Introduction

Bread wheat (Triticum aestivum) is one of the world’s three leading cereal crop species. Intensive selection over many decades has eroded the level of genetic variation available to breeders (Law 1993; Reynolds et al. 2012). Its wild relatives have therefore become an important genetic reservoir, and it is becoming increasingly urgent to broaden wheat’s genetic base by alien introgression. A favored relative in this context is the species Thinopyrum intermedium (intermediate wheatgrass), which harbors genes enabling it to survive high levels of salinity, drought and low temperature, as well as genes ensuring resistance against various fungal and viral diseases (Hohmann et al. 1996; Li and Wang 2009). Sexual hybrids between wheat and Th. intermedium have provided the starting point for a number of attempts to introgress some of these genes into wheat (Hohmann et al. 1996; Ayala-Navarrete et al. 2009; Li and Wang 2009; Georgieva et al. 2011). Numerous partial amphiploid, chromosome addition and substitution lines were obtained and widely used for improvement of wheat against diseases via normal sexual hybridization and chromosome engineering (Friebe et al. 1993; Li and Wang 2009; Georgieva et al. 2011). However, yield or quality penalty often occurs in lines that carry large chromosome fragments from Thinopyrum (Brown 2002; Li and Wang 2009), which prevents the use of certain genes especially that rarely exist in wheat. The development of an alternative introgression method is therefore important, especially combined with the chromosome engineering techniques such as irradiation to shorten the alien chromosome segments.

The salient feature of asymmetric somatic hybridization is that it tends to transfer fragments of the donor genome while retaining the majority of the recipient’s. Furthermore, producing this sort of hybrid is both easier and more rapid than producing sexual hybrids. Over recent years, we have demonstrated the potential of asymmetric somatic hybridization, particularly using a combination of wheat with UV-irradiated tall wheatgrass (Th. elongatum). We have also fused wheat protoplasts with those of Russian wildrye, oat and maize (Zhou et al. 2001a; Xiang et al. 2003, 2010; Xu et al. 2003). One of the derivatives of the wheat/tall wheatgrass hybrid has been released in China as a cultivar (ShanRong No. 3), selected for its expression of an enhanced level of abiotic stress tolerance, presumably inherited from the donor parent (Xia et al. 2003; Liu et al. 2007, 2012; Shan et al. 2008). In this paper, we describe fusion products obtained from the combination of wheat cv. JN177 and the hexaploid grass Th. intermedium.

Materials and methods

Protoplast preparation and fusion product culture

Protoplasts of cv. JN177 were prepared from both embryogenic calli (#176, a regenerable line) and suspension cells (#Cha9, a vigorously dividing but non-regenerable line) (Cheng and Xia 2004; Li et al. 2004). #176 is maintained on a solid medium, while #Cha9 is maintained in a liquid medium. Both media are formulated according to Murashige and Skoog (1962), and supplemented with 2 mg/L 2, 4-dichlorophenoxyacetic acid (Xia and Chen 1996). In preparation for protoplast fusion, #176 protoplasts were recovered after a 6–7 day period of subculture on the solid medium, and the #Cha9 ones after a 3–4 day period of subculture in the liquid medium. Th. intermedium protoplasts were isolated from regenerable calli. Protoplast isolation was performed as described by Xu et al. (2003). #176 and #Cha9 protoplasts were mixed in a 1:1 ratio. The Th. intermedium protoplasts were spread as a monolayer on a 3 cm petri dish, irradiated with 300 uW/m2 UV light for either 0 s (treatment I), 30 s (treatment II) or 60 s (treatment III), mixed with the wheat protoplasts in a 1:1 ratio and fused as described by Xia and Chen (1996). When the regenerated calli had grown to a diameter of 2–5 mm, they were transferred to a proliferation medium, and later to a differentiation medium (Xia and Chen 1996). Regenerated plantlets were transplanted to soil and grown in a greenhouse until maturity.

Identification of peroxidase variation in the fusion products

0.5 g/mL fresh callus was homogenized in 1 M Tris–HCl (pH 8.3), centrifuged (12,000×g, 10 min) and the supernatant electrophoresed through a polyacrylamide gel (4 % stacking and a 10 % separating gel), which was stained following Hu and Wan (1985).

RAPD genotyping

Genomic DNA was extracted from calli following Doyle and Doyle (1990), and used in PCRs primed with one of the decamer primers OPA-01, -06, -08, -17, -19, OPF-03, -05, -12, OPH-04, -20, OPI-10 or OPM-04 (Operon Technology, Alameda, CA). The amplification regime and subsequent visualization of amplicons were as given by Cheng et al. (2004).

