Development and regeneration of wheat-rice hybrid zygotes produced by in vitro fertilization system.

Hybridization plays a decisive role in the evolution and diversification of angiosperms. However, the mechanisms of wide hybridization remain open because pre- and post-fertilization barriers limit the production and development of inter-subfamily/intergeneric zygotes, respectively. We examined hybridization between wheat and rice using an in vitro fertilization (IVF) system to bypass these barriers. Several gamete combinations of allopolyploid wheat-rice hybrid zygotes were successfully produced, and the developmental profiles of hybrid zygotes were analyzed. Hybrid zygotes derived from one rice egg cell and one wheat sperm cell ceased at the multicellular embryo-like structure stage. This developmental barrier was overcome by adding one wheat egg cell to the wheat-rice hybrid zygote. In the reciprocal combination, one wheat egg and one rice sperm cell, the resulting hybrid zygotes failed to divide. However, doubling the dosage of rice sperm cell allowed the hybrid zygotes to develop into plantlets. Rice chromosomes appeared to be progressively eliminated during the early developmental stage of these hybrid embryos, and c. 20% of regenerated plants showed abnormal morphology. These results suggest that hybrid breakdown can be overcome through optimization of gamete combinations, and the present hybrid will provide a new horizon for utilization of inter-subfamily genetic resources.

Introduction

Hybridization and polyploidization (whole‐genome duplication) have been critical to the evolution and diversification of plants and animals (Hilu, 1993; Rieseberg, 1995; Mallet, 2005). Hybridization between two different genera or subfamilies is referred to as intergeneric or inter‐subfamily wide hybridization, respectively. Polyploid individuals possess two or more sets of related chromosomes, and polyploidization is a common phenomenon in angiosperms (Masterson, 1994; Blanc & Wolfe, 2004; Cui et al., 2006; Wood et al., 2009). There are two distinct types of polyploids: autopolyploids and allopolyploids (Kihara & Ono, 1926; Ramsey & Schemske, 1998). Allopolyploids are produced by hybridization followed by chromosome doubling or fusion of unreduced gametes of hybrid plants, and the allopolyploidization appears to be a primary route for polyploidization because the majority of polyploids in angiosperms (> 75%) are allopolyploids (Grant, 1981; Brochmann et al., 2004).

When gamete fusion occurs within a species, the gamete nuclei unite via karyogamy, and the parental genomes in the zygotic nucleus function synergistically to achieve proper zygotic development and embryogenesis (Ohnishi et al., 2014; Toda et al., 2018). When the genomes of two different species are fused, hybridization is initiated. The effects of genome hybridization often arise during the development of hybrid zygotes, resulting in compatible or incompatible embryos. Compatibility between genomes during zygotic embryogenesis is determined by the extent of interspecific conflict arising from genetic and epigenetic reorganization, such as the balance of dominant and recessive alleles or subgenomes, cytonuclear interaction, regeneration of gene silencing (imprinting), and remodeling of chromatin status, etc. (Josefsson et al., 2006; Burton et al., 2013; Bird et al., 2018).

In general, investigations for the hybridization mechanisms have been conducted using hybrid embryos/endosperms that were produced by cross‐pollination between different species or genera. However, cross‐pollination between interspecific or inter‐subfamily pistils and pollen often fails as a result of pre‐fertilization barriers, such as pollen germination failure and lack of pollen tube growth (Morgan et al., 2010). These barriers have impeded the analyses of hybridization mechanisms. In addition, even when successful fertilization occurs from interspecific/inter‐subfamily pollination, embryo rescue is principally essential because the endosperm is highly sensitive to differences in the parental genomes, and the abortion of hybrid endosperm results in the developmental arrest of the hybrid embryo within the seed (Haig & Westoby, 1991).

The in vitro fertilization (IVF) system using isolated gametes is a powerful technique for investigating the mechanism in wide hybridization because the IVF system bypasses pre‐fertilization barriers by fusing gametes isolated from distant plant species (Kranz et al.,1995; Kranz, 1999; Wang et al., 2006; Maryenti et al., 2019; Rahman et al., 2019) and acts as a mediator in monitoring and analyzing developmental profiles of hybrid zygotes without the influence of endosperm. The IVF system involves a combination of three basic microtechniques: isolating and selecting male and female gametes; fusing pairs of gametes to produce zygotes; and culturing zygotes into plantlets (Kranz, 1999). These procedures have been established in rice (Uchiumi et al., 2007), wheat (Maryenti et al., 2019) and maize (Kranz & Lorz, 1993) which, concomitantly, offer the possibility of artificially generating inter‐subfamily zygotes of these three crops with desirable combinations. Chromosomal analyses have been conducted on combinations of wheat pistils (female) and maize pollen (male) cross‐pollinations (Laurie & Bennet, 1988; Zhang et al., 1996), and these studies showed that maize chromosomes are rapidly eliminated during embryogenesis, leading to the haploidization of wheat chromosomes. Genomic combinations derived from wide hybridization between intergeneric or inter‐subfamily gametes are typically unstable, and partial or complete elimination of one of the parental sequences occurs (Kasha & Kao, 1970; Laurie & Bennet, 1989; Sanei et al., 2011; Ishii et al., 2016). The complete elimination of one parental chromosome results in a haploid embryo and plant. In contrast to wheat female × maize male combination, analyses of the reciprocal combination (IVF fusion between maize female and wheat male gametes) found that hybrid zygotes developed into multicellular embryo‐like structures, but then arrested (Kranz et al., 1995), suggesting that the combination of maize egg and wheat sperm cells is incompatible.

