, D.M.F.J., Van Tuyl, J.M., 1999. The effect of ovule age on ovary-slice culture and ovule culture in intraspecific and interspecific crosses with Tulipa gesneriana L. Euphytica 108, 21–28. Van Creij, M.G.M., Kerckhoffs, D.M.F.J., De Bruijn, S.M., Vreugdenhil, D., Van Tuyl, J.M., 2000. The effect of medium composition on ovary-slice culture and ovule culture in intraspecific Tulipa gesneriana crosses, Plant Cell. Tissue Organ Cult. 60, 61–67. Wang, W., Zhou, L., Liu, H., Bao, Z., Zhao, H., 2013. Histological reproductive barriers for intergeneric cross between Sinocalycanthus chinensis and Calycanthus floridus var. oblongifolius. Acta Hortic. Sin. 40, 1943–1950. Xu, N., Bai, H., Yan, X., Xu, J., 2006. Analysis of volatile components in essential oil of Chimonanthus nitens by Capillary gas chromatography–mass spectrometry. J. Instrum. Anal. 25, 90–93. Yao, Q., Xia, B., Ren, Q., Sheng, N., Sun, X., 2007. Heterosis of intergeneric hybrid between Sinocalycanthus chinensis and Calycanthus floridus. J. For. Sci. Technol. 34, 24–26. Zhang, R., Liu, H., Shen, X., 1998. Wax Shrubs in World (Calycanthaceae). China Science and Technology Press, Beijing. Zhou, L., Hao, R., Wu, J., 2006. The pollination biology of Chimonanthus praecox (L.) Link (Calycanthaceae). Acta Hortic. Sin. 33, 323–327. Zhou, L., Hao, R., Wu, J., Mao, Z., 2003. Pollination biology of Chimonanthus nitens Oliver. Acta Hortic. Sin. 30, 690–694.
Contents lists available at ScienceDirect
Scientia Horticulturae journal homepage: www.elsevier.com/locate/scihorti
Reproductive barriers in interspecific hybridizations among Chimonanthus praecox (L.) Link, C. salicifolius S. Y. Hu, and C. nitens Oliver from pollen–pistil interaction and hybrid embryo development Wenpeng Wang a , Lihua Zhou a , Yaohui Huang a , Zhiyi Bao a , Hongbo Zhao a,b,∗ a b
Department of Ornamental Horticulture, School of Landscape Architecture, Zhejiang Agriculture and Forestry University, 311300 Lin’an Hangzhou, China Nurturing Station for State Key Laboratory of Subtropical Silviculture, 311300 Lin’an Hangzhou, China
a r t i c l e
i n f o
Article history: Received 28 April 2014 Received in revised form 24 July 2014 Accepted 28 July 2014 Available online 17 August 2014 Keywords: Chimonanthus praecox C. salicifolius C. nitens Interspecific hybridization Pollen–pistil interaction Hybrid embryo development
a b s t r a c t To investigate interspecific cross-compatibility among Chimonanthus praecox (L.) Link, C. salicifolius S. Y. Hu, and C. nitens Oliver, pollen–pistil interaction by fluorescence microscopy and hybrid embryo development by whole mount clearing technique were observed. The results indicated that for all reciprocal crosses among the three species, pollen grains of the male parents could adhere to and germinate in the stigmas of female parents. For any cross combination, pollen tube growth could reach the embryo sac, and double fertilization could be completed. These findings indicated that no obvious barriers to interspecific hybridization existed at the stages of pollen germination and pollen tube elongation until double fertilization. Although certain proportion (29.6% at 9 d and 36.3% at 15 d after pollination) of abortions (e.g., non-fertilization and retarded or non-existent endosperm development) in normally pollinated embryos of C. praecox at different stages were also observed, a greatly increased proportion (66.1% at 8 d and 76.7% at 15 d after pollination in C. praecox × C. nitens, and 60.0% at 8 d and 75.0% at 15 d after pollination in C. praecox × C. salicifolius) of hybrid embryos did not complete double fertilization or completed double fertilization but underwent abortion due to the retardation of endosperm development. Comparing to normally pollinated embryos of C. praecox, the retardation of endosperm development in hybridizations occurred earlier. Moreover, the occurrence of abortions among different cross combinations was basically consistent. Thus, the retardation of the endosperm development of hybrid embryos after double fertilization (after the globular embryo stage) is an important factor leading to the failure of distant hybridization between Chimonanthus species. © 2014 Elsevier B.V. All rights reserved.
