Background

The wild protospecies of lily are divided into seven groups. The hybrids are produced by cross hybridization within one species, such as Asiatic hybrids, Oriental hybrids, Longiflorum hybrids, Trumpet hybrids. The hybrids are produced by cross hybridization between two species, such as Longiflorum×Asiatic hybrids, Oriental×Trumpet hybrids, Oriental×Asiatic hybrids, Longiflorum×Oriental hybrids (Van, 1987; Zhou et al., 2000). Due to integrate advantages of two species, interspecific hybrids have more obvious superiorities, for example, their stress resistance would be strengthened a lot and ornamental value would become more novel.

Recently, interspecific hybrid breeding has become a goal for lily breeders to strive for. In the end of last century, western scientists had invented the integrated technology in vitro, which meant using pollination in vitro, cut style in vitro or style grafting in vitro to overcome obstacles before fertilization; using ovary slice culture in vitro, ovule culture in vitro or embryo rescue in vitro to overcome obstacles after fertilization, and cultivate quantities of interspecific hybrids (Van et al., 2000). However, the above technologies are all operated in vitro and the process is much more complex so that the success rate is not high (the emergence rate of embryo rescue sometimes is less than 0.1%). Therefore, there are scarcely successful reports even though Chinese lily breeding experts spend years breeding lily interspecific hybrids by these technologies.

Internationally, the interspecific hybrid cultivated in general is back-cross hybrid between F1 generation and one of parents, which is usually triploid hybrid. Zhou Shujun on behalf of lily breeders recently put forward a new theory: lily is a plant of Fritillaria, the amount of polar nucleus DNA after its blastocyst meiosis is the double of body cells. When a triploid Fritillaria plant acted as the female parent, the process of meiosis might be abnormal, but if it hybridized with normal diploid or tetraploid, the hybrid would have aneuploid embryo but euploid endosperm possibly. The euploid endosperm would be well developed because of chromosome (gene) balance, thus let aneuploid embryo fed in early phase and finally overcome obstacles after fertilization to germinate seedlings (Zhou et al., 2011; Zhou et al., 2012).

The nature world has created many aneuploid species due to natural distant hybridization or mutagenesis of environmental factors and they become one of sources for new species evolution. People can also create aneuploid species through distant or ploidy hybridization. Some aneuploid species might have significant changes on traits for addition and loss of chromosomes. For example, someone investigated 38 Chinese traditional chrysanthemum species of ornamental value, including 60.5% aneuploid species (Zhu et al., 2011). It is of great significance for breeding to create aneuploid variation.

This study chose triploid OT lilies as female parents and diploid O lilies as male parents, and deliberately designed hybrid endosperm was euploid. The way to breed quantities of hybrid embryos in early stage with well-developed endosperms would simplify the process of interspecific hybrid breeding and also might create lots of aneuploid variations, which is of significant values for cultivating interspecific lily hybrids.

1 Results and Analysis

1.1 Observation of several types of OT lily chromosome

Through observing chromosome karyotypes, all 6 OT lilies cultivars used in this test were triploid. The number of their chromosomes was 2n=3x=36 (Figure 1).

1.2 Observation of fruit shape and statistics of fruit set rate and plump ovules number

The observation of fruit shape and the statistics of fruit set rate and plump ovules number were conducted on the 60 day after pollination. Parts of cross combinations were shown in Figure 2. Different combinations had different characteristics of fruit shape. Combination S showed its upper part was plump but lower part was wizened; Combination U looked uneven and wrinkled on surface; Combination R was not plump and wrinkled on surface; Combination V was plump and had a bulging body; Combination O was plump but had a wrinkled body; Combination I was plump and had a smooth surface; Combination M was less plump and bulging in the middle; Combination N was more bulging in the middle than its both ends; Combination J was plump evenly and had a smooth surface; Combination Q was less plump; Combination P had a bulging body but some of parts were wizened and wrinkled on surface; Combination L was bulging in the middle and very plump.

In 24 OT(♀)×O(♂) cross combinations, 16 combinations could bear fruits accounting for 66.7%, of which 9 combinations’ fruit set rates had reached over 50% and among those, 8 combinations of Robina and Competition as female parents all had much higher fruit set rates (Table 1). In 8 cross combinations of no fruits, 4 combinations chose Muscat as female parent and 2 combinations chose Table dance, which indicated those two kinds of OT lily cultivars might not be suitable as female parents for hybridization. 5 combinations of Tiber as male parent indicated the O lily cultivar might have poorer pollen viability. 4 combinations using the cut style pollination had less fruits, which indicated the normal stigma pollination method was better than the cut style, for the former fruit set rate was obviously higher than the latter. 5 back-cross combinations didn’t bear fruits and neither did 5 self-cross combinations of OT lilies.

