Introduction

Polyploidization is widespread in plants and has played a major role in the evolution and diversification of the plant kingdom. Unreduced gametes are an important tool for polyploidization and the creation of genetic variation in plant breeding (Dewitte et al. 2009, 2010a, 2010b). Polyploidy is of great value to breeders (Dewitte et al. 2010a; Gallo et al. 2007). But among the rich germplasm resources of Chinese jujube (Z. jujuba Mill.), ‘Zanhuangdazao’ is the unique natural triploid identified up to now, which has the valued attributes of big fruit, high yield and good adaptation. Most of the cultivars of ber (Z. mauritiana Lam.), a close-related species of Chinese jujube, are tetraploid showing similar advantages. Considering the above-mentioned, polyploid induction of Chinese jujube should have attractive prospect.

Polyploidization may be asexual through somatic chromosome doubling and sexual through the formation of 2n gametes following both intra- and inter-ploidy matings (Ton et al. 1977). Asexual polyploidization (traditional chromosome duplication) is performed through the use of harmful mitosis inhibitors like colchicine, oryzalin and trifluralin (Hancock 1997; Wan et al. 1991). The major drawback of this technique is that it largely disables the recombination between chromosomes of different parents in allopolyploids. As a consequence, this form of polyploidization offers limited possibilities to create genetic variability. Sexual polyploidization is far more efficient than somatic chromosome doubling. Sexual polyploidization effected by functioning of 2n gametes in the parental species or in their hybrids has been got in ornamental plants, grain crops, potato, fruit-bearing trees and some other plants (Akutsu et al. 2007; Barba-Gonzalez et al. 2004, 2005; Carputo and Barone 2005; Carputo et al. 2003; Hayashi et al. 2009; Jauhar 2003; Lyrene et al. 2003; Ramanna et al. 2003; Sugiura et al. 2000; Van Laere et al. 2009; Van Tuyl and Lim 2003; Veilleux 1985; Yanagi et al. 2010; Zhou et al. 2008).

Advantages of 2n pollen in polyploidization have been proposed. Firstly, usage of 2n pollen can conquer homogeneity increasing, severe recession and chimaeras occurrence resulted from asexual polyploidization. Secondly, 2n pollen can simply transfer advantage genes from wild species and part-cultivars into cultivars that can be used in breeding and production directly. Thirdly, 2n gametes can transmit the parental heterozygosity to the offspring. The first division restitution (FDR) and the second division restitution (SDR) gametes can transfer at least 75–80% and 33–35% heterozygosity respectively (Camadro and Peloquin 1980; Peloquin et al. 2008). FDR gametes are particularly valuable in plant sexual polyploidization breeding programs (Bretagnolle and Thompson 1995; Ramanna 1983). Fourthly, using 2n gametes can improve self-fruitfulness much as well as the pollen fertility of triploids. Fifthly, endosperm balance number (EBN) can be conquered by 2n gametes, most breeding experiences showed that usage of 2n gametes can conquer this barrier directly and shorten the period of breeding (Johnston et al. 1980). In tuber-bearing solanums, it was reported that EBN in combination with 2n gametes maintains the ploidy integrity of diploid ancestral species, while proving the flexibility for either unilateral or bilateral sexual polyploidization (Carputo et al. 2003). Sixthly, seedless germplasm can be obtained by using 2n gametes. In citrus breeding, 2n gametes have been successfully used for the production of seedless variety, triploid from 2n female gametes (Geraci et al. 1975).

Unreduced gametes are of importance for basic studies in cytogenetics as well as applied plant breeding and facilitate the production of new polyploidy species. According to the previous observation on 223 Chinese jujube cultivars, only four cultivars (‘Linglingzao’, ‘Yingluozao’, ‘Zanhuangchangzao’, ‘Zhongyangmuzao’) produce relative high percentage of 2n pollens and it was proposed that the diameter of 2n pollen was 1.5 times of n pollen (Peng 2008). The cytological basis for the formation of 2n pollen in Chinese jujube has not been clearly understood. The aim of this research was to elucidate the cytological mechanisms that lead to 2n pollen formation and provide new insights on utilization of 2n pollen in polyploidy breeding of Chinese jujube.

Materials and methods

Plant material

‘Linglingzao’ (2n = 2x = 24) which can produce relative high frequency of 2n pollen naturally was used for cytological analysis of meiosis. All the sampled trees were grown under regular management in the field.