1. Introduction Calycanthaceae is a small and relatively primitive family that includes a total of three genera (Chimonanthus Lindley, Calycanthus L., and Sinocalycanthus Cheng et S.Y. Chang) and 7–9 species. This family has a disjunct distribution across East Asia and North America. In spite of a small number of species in the family, these species are highly diverse with respect to relevant traits, such as growth and ecological habits, flowering period, floral structure, odor, floral display, leaf structure, and leaf content, especially abundant diversity in the floral traits. Chimonanthus is endemic to China and the tertiary relict taxa, which include 6 species, mostly grow in mixed evergreen and deciduous broad-leaved forests and
∗ Corresponding author at: School of Landscape Architecture, Zhejiang Agriculture and Forestry University, 311300 Lin’an, Zhejiang, China. Tel.: +86 571 63748611. E-mail addresses: [email protected], [email protected] (H. Zhao). http://dx.doi.org/10.1016/j.scienta.2014.07.040 0304-4238/© 2014 Elsevier B.V. All rights reserved.
evergreen broad-leaved forests in warm temperate and subtropical regions (Zhang et al., 1998; Li and Li, 2000). The plants in this genus are widely distributed and are most abundant in the Yangtze River basin. Chimonanthus praecox (L.) Link is a traditionally ornamental fragrant plant that flowers in winter (from December to February) in China (Zhang et al., 1998; Kozomara et al., 2008). The flowers have an elegant aroma and flavor, with abundant variations in floral traits. C. nitens Oliver is a fragrant evergreen shrub that has leathery leaves with smooth surfaces, white tepals with a flowering period from September to November. C. salicifolius S. Y. Hu is a semi-evergreen shrub with narrow, willow-like leaves that have a rough upper surface and a lower surface covered in white powder. This plant’s flowers are similar to the flowers of C. nitens and also have a flowering period from September to November. These three species exhibit rich variations in growth habits, floral traits, flowering period, and stress resistance. In particular, the leaves of C. nitens and C. salicifolius are rich in alkaloids and flavonoids
86
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
(Xu et al., 2006; Ouyang et al., 2010; Shu et al., 2010), which not only possess extremely high medicinal value but also confer good insect resistance. The use of distant hybridization to introduce heterologous genes and thereby create novel germplasm with advantages from both parents to expand genetic pools is currently a primary method (Eeckhau et al., 2007). With respect to the distant hybridization in Calycanthaceae, Lasseigne et al. (2001), first reported obtaining intergeneric hybrids of S. chinensis × Calycanthus floridus, and Yao et al. (2007), have also reported obtaining intergeneric hybrids of these two species (S. chinensis × C. floridus). We have utilized conventional hybridization approaches to obtain C. floridus var. oblongifolius × S. chinensis hybrids, which first flowered in 2013 (unpublished work). Although these intergeneric hybrids have been obtained, the observed seed sets in hybridizations have been extremely low, indicating the presence of certain reproductive barriers in the intergeneric hybridization. Therefore, Wang et al. (2013), examined the underlying mechanisms of barriers to the hybridization of S. chinensis var. oblongifolius and C. floridus and found that reciprocal crosses involved different incompatibility mechanisms. When S. chinensis was the female parent, the main barrier to successful hybridization was the abnormal disintegration of the early hybrid embryo after fertilization. When C. floridus var. oblongifolius was the female parent, the main barriers to hybridization were due to the combination effects of high proportion of developmental abnormalities of female pistils and the abortion of early hybrid embryos. No prior studies have reported on the cross-compatibility of interspecific hybridizations within Chimonanthus. Interspecific hybridization may represent a method of successfully transferring or aggregating desirable traits to create new varieties with high ornamental value and strong stress resistance. However, in numerous tests of interspecific hybridization, conventional hybridization experiments involving the aforementioned these three species could produce the enlargement of fruit during the early post-pollination stages but could not successfully obtain hybrid seeds. Therefore, in this paper, the reproductive barrier of interspecific hybridization within Chimonanthus is investigated from the perspectives of pollen–pistil interaction (pollen germination, pollen tube elongation, and fertilization), and hybrid embryo development after pollination. Using this approach, the present study clarified the underlying mechanisms of reproductive barriers to hybridization, thereby providing a reference for future breeding efforts involving the distant hybridization of Chimonanthus. 