The statistics of plump seeds number in different cross combinations before embryo rescue was shown in Table 2. It was found that plump seeds: the endosperm was thick and solid, the embryo was yellow and clavate; not plump seeds: the endosperm was wizened, the embryo was small or not well developed. Most of combinations bearing fruits could have plump seeds, the rate (plump seeds number/ovules amount) was between 2.49%-16.78%. A few combinations (such as Competition×Tiber, Robina×Tiber): fruits grew normally in early stage but became wizened in the later and finally none plump seeds. The full-shaped fruits produced more plump seeds.

1.3 Hybrid embryo rescue

The statistics of ovule emergence rates in different cross combinations on the 60 day after embryo rescue was shown in Figure 3 and Table 3. It was found that 4 cross combinations including Competition×Siberia, etc.: the ovule emergence rate was the highest, between 5.26%~18.56%, when the embryo rescue was conducted on the 60 d after pollination. The rate was obviously lower when the embryo rescue was conducted on the 65 d after pollination. Almost no ovule would sprout when it was conducted on the 70 d or 75 d after pollination. The emergence rate of Robina×Siberia cross combination was the highest.

2 Discussion

2.1 The six triploid OT lily cultivars

At first, people get OT lily interspecific hybrids by integrated technology in vitro (Van et al., 2000), it is complex with low success rate and most obtained F1 generations are diploid OT cultivars. After accumulating some diploid OT cultivars by strenuous efforts, people start to use methods of sexual polyploidy and ploidy hybridization in order to get triploid OT cultivars (Zhang et al., 2012). First, in order to make diploid OT lilies regain fertility, the way to guide them become tetraploid is applied. Then tetraploid as female parent and O lilies as male parent, back-cross tests are conducted to produce allotriploid OT lilies, this method is called as ploidy hybridization. But in some cases, diploid OT lilies produce some special egg cells whose chromosome number is 2n. Then this diploid as female parent and O lilies as male parent, back-cross tests are also conducted with embryo rescue in vitro to produce OT lilies, this technology is called as sexual polyploidization (Zhou et al., 2008). Compared with diploid OT lily, triploid OT lily is better in ornamental traits, stress resistance and more popular in market. This time, 6 OT lily species were introduced from Netherland and they were all triploids after observation, which could provide cytology basis for breeding later.

2.2 The pollination technology

The lily style is longer, pre-fertilization obstacle mainly manifests as the incompatibility between stigma and style. One of common ways to overcome pre-fertilization obstacle is the cut style pollination (Van et al., 1988; Chen et al., 2007). But someone found that sometimes the method could make the seed set rate reduced and they thought it might be caused by immature pollen tube untimely reached in embryo sac (Van et al., 1987); or maybe because pollen tube didn’t absorb nutrition or get signal guiding in the style, which obviously happened in the compatible pollination (Shao et al., 2014). In this study, the fruit and seed set rate of the cut style pollination were both lower than those of the normal stigma pollination, it may be because triploid OT lilies had at least one set of genome the same as male parent, and there were fewer incompatible factors in the stigma and style, so that when using the normal stigma pollination method, it would not meet too large incompatible obstacles. However, the cut style pollination method would limit pollen tube to absorb nutrition and get signals when it passed through the style, so that the seed set rate was lower. In this study, OT lilies self-cross and O(♀)×OT(♂) back-cross both had no fruits, it is because triploid OT lilies had undergone abnormal meiosis and couldn’t produce functional pollens.

2.3 The hybridization of triploid OT×diploid O lilies

Lily has a Fritillaria embryo sac of four cells and eight nucleuses (Maheshwari, 1971). Based on hybridization experiments and cytological study, Zhou Shujun thought the lily of Fritillaria embryo sac usually produced aneuploid egg cells and euploid polar nucleus when the abnormal meiosis was carried out. Allotriploid lily generally underwent abnormal meiosis to form aneuploid egg cells and hexaploid polar nucleus (Zhou, 2014). After double fertilization, if the endosperm contained five identical genomes, it meant that the endosperm developed well and would produce a good ‘nurturing’ effect on aneuploid embryos (Zhou et al., 2012). Based on this theory, Zhou Shujun crossed triploid and diploid lilies to create a large number of interspecific aneuploid hybrids (Yuan et al., 2013; Zhou et al., 2014). In this research, triploid OT lilies were crossed with diploid O lilies, and 12 combinations successfully got hybrid seedlings, which supported Zhou Shujun’s theory.