Mitotic chromosome spreads and genomic in situ hybridization (GISH)

Calli and the seedling root tips of hybrid and each of the parents were immersed in ice water for about 24 h, then fixed in 3:1 ethanol:acetic acid for about a week. Chromosome counts were obtained using the conventional Feulgen method. For the purposes of GISH, the probe was Th. intermedium genomic DNA labeled by DIG-nick translation (Roche catalog No. 11745816910), and blocking DNA was provided by DNA extracted from #Cha9 and #176. The probe to blocking DNA ratio was between 1:100 and 1:120. The GISH protocol used followed Xiang et al. (2003).

Results

Putative hybrid fusion products

Granular calli were induced from each of the three combinations after about 4 weeks of culture in liquid P5 medium (Xia and Chen 1996) in the dark at 25 °C (Fig. 1b–e). In all, 124 small calli were recovered (Table 1). The calli increased in size once they had been transferred onto proliferating medium (Fig. 1f–i) and plantlets formed following the second transfer to the regeneration medium (Fig. 1j–m). The largest number of calli (64) was obtained from combination I (no exposure to UV). Of these, nine regenerated green plantlets, characterized by soft leaves and a fasciculate growth habit (Fig. 1j, k). Combination II (30 s UV irradiation) produced 49 calli, of which just four were regenerable. Upon transfer to the differentiation medium, one of these four calli (II-3) produced green plantlets with strong roots, and the resulting plants showed a wheat-like phenotype (Fig. 1l, m). A dozen putative hybrid plants recovered from calli I-34 and II-3 were potted into soil and grown in a greenhouse; only those regenerated from II-3 were self-fertile (Fig. 1n–p), and the morphology of its seeds was wheat-like (Fig. 1l–q). The 11 calli recovered from combination III (60 s UV irradiation) grew rapidly, but were non-regenerable. Fusion of either #176 or #Cha9 with Th. intermedium formed no calli, and stopped growing after a few rounds of cell divisions. Unfused parental protoplasts formed cell clusters, but no calli were generated (Fig. 1a).

Fig. 1
figure 1

Morphology of the somatic hybrid products, the recipient and the donor calli at 30 days after fusion. a Mixed #Cha9, #176 and Th. intermedium protoplasts, b, c combination I, d combination II, e combination III. Calli at 60 days after fusion. Regenerants f, j formed from callus I-34, g, k from I-54, h, l from II-3, i, m from II-6. n–p Spikelets formed by the regenerant from callus II-3. q Grain set by the regenerant from callus II-3 and its parents

Full size image

Genotypic analysis

Peroxidase profiles

To characterize the hybrid nature of the regenerated calli, peroxidase assay was conducted. All the 124 calli from combination I, II and III showed the similar peroxidase patterns to that of parent wheat. Among them, 42, 28 and 4 calli of combination I, II and III still include specific bands of Th. intermedium, which were subsequently identified as hybrids. As shown in Fig. 2, the profiles of calli from combination I (e.g. I-2, -3, -10, -14, -24, -34 and -52), combination II (e.g. II-1, -14) and combination III (III-2, -6) included the parent bands. This result showed that the calli had retained most, if not all, of the wheat genome, and that parts of the donor genome had been introgressed in the fusion derivatives.

Fig. 2
figure 2

Peroxidase profiles of Calli formed from a combination I, b combinations II and III. Zh: Th. intermedium. Right-doubled arrow recipient isozyme, right arrow donor isozyme

Full size image

RAPD profiles

All the presumptive hybrid calli identified by the isozyme analysis were analyzed by the RAPD profile further to confirm their hybrid nature. Overall, 35 of the combination I, 24 of the combination II and 4 of the combination III calli contained DNA derived from Th. intermedium, and a few carried RAPD fragments not present in either parent (Table 1). The frequency of donor fragments decreased as the UV dosage increased (Table 1). The RAPD profiles of the regenerants included not only fragments derived from Th. intermedium and both #Cha9 and #176, but also some not present in the profiles of any of the parents (e.g., calli I-2, -10, -14, -24 and -34, see Fig. 3a, c). Template extracted from different #176 calli or samples of #Cha9 cells also produced variable RAPD profiles due to the somaclonal variations (Fig. 3). The OPA-06 amplicons of calli I-10 and -28 each included fragments specific to #Cha9, while the I-2, -14 and -24 amplicons included a fragment specific to #176; meanwhile the I-34 amplicon included both #Cha9 and #176 specific fragments (Fig. 3a). The results implied that all three protoplasts were involved in some of the fusion events. Overall, the proportion of the RAPD fragments assignable to the wheat parents varied from 75.0 to 89.5 %, to the Th. intermedium parent from 2.6 to 15.9 %, and the remainder related to fragments not present in any of the three parent’s profiles (Table 2). Th. intermedium and non-parental fragments were more frequent in combination I calli than in those from combinations II or III. Clone I-34 carried the largest number of Th. intermedium fragments: on the basis of its RAPD profiles, about 16 % of its genome was estimated to have been inherited from Th. intermedium.