2. Materials and methods 2.1. Materials The materials of C. praecox ‘CHg08’ (Fig. 1a), C. nitens (Fig. 1b), and C. salicifolius (Fig. 1c) were from the germplasm resource of Zhejiang Agriculture and Forestry University. C. praecox ‘CHg08’ is an early-blooming germplasm whose anthesis is from the end of October to the beginning of December. C. nitens and C. salicifolius are introduced from the field respectively in Lishui, Zhejiang and Qiyunshan, Anhui. These plants had been cultivated for above 8 years and were capable of normal flowering. Hybridization pollinations were conducted during the October–December flowering periods in 2010–2012. 2.2. Cross-pollination and sample fixation Reciprocal crosses was conducted for each pairwise combination of the three aforementioned Chimonanthus species according
to Wang et al. (2013), and a normal pollination (natural pollination) of C. praecox ‘CHg08’ using the conspecific pollen from other strains was as a control. The following specific operating procedures were employed. On a clear day, well-developed flower buds on a female parent were selected. Emasculation was performed on the relaxed flower buds by entirely removing the upcoming inner and outer tepals as well as the stamens (without damaging the stigma). Anthers just before dehiscence from the male parents were collected and dehydrated under incandescent lamp for 2 h, and then pollen were picked up and used to pollinate the stigmas of female parents. After pollination, flowers were promptly bagged and labeled. Pistils were collected at 1, 2, 6, 12, 24, 36, 48, 72, and 96 h after pollination, fixed with FAA (formalin: acetic acid: 50% alcohol, 1:1:8) in a 5 mL centrifuge tube for more than 24 h, and observed by fluorescence microscopy. In addition, to observe hybrid embryo development, for 3–10 d after pollination, sampling was performed once per day; for 10–20 d after pollination, sampling was performed at 11 d, 13 d, 15 d, 17 d and 20 d; beyond 20 d after pollination, sampling was performed every 5 d until shedding occurred. 2.3. Fluorescence microscopy A slightly modified version of the fluorescence microscopy method described by Zhou et al. (2006), was utilized in this study. The developing fruits were gently cut with a razor blade, and the whole pistils including stigmas and ovaries were separated from fruits and collected with tweezers. Each sample was softened and bleached with 2 mol L−1 NaOH for 2 h and then soaked in 0.1% aniline blue staining solution for 10 h (overnight). Samples were mounted in glycerol, and sections were prepared. The adhesion to and germination of the pollen on the stigma and pollen tube elongation was observed under the fluorescence microscope (Leica DM4000). 2.4. The observation of embryo development using the whole mount clearing technique The whole mount clearing technique allows for the rapid observation of internal organization and structural development within certain materials that are small in size, difficult to prepare by section, or inefficient to observe by section (Jana et al., 2006; Hao and Qiang, 2007). The following specific procedures were employed in this process. (1) Samples (developing fruits) were fixed for more than 24 h was gently cut with a razor blade, and the pistils were separated with tweezers and placed into a centrifuge tube; (2) the pistils were successively treated with a 75–85–95–100% I–100% II ethanol gradient for 2 h per ethanol solution, (3) were treated with an ethanol and methyl salicylate mixture (1:1 by volume) for 3 h, (4) and were cleared three times with methyl salicylate for 12 h for each of the first two clearing treatments and for 24 h for the final clearing treatment. (5) Then samples were placed on a concave glass slide and observed and photographed under the differential interference microscope (Nikon Eclipse 80i). 3. Results 3.1. Pollen germination and pollen tube elongation of hybridization between C. praecox and C. salicifolius In C. praecox × C. salicifolius, pollen grains adhered to the stigma by 6 h after pollination (Fig. 2a); pollen germination had begun at 12 h after pollination (Fig. 2b); the pollen tube was continuing to grow downward at 24 h after pollination (Fig. 2c); and the pollen tube had reached the embryo sac at 48 h after pollination (Fig. 2d). When C. salicifolius was the female parent, observations again
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
87
Fig. 1. Flower morphology of C. praecox (a), C. nitens (b) and C. salicifolius (c).
indicated that the pollen tube reached the embryo sac by 48 h after pollination (Fig. 2e).
and the pollen tube had approached the embryo sac by no later than 48 h after pollination (Fig. 4f).