In comparison with the extensive body of work on wheat–maize hybridization, the number of studies on wheat–rice hybridization is small. To the best of our knowledge, Bakos et al. (2005) attempted cross‐pollinations between wheat pistils and rice pollen and reported that fertilization or zygotic development was not achieved. These strongly suggest that cross‐pollination is not suitable for producing wheat–rice hybrid zygotes. Therefore, in this study, by combining rice and wheat in vitro fertilization systems, we produced allopolyploid wheat–rice hybrid zygotes with a variety of gamete (genome) combinations and analyzed their developmental profiles to acquire some insight into the hybrid nature between wheat and rice genomes. In addition, because wheat and rice are two of three major energy‐producing crops, utilization of hybrid genetic resources in plants regenerated from allopolyploid wheat–rice hybrid zygotes is likely to have a positive impact on agriculture.

Materials and Methods

Plant materials and gamete isolation

The wheat plants (Triticum aestivum L. cv Fielder) used in this study were grown in an environmental chamber under 16 h 23°C : 8 h 20°C, light : dark conditions and a photosynthetic photon flux density of 100–150 µmol m−2 s−1. Oryza sativa L. cv Nipponbare plants were grown in an environmental chamber under 13 h : 11 h, light : dark conditions at 26°C. Transformed rice plants expressing the histone H2B‐GFP fusion protein were prepared as previously described (Abiko et al., 2013). Tetraploid rice plants expressing the histone H2B‐GFP fusion protein were prepared as previously described (Toda et al., 2018). The isolation of egg and sperm cells from rice and wheat flowers was conducted as previously described (Uchiumi et al., 2006; Maryenti et al., 2019).

Production of wheat–rice hybrid zygotes

Electrofusion of isolated gametes from wheat and/or rice was performed as previously described (Maryenti et al., 2019) with some modifications. Schematic illustrations of the procedures to generate isogenic and inter‐subfamily zygotes are shown in Fig. 1. Inter‐subfamily zygotes consisting of one rice egg cell and one wheat sperm cell were generated by fusing a rice egg (Re) expressing H2B‐GFP with a wheat sperm with a single DC pulse as in the fusion between rice gametes (Fig. 1a; Uchiumi et al., 2007) to produce a ReWs hybrid zygote (Fig. 1c). Inter‐subfamily 2ReWs zygotes using two rice egg cells and one wheat sperm cell were produced in two steps. Two rice egg cells (2Re) expressing H2B‐GFP were initially fused, and then the fusion product was fused with a wheat sperm cell (Ws) (second fusion, Fig. 1d). A ReWsWe zygote was produced by initially fusing one rice egg (Re) expressing H2B‐GFP with a wheat sperm (Ws), and then the fusion product was fused with a wheat egg (We) (second fusion; Fig. 1e).

Fig. 1

Schematic illustrations of the procedures for generating inter‐subfamily polyploid zygotes by in vitro fertilization. (a) Procedure to generate a rice zygote. (b) Procedure to generate a wheat zygote. (c) ReWs zygotes were generated by fusing one rice egg (Re) expressing H2B‐GFP and one wheat sperm (Ws). (d) 2ReWs zygotes were formed by first fusing two rice eggs (2Re) expressing H2B‐GFP and then the fused rice (2Re) was further fused with a wheat sperm (Ws). (e) ReWsWe zygotes were generated in two steps. First, a rice egg (Re) expressing H2B‐GFP and a wheat sperm (Ws) cell were fused. Thereafter, the fusion product (ReWs) was fused with a wheat egg (We) (second fusion). (f) WeRs zygotes were formed by fusing one wheat egg (We) and one rice sperm (Rs) expressing H2B‐GFP. (g) WeRs(2n) zygotes were generated by fusing a wheat egg (We) and a diploid rice sperm (Rs(2n)) which was isolated from a tetraploid plant expressing H2B‐GFP. (h) ReRsWeWs zygotes/Double zygote (DZ) were generated in three steps. First, a rice egg (Re) and rice sperm (Rs) were fused. Thereafter, the fused product (ReRs) was further fused with a wheat sperm (Ws) and finally a wheat egg (We). The large and small pink circles indicate the nuclei of rice egg and sperm cells, respectively. The large and small yellow circles indicate the nuclei of wheat egg and sperm cells, respectively. The yellow bolt indicates the point of electrofusion.