Table 1 The success rate of regenerating plants from somatic hybridization
Full size table
Fig. 3
figure 3

RAPD analysis of putative hybrids and their parents. Profiles resulting from priming with a OPA-06, b OPA-08, c OPA-19, d OPF-05. I-2, -10, -14, -24, -28, -34, -54, and -62 derived from combination I, II-1, -3, -4, and -6 from combination II, III-1, -3, and -11 from combination III. M: lambda DNA digested with EcoRI and HindIII. The right-pointing arrows indicate fragments derived from each parent, arrowheads identify fragments not present in the parental profiles

Full size image
Table 2 RAPD analysis of the fusion products
Full size table

Chromosome number and GISH karyotype

The somatic chromosome number of #Cha9, #176 and Th. intermedium protoplasts was, respectively, 24, 34 and 42 (Fig. 4a–c), while that of the hybrids from combination I, II and III regenerants varied from 40 to 60 (Fig. 4d–i; Table 3). Chromosome fragments were common in the fusion products’ mitotic cells (Fig. 4d, h). One or two small chromosomes and/or chromosome fragments were present in 2.6 % cells of combination I, 7.01 % of combination II and 8.8 % of combination III. Multi-centromere chromosomes were seen in the cells of some of the regenerants (Fig. 4f). The GISH treated somatic chromosomes of regenerants from calli I-34 and II-3 were compared to those of the parents in Fig. 5. No intact Th. intermedium chromosomes were identified, but it was possible to detect many small segments incorporated into large wheat chromosomes (Fig. 5c–f), consistent with the notion that the somatic hybridization process was successful in introgressing only a limited amount of the Th. intermedium genome.

Fig. 4
figure 4

Karyotypic analysis. Arrow indicates chromosome fragments, arrowhead indicates polycentric chromosomes. a #176, 2n = 34, b #Cha9, 2n = 24, c Th. intermedium, 2n = 42, regenerant from d I-2, 2n = 48, e I-34, 2n = 52, f II-3, 2n = 50, g II-6, 2n = 54, and Calli from h III-1, 2n = 44, i III-11, 2n = 45

Full size image
Table 3 Variation in somatic chromosome number among the parents and fusion products
Full size table
Fig. 5
figure 5

GISH analysis using genomic DNA of Th. intermedium as probe a Wheat cv. JN177, b Th. intermedium, regenerants from c, d I-34, e, f II-3. Arrow indicates chromosome fragments of Th. intermedium

Full size image

Discussion

Until now, plant regeneration from somatic hybridization calli has been difficult to achieve, and most regenerants have proven to be sterile and/or morphologically abnormal (Gamborg and Holl 1977; Bauer-Weston et al. 1993; Fahleson and Glimelius 1999; Xia 2009; Eeckhaut et al. 2013). The creation of asymmetric hybrids offers a potential way to avoid these problems. The extent of the asymmetry is a function of the severity of the irradiation dose, the phylogenetic distance between the parental species, the chromosome number of each parent, and differences in the cell cycle time of each of the parental materials (Dudits et al. 1987; Sears 1993; Nakanoa et al. 2006; Xia 2009). We have explored the use of a three cell system, based on two distinct types of recipient cells, both obtained from the same cultivar; one was a rapidly dividing but non-regenerable cell line (#cha9) and the other grew slowly but was regenerable (#176). Fusion of donor cells with #cha9 or #176 individually failed to regenerate green plants because of significant loss of chromosomes from the #Cha 9 (2n = 24) cell line, and the chromosome elimination of #176 (2n = 34) (Li et al. 2004; Xiang et al. 2010). Fusing these two different protoplasts with UV-irradiated donor protoplasts has been an effective strategy for the production of green regenerants from a range of donors, including oat (Xiang et al. 2003, 2010), maize (Xu et al. 2003), Russian wildrye (Li et al. 2004), Italian ryegrass (Cheng and Xia 2004) and foxtail millet (Cheng et al. 2004; Xiang et al. 2004), but no hybrid progenies were produced in these combinations. In the present study, regeneration was possible from about 10 % (13/124) of the calli, and the fertile hybrid plants were obtained between the fusion of wheat and Th. intermedium using this fusion system. The regenerants were genotypically largely wheat, with a small number of donor chromosome fragments incorporated into the largely wheat chromosomes (Fig. 5). Both the recipient cells lines were highly aneuploid (#Cha9 contained on average 24 chromosomes and #176 34 chromosomes, compared to the euploid number of 42, see Table 3), but in combination, they conferred both the ability to grow vigorously and to regenerate. The implication is that different chromosomes were missing in each of the two cell lines, so that together they provided a full complement.