3.2. Pollen germination and pollen tube elongation of hybridization between C. praecox and C. nitens
3.4. Normal embryo development and abortion in C. praecox after natural pollination
In C. praecox × C. nitens, pollen germination had begun at 12 h after pollination (Fig. 3a); the pollen tube had grown to reach the middle of the style by 24 h after pollination (Fig. 3b); the pollen tubes in certain samples had reached the embryo sac by 48 h after pollination (Fig. 3c); and in most observations of styles, pollen tubes had grown to reach the embryo sac by 72 h after pollination (Fig. 3d). When C. nitens was the female parent, observations again indicated that the pollen tube had reached the embryo sac by 72 h after pollination (Fig. 3e).
The whole clearing technique was applied to observe the embryo development of C. praecox under natural pollination (xenogamy with conspecific pollen). The results indicated that eggs and central cells were observed at 6 d after pollination (Fig. 5a), and that double fertilization in embryo sac was observed at 9 d after pollination (Fig. 5b). At this time, the egg was located at the end of the micropyle, next to two synergids; the central cell was in the middle, with three antipodal cells above the central cell. Globular embryos were observed at 20 d after pollination (Fig. 5c); at this time, endosperm cells had formed. The longitudinal division of the central endosperm cells had formed two strips of endosperm cells. Eventually, endosperm cells filled the entire long, narrow embryo sac cavity (Fig. 5e); during this process, nucellar tissue was gradually digested and absorbed. Heart-shaped embryos appeared at 25 d after pollination (Fig. 5d), and mature embryos were observed at approximately 35 d after pollination (Fig. 5f). Under natural pollination for C. praecox, a certain percentage of embryo abortions were observed at various time points after pollination (Table 1). At 9 d after pollination, double fertilization was not observed in approximately 29.6% of the sampled materials (Fig. 6a). At 15 d after pollination, 36.3% of fertilized eggs in the embryo sac
3.3. Pollen germination and pollen tube elongation of hybridization between C. salicifolius and C. nitens When C. salicifolius was the female parent, pollen had begun to germinate by 12 h after pollination (Fig. 4a); pollen tube growth was continuing at 48 h after pollination (Fig. 4b); and the pollen tube had grown sufficiently to approach the embryo sac by 72 h after pollination (Fig. 4c). When C. nitens was the female parent, the similar results was observed; in particular, a pollen tube was clearly visible at 24 h (Fig. 4d) and 36 h (Fig. 4e) after pollination,
Fig. 2. Growth of pollen tubes in styles after pollination between C. praecox and C. salicifolius. (a)–(d) Pollen germination and pollen tube growth at 6, 12, 24 and 48 h, respectively, after pollination for the C. praecox × C. salicifolius cross. e Pollen tube growth at 48 h after pollination for the C. salicifolius × C. praecox cross. Bar = 100 m.
88
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
Fig. 3. Growth of pollen tubes in styles after pollination between C. praecox and C. nitens. (a)–(d) Pollen germination and pollen tube growth at 12, 24, 48 and 72 h, respectively, after pollination for the C. praecox × C. nitens cross. e Pollen tube growth at 72 h after pollination for the C. nitens × C. praecox cross. Bar = 100 m.
Fig. 4. Growth of pollen tubes in styles after pollination between C. salicifolius and C. nitens. (a)–(c) Pollen germination and pollen tube growth at 12, 48, and 72 h, respectively, after pollination for the C. salicifolius × C. nitens cross. (d)–(f) Pollen tube growth at 24, 36, and 48 h, respectively, after pollination for the C. nitens × C. salicifolius cross. Bar = 100 m.