2.4 The time for embryo rescue

In this research, the embryo rescue was conducted on the 60 d after pollination, which had significantly higher germination rate than that on the 65 d, 70 d, 75 d after pollination. It may be because dormancy factors in the ovule increased (such as abscisic acid increased) with the delay of time for embryo rescue, or the ability of endosperm to supply embryo growth nutrients got weaker and weaker, the embryo became more and more rigid, thus resulting in the reduction of embryo germination ability.

3 Materials and Methods

3.1 Experimental materials

Lily varieties used in the experiment included Robina, Competition, Muscat, Myth, Palazzo, Table dance and Tiber, Siberia, Cobra, Justina. They were purchased from Beijing Plant Horticulture Co., Ltd. The first six varieties were OT lily, used as female parents; the latter four were O lily, used as male parents.

3.2 Observation of chromosome ploidy and karyotype

In the morning, root tips of 0.5-1.0 cm in length were taken from box planting lilies, rinsed with distilled water and pretreated with 0.04% colchicine for 24 h; then placed in Carnoy’s Fluid (absolute ethanol: glacial acetic acid = 3:1) for 24 h. Later, the lily root tips were added to 1 mol/L HCl, placed in water bath at 60°C and dissociated for 15 min; put some materials on the glass slide and covered it after Carbol fuchsin dyeing; used microscope to observe cells at the chromosome metaphase, took photographs and did measurement analysis and finally chromosome counting and karyotype analysis were completed based on Li Maoxue’s standard (Li and Zhang, 1996).

3.3 Cross pollination and embryo rescue

Male and female seedballs were cultivated in 60 cm (length) × 40 cm (width) × 18 cm (height) bulb export boxes by the way of box planting, and the cultivation medium was a mixture of import peat and perlite at 3:1 ratio. As the male grew longer than the female, to meet the flowering date, the male would be planted earlier than the female and both would be planted in batches.

3.4 Cross pollination

When the maternal flower buds began to show color, emasculation and bagging started. When there were mass of mucus in the maternal stigma, pollination started. Normal stigma pollination method (NSM) and cut style pollination method (CSM) were used. Normal stigma pollination method: the paternal pollen is directly pollinated to the maternal stigma; Cut style pollination method: the stigma and style are cut off by a knife, remained only 5 mm and smeared with pollen, immediately bagging and hanging labels after pollination.

There were 34 cross combinations (Table 4). 24 OT(♀)×O(♂) positive-cross combinations (see combination A-X) used NSM, besides combination H, J, N, R also used CSM. The remaining 10 combinations were control groups, including 4 O(♀)×OT(♂) back-cross combinations (see CK1-CK4) and 6 OT(♀)×OT(♂) self-cross combinations (see CK5-CK10). Each combination had 20 flowers for hybridization.

3.5 Hybrid embryo rescue

The embryo rescue was conducted respectively on the 60 d, 65 d, 70 d, 75 d after pollination. The method was shown below: take off the fruit, wash its surface with clear water, cut open the fruit along its ventral suture by a scalpel, and take out ovules. The ovules were surface sterilized with 75% alcohol and 1% NaClO for a short time, after sterile water rinsing, the ovules were directly inoculated into the embryo rescue medium of “MS+1.0 mg/L BA+0.5 mg/L NAA+30 g/L sucrose” for culturing. The germination rate of ovules was recorded after 60 d of embryo culture.

Authors’ contributions

ZZQ and CX was the executor of the experimental design and research, at the same time completed the data analysis and the first draft writing. CJT and JYH participated in the experimental design and experiment results analysis. ZKZ was the designer and director of this research, to guide the experimental design, data analysis, thesis writing and modification. All authors read and approved the final manuscript.

Acknowledgments

This research is jointly supported by Beijing science and technology program-2017 innovation base cultivation and development special (Z171100002217036), the Beijing Municipal Education Commission (CEFF-PXM2018_014207_000024), Quality construction of talent training-high level personnel cross training-real training plan (PXM2018–014207-000022)，Beijing Natural Science Fund-Municipal Education Commission jointly funded projects (KZ201810020029) and Beijing Laboratory of Urban and Rural Eco-environment Project.