The WeRs hybrid zygotes, consisting of one wheat egg (We) and one rice sperm (Rs) expressing H2B‐GFP (Fig. 1f), were fused by multiple DC pulses as in the fusion between wheat gametes (Fig. 1b; Maryenti et al., 2019). The WeRs(2n) zygotes were produced by fusing one wheat egg (We) and one diploid rice sperm (Rs(2n)) isolated from a tetraploid rice plant expressing H2B‐GFP (Fig. 1g).

Inter‐subfamily polyploid zygotes (DZ) were generated in three steps. A rice egg (Re) was initially fused with a rice sperm (Rs), and then the resulting zygote was fused with a wheat sperm (Ws). This fused cell was then fused with a wheat egg (We) (third fusion, Fig. 1h), resulting in production of wheat–rice hybrid double zygote (DZ) (Fig. 1h).

Culture and observation of IVF‐produced inter‐subfamily zygotes

Wheat–rice inter‐subfamily zygotes were cultured into early embryos as previously described (Maryenti et al., 2019). The embryos were cultured into calli, and the calli were regenerated into plantlets according to Ishida et al. (2015) and Maryenti et al. (2019). Zygotes/embryos expressing H2B‐GFP were observed under a BX‐71 inverted fluorescence microscope (Olympus, Tokyo, Japan) using excitation at 460–490 nm and emission at 510–550 nm wavelengths (U‐MWIBA2 mirror unit; Olympus). Digital images were obtained using a cooled charge‐coupled device (CCD) camera (Penguin 600CL; Pixera, Los Gatos, CA, USA) and InStudio software (Pixera).

Fluorescence in situ hybridization of IVF‐produced inter‐subfamily globular‐like embryos

Globular‐like embryos at 3 and 4 d after fusion (DAF) were fixed with 3 : 1 (v/v) ethanol : glacial acetic acid at 4°C. The fixed embryos were placed onto a glass slide and stained with 2% (w/v) carmine (Merck, Darmstadt, Germany) in 45% acetic acid for 5 min. The staining solution was removed with Kimwipes (Kimberly‐Clark Worldwide, Irving, TX, USA), and the embryos on glass slides were washed in distilled water for 1 min. Embryos were then treated with 2% (w/v) of pectolyase Y23 (Kyowa Kasei, Osaka, Japan) and 2% (w/v) cellulase ‘ONOZUKA’ R‐10 (Yakult, Tokyo, Japan) for 5–10 min at 37°C under humid conditions. Slides were wiped carefully with Kimwipes to remove the enzyme solution and 10 μl of 45% acetic acid was dropped onto the slides containing the treated embryos. The embryos were squashed with coverslips in 45% acetic acid and stored at −80°C overnight. The coverslips were then removed, and the slides were dried at room temperature.

Wheat chromosomes were probed with wheat genomic DNA labeled with ChromaTide Alexa Fluor 488‐5‐dUTP (Thermo Fisher Scientific, Waltham, MA, USA) using a nick‐translation system (Thermo Fisher Scientific) according to the manufacturer’s instructions. Rice centromere‐specific oligo probes were designed based on centromeric satellite DNA sequence, and 5′ end‐labeled with ATTO 565 (CAAAAMTCATGTTTKGGTGN and GGACMTAWWGKAGTGKATN). Embryos on glass slides were denatured with 0.2 M NaOH in 70% ethanol for 5 min at room temperature and then dehydrated sequentially in 70%, 90% and 99.5% ethanol at room temperature. Hybridization solution (10 μl) containing 50% formamide, 10% dextran sulfate, and 50–100 ng of labeled probe in 2× saline sodium citrate (SSC) was applied to each slide. Slides were covered with 22 × 22 mm coverslips and sealed with paper bond glue (Kokuyo, Osaka, Japan) and hybridized at 37°C for 24–36 h. Slides were then washed sequentially at room temperature with 0.1% Triton X‐100 in 2× SSC for 5 min and 2× SSC for 5 min. The slides were dehydrated sequentially in 70%, 90%, and 99.5% ethanol at room temperature. Nuclei were counterstained with Vectashield (Vector Laboratories, Burlingame, CA, USA) containing 1 ng µl−1 4′,6‐diamidino‐2‐phenylindole (DAPI).