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
89
Fig. 5. Embryogenesis of C. praecox at different stages after natural pollination. (a) Unfertilized embryo sac (at 6 d after pollination). (b) Double fertilization (at 9 d after pollination, with synergids and antipodal cells visible within the embryo sac). (c) Globular embryo (at 20 d after pollination). (d) Heart-shaped embryo (at 25 d after pollination). (e) The formation of endosperm cells in a long and narrow shape (at 25 d after pollination). (f) Mature embryo (at 35 d after pollination). cc central cell, ec egg cell, ant antipodal cell, sn sperm nucleus, syn synergid. Bar = 50 m. Table 1 The number of normal and aborted embryos in C. praecox at different stages after natural pollination. Days after pollination
No. of embryos
No. of normal embryos
Percentage of normal embryo (%)
No. of aborted embryos
Percentage of aborted embryo (%)
5 9 15 20 25
43 27 33 25 36
36 18 14 8 8
83.7 66.7 42.4 32.0 22.2
4 8 12 12 18
9.0 29.6 36.3 48.0 50.0
cavity stopped developing without undergoing divisions (Fig. 6b). At 25 d after pollination, approximately 50.0% of fertilized eggs in the embryo sac had divided and enlarged, but fertilized polar nuclei had not yet split (Fig. 6c). These findings suggest that at this time, future nutrients required for embryo development could not be supplied leading to eventual embryo disintegration (Fig. 6d).
25 d, the globular embryos had gradually degraded, and fertilized polar nuclei had not developed (Fig. 7d). At this stage, the setting fruits turned yellow and then fell off, and the inner ovaries gradually withered. Embryo development for C. praecox × C. salicifolius hybrids was essentially the same as C. praecox × C. nitens hybrids (Table 3), and the abortions also included mainly no fertilization (Fig. 7e), and endosperm development retardation (Fig. 7f).
3.5. Hybrid embryo development in C. praecox × C. nitens and C. praecox × C. salicifolius
4. Discussion
In C. praecox × C. nitens, double fertilization occurred within the embryo sacs by 8 d after pollination (Fig. 7a), and cases with normal double fertilization accounted for 30.4% of the observed samples (Table 2). At 15 d after pollination, normal globular embryos were observed in only 20.5% of embryo sacs (Fig. 7b and Table 2); however, no endosperm development was detected (Fig. 7c). By
Reproductive barriers in plants can mostly be divided into prefertilization barriers and post-fertilization barriers (Arnold, 1998). Pre-fertilization barriers result from interactions between stigmas and pollens, a lack of pollen germination, or the premature stoppage of pollen tube growth after germination, preventing the pollen tubes from reaching the embryo sac to complete fertilization; the
Fig. 6. The embryo abortion of natural pollination in C. praecox. (a) No double fertilization (at 9 d after pollination). (b) The fertilized egg did not develop (at 15 d after pollination). (c) The fertilized egg has developed, but fertilized polar nuclei did not develop (at 25 d after pollination). (d) Embryo disintegration (at 30 d after pollination). eg egg, pn polar nuclei. Bar = 50 m.
90
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
Fig. 7. Embryogenesis in C. praecox × C. nitens ((a)–(d)) and C. praecox × C. salicifolius after pollination ((e) and (f)). (a) Double fertilization (at 8 d after pollination). (b) Globular embryo (at 15 d after pollination). (c) Lack of endosperm development (at 15 d after pollination). (d) Globular embryo had disintegrated and endosperm did not develop (at 25 d after pollination, with arrows indicating fertilized polar nuclei). (e) Unfertilized embryo sac (at 10 d after pollination). (f) Retardation of endosperm development (at 15 d after pollination). cc central cell, ec egg cell, sn sperm nucleus, pn polar nuclei. Bar = 50 m. Table 2 The number of normal and aborted embryos at different development stages of hybrid embryo in C. praecox × C. nitens. Days after pollination
No. of embryos
No. of normal embryos
Percentage of normal embryo (%)
No. of aborted embryos
Percentage of aborted embryo (%)
6 8 13 15 25
56 56 49 73 55
12 17 9 15 0
21.4 30.4 18.4 20.5 0.0
32 37 37 56 55
57.1 66.1 75.5 76.7 100.0