References

Chen Q., Mu D., Yi M.F., Ming J., and Liu C., 2007, Effects of different pollination methods on bypassing pre-fertilization barriers in lily breeding, Zhongguo Nongye Daxue Xuebao (Journal of China Agricultural University), 12(4): 35-40

Li M.X., and Zhang Z.P., eds., 1996, Crop chromosome and its research techniques, China Agriculture Press, Beijing, China, pp. 1-37

Maheshwari P., ed., 1971, An introduction to the embryology of angiosperms, McGraw-Hill Book Company, New York, US, pp. 453

Shao Y., Wen W.H., Cui J.T., Jia Y.H., and Zhang K.Z., 2014, Study on the pollination methods and embryo rescue techniques of the crosses among Lilium oriental hybrids, Xibei Zhiwu Xuebao (Acta Botanica Boreali-Occidentalia Sinica), 34(6): 1119-1124

Van R.P.M., Keijzer C.J., Wilms H.J., and Keijzer C.J., eds., 1987, Pollen tube growth in an interspecific cross between two Lilium species, (In): Williams E.G., Knox R.B., and Irvine D., (eds.), Pollination: Proceeding of the International Symposium, Melbourne, Australia, pp. 240-241

Van T.J.M., Straathof T.P., Bino R. J., and Aam K., 1988, Effect of three pollination methods on embryo development and seedset in intra- and interspecific crosses between seven Lilium species, Plant Reproduction, 1(2): 119-123

https://doi.org/10.1007/BF00189271

Van T.J.M., Van D.A., Chi H.S., Lim, K.B., Villemoes S., and Kronenburg B.C.E.V., 2000, Breakthroughs in interspecific hybridization of lily, Acta Horticulturae, 508: 83-90

Yuan G.L., Xu P., Gong H.X., Zhou F.H., and Zhou S.J., 2013, Lily (Lilium) introgression breeding using triploid cultivars as maternal parents, Nongye Shengwu Jishu Xuebao (Journal of Agricultural Biotechnology), 21(11): 1321-1327

Zhang X.X., Ren G.L, Li K.H., Zhou G.X., and Zhou S.J., 2012, Genomic variation of new cultivars selected from distant hybridization in Lilium, Plant Breeding, 131(1): 227-230

https://doi.org/10.1111/j.1439-0523.2011.01906.x

Zhou H.G., Zhang X.L., Zhou Y., and Ning Y.F., 2000, A study on the cross compatibility of Lilium cultivars, Jiyinzuxue Yu Yingyong Shengwuxue (Genomics and Applied Biology), 19(4): 223-227

Zhou S.J., Li K.H., and Zhou G.X., 2012, Analysis of endosperm development of allotriploid×diploid/tetraploid crosses in Lilium, Euphytica, 184(3): 401-412

https://doi.org/10.1007/s10681-011-0609-5

Zhou S.J., 2014, Technical ways of breeding modernlily (Lilium) cultivars and the mechanisms of the special phenomena of their hybridizations, Nongye Shengwu Jishu Xuebao (Journal of Agricultural Biotechnology), 22 (10): 1189-1194

Zhou S.J., Ramanna M.S., Visser R.G.F., and Van T.J.M., 2008, Genome composition of triploid lily cultivars derived from sexual polyploidization of Longiflorum×Asiatic hybrids (Lilium), Euphytica, 160: 207-215

https://doi.org/10.1007/s10681-007-9538-8

Zhou S.J., Yuan G.L., Xu P., and Gong H., 2014, Study on lily introgression breeding using allotriploids as maternal parents in interploid hybridizations, Breeding Science, 64(1): 97-102

https://doi.org/10.1270/jsbbs.64.97

PMid:24987295

PMCid:PMC4031115

Zhou S.J., Zhou G.X., and Li K.H., 2011, Euploid endosperm of triploid×diploid/tetraploide crosses results in aneuploid embryo survival in Lilium, Hortscience, 46(4): 558-562

http://hortsci.ashspublications.org/content/46/4/558.full.pdf+html

Zhu M.L., Liu Q.Q., and Dai S.L., 2011, Karyotype analysis of 38 large-flowered Chrysanthemum cultivars from China, Zhiwu Xuebao (Bulletin of Botany), 46(4): 447-455