Cytological analysis of meiosis

During May 15 to 22, 2009, flower buds were collected and fixed in ethanol:acetic acid (3:1) fixative for 24 h, then transferred into 70% ethanol and stored in refrigerator at 4°C. Before observation, anthers were dissected from the buds with a dissecting needle, then microsporocytes were prepared by squashing anthers with a nipper, and stained with propanoic acid-iron-hematoxylin-chloral hydrate (PIHCH). Chromosome number was determined at diakinesis and all phases of meiosis were observed with light microscope OLYMPUS. Photomicrographs were taken under binocular light microscope (×1000 magnification) with image-forming system Stk ‘MDC200(LY-HPCCD)’.

Observation of 2n pollen

Anthers from opening flowers were collected and squashed in a drop of 0.5% triphenyltetrazolium chloride (TTC) solution on a microscope slide. After a minimum of 1 h, the preparation was observed under a binocular light microscope (×100 magnification). Viable pollen grains were stained into full red, whereas aborted pollen grains were unstained. More than 5000 pollen grains were measured. Pollen which is 1.5 times of n pollen in diameter was regarded as 2n pollen.

Estimation of the frequency of 2n pollen

The number of dyads, triads, and tetrads were counted respectively. The number of observed dyads, triads, and tetrads during meiosis enabled an evaluation of the theoretical frequency of 2n pollen. Based on the rule that each dyad led to two unreduced pollen grains, each triad produced one unreduced pollen grain and two reduced pollen grains, and each tetrad gave rise to four reduced pollen grains, the frequency of 2n pollen (F) was estimated as below:

$$ {\text{F}} = \frac{2D + Tr}{2D + 3Tr + 4Te} \times 100\% $$

D = number of dyads, Tr = number of triads, Te = number of tetrads.

Statistical analysis

Correlation analysis was carried out with Statistical Product and Service Solutions version 16 (SPSS Inc, USA).

Results

Microsporogenesis

During microsporogenesis, chromosomes paired in 12 bivalents at diakinesis (Fig. 1a–d) and first meiosis was normal. However, abnormal spindle orientation (parallel and tripolar spindles) were observed at metaphase II (Fig. 1i, m, q). The frequencies of perpendicular, tripolar and parallel spindle were 72.93, 17.22, and 9.85% respectively (Table 1). While the frequencies of tetrad (Fig. 1g, h), triad (Fig. 1k, l) and dyad (Fig. 1s, t) were 80.99, 16.40, and 2.61% respectively at tetrad stage (Table 1). The correlation analysis showed that there was a significant association (r = 0.98, P < 0.01) between the frequency of tripolar spindle and that of triad (Table 2), which indicated that tripolar spindles resulted in triads. The correlation between parallel spindle and dyad and that between perpendicular spindle and tetrad were not significant, which indicated that not all the parallel spindles resulted in dyads formation.

Fig. 1
figure 1

Dyads and triads formation of Chinese jujube (Ziziphus jujuba Mill. ‘Linglingzao’). Bar 5 μm. a–d Diakinesis with 12 bivalents. e Normal perpendicular spindles. f Anaphase II with two sets of chromosomes being pulled to four poles. g–h Tetrads as a result of perpendicular spindles formation. i Tripolar spindles at Metaphase II. j–l Triads as a result of tripolar spindles formation. m Parallel spindle I with a distance between two sets of chromosomes. n Anaphase II with sets of chromosomes being separated. o Tetrads resulted from parallel spindle I. p Tetraspore. q Parallel spindle II (Fused spindle). r Tripolar, parallel and fused spindle. s – t Dyads following the formation of fused spindles

Full size image
Table 1 Frequency of different type spindles and sporads in ‘Linglingzao’
Full size table
Table 2 Correlation between the frequencies of different types of spindles and sporads
Full size table

Actually, two types of parallel spindles were observed. For parallel spindle I (Fig. 1m), a long distance between the two sets of chromosomes led to tetrads formation (Fig. 1o). Differed with the regular perpendicular tetrads, the tetrads resulted from the parallel spindle I were arranged at the same division plane. For parallel spindle II (fused spindle), the two sets of chromosomes were very close, even integrated with each other (Fig. 1q). Fused spindles were similar to spindles at metaphase I. However, fused spindles could be distinguished based on the synchronization of meiosis (Fig. 1r). The correlation analysis showed a significant association (r = 0.84, P < 0.05) between the frequency of parallel spindle II and dyads (data not show). This result further indicated that fused parallel spindles (Fig. 1q) led to dyads formation (Fig. 1s, t).

2n pollen formation

Each triad gave rise to one 2n microspore and two n microspores, while each dyad produced two 2n microspores. 2n microspore formed 2n pollen grain. The frequencies of 2n pollen from different types of sporads were estimated (Table 3). The average frequency of 2n pollen from both dyad and triad was estimated as 5.71%, which was not significantly different with the observed one, 6.15% (Table 3). It indicated that the 2n pollens in ‘Linglingzao’ are resulted from the formation of triads and dyads at tetrad stage which originated from the abnormal spindle orientation at metaphase II.