primary post-fertilization barrier is deficiencies in the viability of hybrid embryos (Campbell et al., 2003; Martin et al., 2005). Prefertilization barriers can be overcome by improving pollination techniques, whereas post-fertilization barriers can be managed through the in vitro culture of hybrid embryos, which can improve the success rate of hybridization efforts (Van Tuyl et al., 1991). Pollen viability and stigma receptivity are the first considerations for successful hybridization. Relevant research on the pollen viability and stigma receptivity of C. praecox and C. nitens has been conducted (Zhou et al., 2003, 2006), and the results of these prior studies indicated that these two species exhibited high pollen viability and stigma receptivity for approximately 5 d and that pollen viability and stigma receptivity issues did not significantly impact hybrid seed setting. The considerations of whether pollen can germinate on the stigma and whether the germinated pollen tube can grow normally in the style to reach the embryo sac also have important effects on the success of hybridization. In the hybridization of tobacco (Nicotiana), Christopher et al. (2008), found that the pollen germination rate and the length of pollen tubes were positively correlated with the length of the male parent’s pistils. When a variety with short pistils was used as the male parent and a variety with long pistils was used as the female parent, the length of pollen tube growth would be insufficient to reach the embryo sac; thus, fertilization could not be completed (Christopher et al., 2008). In interspecific hybridization between Kunzea pomifera and K. ericoides (Myrtaceae), the main reason for hybridization failure was that the tip of the pollen tube swelled and ceased growing (Page et al., 2010). In Solanaceae, one type of interspecific
reproductive barrier (IRB), called unilateral incompatibility/incongruity (UI), was present (Mutschler and Liedl, 1994). Solanum pennellii rejected pollen from domesticated tomato, whereas domesticated tomato accepted pollen from S. pennellii; pollen rejection occurs in the upper part of the style in S. pennellii when crossed with S. lycopersicum pollen (Liedl et al., 1996). For interspecific hybridization involving the three examined Chimonanthus species, for both halves of all reciprocal crosses, pollens of male parents could adhere to and germinate on the stigmas of female parents, and pollen tubes were formed that reached the embryo sacs in 48 to 72 h after pollination (Figs. 2–4), and double fertilization occurred within the embryo sac by 8 d after pollination (Fig. 7a). The reproductive process was very similar with the intraspecific pollination in C. praecox (Zhou et al., 2006). These findings indicated that there were no pre-fertilization barriers to interspecific hybridization among the examined Chimonanthus species. The pistil and ovule development of the female parents are closely correlated with the seed-setting rates of hybrids. In particular, a high proportion of pistil dysplasia will greatly reduce the hybrid seed-setting rates. Under natural conditions, C. praecox has a winter flowering period. To ensure reproductive success by attracting pollinators and appropriately adapting to the harsh environment, it produces a large number of flowers, thereby enhancing their probabilities of successful reproduction; therefore, to properly regulate reproductive allocation, abortion will gradually occur in a large number of ovaries after flowering, and normal development will ultimately only occur in a certain percentage of ovaries.
Table 3 The number of normal and aborted embryos at different development stages of hybrid embryo in C. praecox × C. salicifolius. Days after pollination
No. of embryos
No. of normal embryos
Percentage of normal embryo (%)
No. of aborted embryos
Percentage of aborted embryo (%)
5 8 10 15 25
42 40 54 48 64
14 11 10 8 0
33.3 27.5 18.5 16.7 0.0
16 24 38 36 64
38.1 60.0 70.4 75.0 100.0
W. Wang et al. / Scientia Horticulturae 177 (2014) 85–91