Chromosome elimination frequently accompanies the fusion of highly differentiated genomes (Feldman et al. 1997; Kashkush et al. 2002; Xia 2009). Regenerants typically carry fewer chromosomes than predicted from the sum of each parent’s complement (Kisaka et al. 1997; Li et al. 2004; Xiang et al. 2010). The factors which probably govern the extent of chromosome elimination have been discussed by Xia (2009). In the present experiments, all the regenerants, whether derived from symmetric or asymmetric fusion, exhibited a degree of chromosome elimination: in three combinations, their somatic chromosome number was in the range 40–60, while the sum of the three parental complements was 100 (Table 3). GISH analysis demonstrated that the regenerants’ genomes were a mosaic of wheat and Th. intermedium chromosomes, in some cases involving small donor fragments inserted interstitially within a largely wheat chromosome, and in others involving the translocation of quite large donor fragments (Fig. 5). Whether or not the fusion was asymmetric, the overall chromosome complement was mostly inherited from one and/or the other recipient line. However, chromosome elimination was clearly not restricted to the donor genome, since the total wheat complement contributed by the two wheat parents was 58, a number far greater than the observed range of somatic chromosome number in the hybrid derivatives. UV irradiation of the donor cells encouraged the elimination of donor chromosomes. UV irradiation is known to fractionate chromosomes (Hall et al. 1992), and it is clear that increasing the UV dosage reduces the extent of donor material transferred into the recipient (Xia et al. 2003; Cheng et al. 2004; Cui et al. 2009). The overall somatic chromosome number was rather insensitive to the dosage of UV, but chromosome fragments were more frequently induced as the dosage was raised (Figs. 3, 5; Table 2). The suggestion is that a balanced genome is necessary to achieve regeneration and subsequent self-fertility.

Phylogenetic distance is known to influence regeneration capacity and fertility. In the combination wheat/oat, it was only possible to regenerate albino plants (Xiang et al. 2003). In other wide combinations (wheat with either maize (Xu et al. 2003), Italian ryegrass (Cheng and Xia 2004), foxtail millet (Chen et al. 2004; Xiang et al. 2004) and Russian wildrye (Li et al. 2004), green plants were regenerated, but the plants were all sterile. So far, fertile somatic hybrid plants have only been successfully produced from the combinations wheat/Haynaldia villosa (Zhou et al. 2001b) and wheat/Th. elongatum (Xia et al. 2003; Cheng et al. 2004; Cui et al. 2009), both of these donors being, like wheat, Triticeae species. In the wheat/Th. elongatum case, symmetric hybridization produced tall, perennial plants with an appearance quite similar to that of the donor. However, following the asymmetric route, some derivatives exhibited a wheat-like phenotype and others an intermediate one (Xia et al. 2003; Cheng et al. 2004). Cytological analysis of these regenerants has confirmed that the genome of the wheat-like lines had been predominantly inherited from the wheat parent (Xia et al. 2003), whereas in the wheatgrass-like ones, only a few wheat chromosome fragments were evident (Cui et al. 2009). The former type is similar, in terms of both phenotype and genome constitution, as pertained for the fertile derivative line II-3 described here (Fig. 5).

In conclusion, besides two wheat cell lines provided a full complement, the phenogenetic relationship of wheat with donor cereal was another key factor to influence the hybrid fertile. We provide a possible route to transfer segments of Th. intermedium chromosome(s) into wheat background and produce fertile derivatives using a three cell fusion system. The Th. intermedium genotype includes a number of potentially useful agronomic traits and has proven to be a valuable source for resistance to various diseases in wheat (Li and Wang 2009; Nevo and Chen 2010). The materials obtained from somatic hybridization are potentially interesting as sources of genetic variation of relevance to wheat improvement. The priority now will be to identify what traits have been successfully transferred and to monitor the expression of donor genes in the hybrid plants.