Table 3 Frequency of 2n pollen observed and estimated
Full size table

Discussion

The formation of 2n gametes is a common phenomenon in the plant kingdom. 2n gametes may result from a variety of different cytological abnormalities (Bretagnolle and Thompson 1995; Harlan and De Wet 1975; Veilleux 1985).

Five main cytological mechanisms of 2n gamets formation were summarized: (i) premeiotic chromosome doubling during the process from mitosis to meiosis, disturbances appeared and resulted in chromosomes of archesporial cell being doubled followed by 2n gametes formation; (ii) meiosis disturbances, most 2n gametes were formed by this reason. Meiosis abnormalities led to 2n gametes of different types, second division with crossing-over (SDR-CO), first division with crossing-over (FDR-CO), first division restitution without crossing-over (FDR-NCO), transmitting ±40, ±80, 100% of the parental heterozygosity to the offspring, respectively (Crespel et al. 2002; El Mokadem et al. 2002); (iii) abnormal cytokinesis, in Paspalum, 52 accessions presented failure of cytokinesis during first or second meiosis and resulted in dyads and triads formation respectively, followed by 2n pollen formation (Pagliarini et al. 1999). In Brachiaria, it was reported that the failure of cytokinesis at metaphase I and total absence of cytokinesis resulted in unreduced pollen formation (Gallo et al. 2007); (iv) desynaptic mutants, desynaptic mutants of diploid potato (2n = 2x = 24) were found to produce a fairly high frequency of fertile 2n pollen grains (Ramanna 1983); (v) abnormal spindles of meiosis, this often happened in simultaneous dicotyledonous plants. Parallel or tripolar spindles during metaphase II of microsporogenesis were responsible for the production of 2n pollen. Triads formation resulted in 2n pollen formation in Brazilian gingseng (Taschetto and Pagliarini 2003) and rose (Crespel et al. 2006). In Populus tomentosa Carr, parallel, fused and tripolar spindles were considered to be leading to 2n pollen formation (Zhang and Kang 2010; Zhang et al. 2009). Parallel and tripolar spindles have been reported leading to 2n pollen formation in the wild potato Solanum okadae (Camadro et al. 2008).

Unreduced gametes formation of Chinese jujube was due to the abnormal configuration of spindles according to this research. Detailed studies have been made on pollen mother cells (PMCs) at all stages of meiosis to determine whether or not meiotic disturbances were responsible for 2n gametes formation. During the observation of microsporogenesis of Chinese jujube, chromosomes paired in 12 bivalents at diakinesis and the first meiotic division was normal, whereas the second division was characterized by frequent abnormal spindle orientation (parallel and tripolar spindles), as have been reported in some plants (Bretagnolle and Thompson 1995; Camadro et al. 2008; Veilleux 1985; Xu et al. 2008). However, the occurrence of parallel spindles at the second division does not necessarily mean that 2n pollen has been formed (Ramanna 1979; Tavoletti et al. 1991). The ‘fused spindle’ at metaphase II has been associated with the formation of dyads (Ramanna 1983; Veilleux et al. 1982; Xu et al. 2008). In this study, a significant association correlation (r = 0.84, P < 0.05) between fused spindles and dyads was found. This meant only parallel spindle II (fused spindle) can form dyads thereby result in 2n gametes, while parallel spindles I (with a long distance between two sets of chromosomes) seemed to be leading tetrad formation followed by n gametes in Chinese jujube.

Most researches indicated that formation of 2n gamete was controlled by genes (Calderini and Mariani 1997; Katsiotis and Forsberg 1995). d’Erfurth et al. (2008) isolated and characterized a gene (AtPS1) involved in the abnormal orientation of spindles at meiosis II that controlled 2n gamete formation apparently in Arabidopsis thaliana. It was reported that the mechanism of 2n pollen formation of Solanum is under control of Mendelian recessive ps (Camadro et al. 1980). Jumbo pollen formation in alfalfa is dependent on the homozygous recessive condition of the gene jp that causes complete failure of post-meiotic cytokinesis which results in 100% of 4x pollen (Tavoletti et al. 2000). Desynapsis was found to be controlled by a single recessive gene ds (Ramanna 1983). Although the production of 2n gametes is under genetic control, it can be also affected by environmental conditions. The variation of 2n pollen production could be related to environmental fluctuations (Crespel et al. 2006).

Depending on the cytological mechanisms involved in the production of 2n gametes in Chinese jujube, such gametes are generally categorized as being equivalent to FDR. FDR gametes could transmit at least 75% of parental genotype to their progenies. Further research is needed to reveal the actual level of heterozygosity of the unreduced gametes and their contributions to the progeny.