Under natural pollination conditions, high proportions of disorders and/or lack of coordination in embryo and endosperm development occur at different stages after pollination. These issues are the main causes of embryo abortion and contribute to the fact that the natural seed-setting rate for C. praecox is only 9.85% (Zhou et al., 2006). This study found that a certain percentage of abortions occurred after pollination among un-hybridized C. praecox embryos (Table 1 and Fig. 6) but that among hybrid embryos created by interspecific cross-pollination, there were greatly increased proportions of abortions and developmental disorders in various developmental stages (Tables 2 and 3 and Fig. 7). Sun et al. (2009), investigated reproductive disorders associated with the hybridization of Chrysanthemum morifolium ‘Aoyuntianshi’ and C. japonense and found that only 10% of embryos could develop to the cotyledon stage but that a large number of embryos underwent abnormal degradation during development; this phenomenon is the main reason for the low seed-setting rate in the hybridization. The results in the Trifolium genus revealed that in the hybridization of T. sarosiense and T. pratense, hybrid embryos underwent degradation at 7 d during the course of development into a globular embryo; in addition, hybrid fruit could not be obtained because the expansion of endothelial cells caused the shrinkage of embryo sac cavities (Jana et al., 2006). In this study, hybridizations involving C. praecox as the female parent and C. salicifolius or C. nitens as the male parent revealed that after pollination, the development of the embryos and the endosperms was uncoordinated and endosperm development was retarded; as a result, embryo degradation occurred in the globular stage, which was the main cause of hybridization failure. 5. Conclusions In summary, during interspecific hybridization in Chimonanthus, there were no reproductive barriers during the stage of pollen–pistil interactions; instead, reproductive barriers primarily occurred after fertilization. In the hybrids, the parent genomes interacted and underwent reorganization at different stages of development; as a result, there was a greatly increased occurrence of disorders and lack of coordination in the development of embryos and endosperms at various developmental stages. This phenomenon was the main reason for the lack of seed setting by interspecific hybrids. The artificial culture in vitro of immature embryos at appropriate stages could effectively overcome embryo abortion- or dysplasia-induced hybrid seed problems related to the lack of an embryo or to seed germination difficulties; in this case, embryo age, culture media, and culture methods have important impacts on the success of hybrid embryo culture (Van Tuyl et al., 1991; Van Creij et al., 1999, 2000; Ahn et al., 2003). Cheng et al. (2010) employed ovary rescue to create interspecific hybrids between Chrysanthemum morifolium ‘rm20-12’ and its wild relative C. nankingense, whose cold tolerance was significantly superior to that of the parents. Therefore, in future interspecific hybridizations among Chimonanthus species, an individual with a high seedsetting rate should be selected as the female parent; in addition, a combination of embryo rescue should be utilized before the abortion of the hybrid (before and after the globular embryo stage) to overcome the hybridization barriers and thereby obtain hybrid offspring. Acknowledgements This work was supported by the Program for the National Natural Science Foundation of China (Grant No. 31101571 and 31170656) and Zhejiang Provincial Natural Science Foundation of China (Grant No. Y3100332 and Y3110357), and Zhejiang Provincial
91
Major Program of New cultivar breeding (Grant No. 2012C12909-9 and 2012C12909-19). References Ahn, M.S., Lee, K.J., Jeong, J.S., 2003. Investigation on pollination methods for the intergeneric hybrid between Hemerocallis and Lilium. Acta Hortic. 620, 305–310. Arnold, M.L., 1998. Natural Hybridization and Evolution. Oxford University Press, Oxford. Campbell, D.R., Alarcon, R., Wu, C.A., 2003. Reproductive isolation and hybrid pollen disadvantage in Ipomopsis. J. Evol. Biol. 16, 536–540. Cheng, X., Chen, S., Chen, F., Fang, W., Deng, Y., She, L., 2010. Interspecific hybrids between Dendranthema morifolium (Ramat.) Kitamura and D. nankingense (Nakai) Tzvel. achieved using ovary rescue and their cold tolerance characteristics. Euphytica 172, 101–108. Christopher, B.L., Lawrence, E.P., Bruce, A.M., Timothy, P.H., 2008. Post-pollination hybridization barriers in Nicotiana section Alatae. Sex. Plant Reprod. 21, 183–195. Eeckhau, T.T., Keyser, E.D., Huylenbroeck, J.V., Riek, J.D., Bock-staele, E.V., 2007. Application of embryo rescue after interspecific crosses in the genus Rhododendron. Plant Cell, Tissue Organ Cult. 89, 29–35. Jana, R., Barbara, J., Hana, J., 2006. Identification of barriers to interspecific crosses in the genus Trifolium. Euphytica 151, 39–48. Hao, J., Qiang, S., 2007. The examples and analysis of whole mount clearing techniques in plant biology. Chin. Bull. Bot. 24, 490–497. ´ Lj., Vinterhalter, D., 2008. In vitro propagaKozomara, B., Vinterhalter, B., Radojevic, tion of Chimonanthus praecox (L.), a winter flowering ornamental shrub. In Vitro Cell. Dev. Biol. Plant 44, 142–147. Lasseigne, F.T., Fantz, P.R., Raulston, J.C., Straley, G.B., 2001. Sinocalycalycanthus raulstonii (Calycanthaceae): a new intergeneric hybrid between Sinocalycanthus chinensis and Calycanthus floridus. HortScience 36, 765–767. Li, Y., Li, B., 2000. The origin, evolution and distribution in Calycanthaceae. Guihaia 20, 295–300. Liedl, B.E., McCormick, S., Mutschler, M.A., 1996. Unilateral incongruity in crosses involving Lycopersicon pennellii and L. esculentum is distinct from selfincompatibility in expression, timing and location. Sex. Plant Reprod. 9, 299–308. Martin, N.H., Bouck, A.C., Arnold, M.L., 2005. Loci affecting long-term hybrid survivorship in Louisiana irises: implications for reproductive isolation and introgression. Evolution 59, 2116–2124. Mutschler, M.A., Liedl, B.E., 1994. Interspecific crossing barriers in Lycopersicon and their relationship to self-incompatibility. In: Williams, G., Clarke, A.E., Knox, B.R. (Eds.), Genetic Control of Self-incompatibility and Reproductive Development in Flowering Plants. Klewer, Netherlands, pp. 164–188. Ouyang, t., Mai, X., Xia, H., Chen, D., Wei, J., Tong, W., 2010. Analysis of chemical constituents of volatile oil from Chimonathus salicifolius S. Y. Hu by GC–MS. J. Nanchang Univ. (Nat. Sci.) 34, 78–81. Page, T., Moore, G.M., Will, J., Halloran, G.M., 2010. Breeding behaviour of Kunzea pomifera (Myrtaceae): self-incompatibility, intraspecific and interspecific crosscompatibility. Sex. Plant Reprod. 23, 239–253. Sun, C., Chen, F., Fang, W., Liu, Z., Ma, J., Teng, N., She, L., 2009. Cellular mechanism of reproductive barrier during cross breeding between Dendranthema grandiflorum ‘Aoyuntianshi’ and D. japonense. Sci. Agric. Sin. 42, 2085–2091. Shu, R., Li, S., Hu, H., Zhang, P., 2010. Studies on chemical constituents of Chimonanthus nitens Oliver. Chin. Pharm. J. 145, 1134–1135. Van Tuyl, J.M., Van Dien, M.P., Van Creij, M.G.M., Van Kleinwee, T.C.M., Franken, J., Bino, R.J., 1991. Application of in vitro pollination, ovary culture, ovule culture and embryo rescue for overcoming incongruity barriers in interspecific Lilium crosses. Plant Sci. 74, 115–126. Van Creij, M.G.M., Kerckhoffs, D.M.F.J., Van Tuyl, J.M., 1999. The effect of ovule age on ovary-slice culture and ovule culture in intraspecific and interspecific crosses with Tulipa gesneriana L. Euphytica 108, 21–28. Van Creij, M.G.M., Kerckhoffs, D.M.F.J., De Bruijn, S.M., Vreugdenhil, D., Van Tuyl, J.M., 2000. The effect of medium composition on ovary-slice culture and ovule culture in intraspecific Tulipa gesneriana crosses, Plant Cell. Tissue Organ Cult. 60, 61–67. Wang, W., Zhou, L., Liu, H., Bao, Z., Zhao, H., 2013. Histological reproductive barriers for intergeneric cross between Sinocalycanthus chinensis and Calycanthus floridus var. oblongifolius. Acta Hortic. Sin. 40, 1943–1950. Xu, N., Bai, H., Yan, X., Xu, J., 2006. Analysis of volatile components in essential oil of Chimonanthus nitens by Capillary gas chromatography–mass spectrometry. J. Instrum. Anal. 25, 90–93. Yao, Q., Xia, B., Ren, Q., Sheng, N., Sun, X., 2007. Heterosis of intergeneric hybrid between Sinocalycanthus chinensis and Calycanthus floridus. J. For. Sci. Technol. 34, 24–26. Zhang, R., Liu, H., Shen, X., 1998. Wax Shrubs in World (Calycanthaceae). China Science and Technology Press, Beijing. Zhou, L., Hao, R., Wu, J., 2006. The pollination biology of Chimonanthus praecox (L.) Link (Calycanthaceae). Acta Hortic. Sin. 33, 323–327. Zhou, L., Hao, R., Wu, J., Mao, Z., 2003. Pollination biology of Chimonanthus nitens Oliver. Acta Hortic. Sin. 30, 690–694.