Fluorescence imaging was performed using an Olympus BX61 microscope equipped with an ORCA‐ER CCD camera (Hamamatsu Photonics, Hamamatsu, Japan). All images were acquired in grayscale with CellSens Dimension 1.11 software (Olympus Soft Imaging Solutions) and pseudo‐colored and merged in Adobe Photoshop 2020 (Adobe, San Jose, CA, USA).

Flow cytometry analyses

The nuclear DNA content (ploidy level) of plants regenerated from inter‐subfamily polyploid zygotes was measured by flow cytometry using a CyFlow CCA (Partec, Muenster, Germany) and a QuantumStain NA UV2 Kit (Quantum Analysis, Muenster, Germany). Fresh leaves (25–50 mm2) were chopped with a sharp razor in 100 μl of kit solution, and 100 μl more was added after chopping. The samples were incubated for 5 min, and then the crushed tissue suspension was filtered through a 30 μm nylon mesh (Partec). The samples were then loaded into the ploidy analyzer. Leaf samples from parsley (Petroselinum crispum spp.) or pothos (Epipremnum aureum spp.) were used as internal controls. Fluorescence intensities were measured and plotted in real time. The peak and mean values were calculated using FloMax software.

Chromosome counting

Seeds were germinated on a moistened filter paper with 2 nM of 8‐hydroxyquinoline. Root lengths of c. 2 cm were collected from the germinated seeds and transferred into a 2 ml tube filled with double distilled water. The tube was then placed on ice at 4°C overnight for preparation of metaphase chromosomes. Pretreated roots were fixed with 3 : 1 (v/v) ethanol : glacial acetic acid at room temperature for 3 d, and the fixed roots were used for slide preparation as described in Ishii et al. (2012). The slides were observed with an Olympus BX41 microscope and the images were taken with a DP21 camera (Olympus) and the chromosome numbers were counted using ImageJ.

Genomic PCR for rice simple sequence repeat (SSR) markers and a rice centromere‐specific repeat

A set of total 50 SSR markers that are evenly distributed across the whole 12 chromosomes of rice were obtained from the Gramene marker database (https://archive.gramene.org/markers/microsat/50_ssr.html) developed from O. sativa L. cv Nipponbare genome sequencing. Among the 50 SSR markers, SSR markers on short or long arms of each 12 rice chromosomes (a total of 24 SSR marker sets) were selected. The detailed information about the 24 SSR markers and PCR primer sets for the markers used in the study was presented in Supporting Information Table S1. A marker of rice centromere‐specific repeat, CentO, and the specific primer set (5′‐CAAAAMTCATGTTTKGGTGN‐3′ and 5′‐GGACMTAWWGKAGTGKATN‐3′) for the marker were used according to Zhang et al. (2005).

Genomic DNAs were isolated from the leaves of ReWsWe 3A, WeRs(2n) 2B, and DZ 5B hybrid plants, wheat plants and rice plants using Nucleo Spin II Kit (Macherey‐Nagel, Düren, Germany) according to the manufacturer’s instructions. For PCR, 0.5 µl of DNA (1 ng µl−1) was used as the template in a 20 µl PCR reaction with 0.3 µM of primers using KOD‐FX DNA polymerase (Toyobo, Osaka, Japan). The touchdown PCR method was used for SSR marker detection with the condition as follows: an initial denaturation at 95°C for 5 min, followed by 22 cycles each of denaturation at 95°C for 30 s, annealing at 70°C for 1 min (−1°C/cycle) and extension at 72°C for 30 s; another 13 cycles of 95°C for 30 s, 57°C for 30 s and 72°C for 30 s with a final extension at 72°C for 1 min. A PCR for rice centromere‐specific marker was performed as follows: an initial denaturation at 95°C for 5 min, followed by 35 cycles of 95°C for 30 s, and 57°C for 30 s with a final extension at 72°C for 1 min. In addition, a partial region of wheat CP3 gene (MK647992.1) was amplified using the primer set (5′‐CTTCAAGTCCAACGTCCACTT‐3′ and 5′‐GCGCGGAACTCGGCCTG‐3′) and touchdown PCR cycle as described earlier.

Results

Production and development of inter‐subfamily polyploid zygotes of ReWs and 2ReWs

The procedures for producing isogenic rice and wheat zygotes using isolated gametes are presented in Fig. 1(a,b). The first combination of an inter‐subfamily polyploid zygote was conducted with one rice egg (Re) and one wheat sperm (Ws), and the resulting zygote was termed a rice egg–wheat sperm (ReWs) zygote (Fig. 1c). To visualize the nucleus, egg cells isolated from transgenic rice plants expressing an H2B‐GFP under regulation of the ubiquitin promoter were used (Abiko et al., 2013). The zygotes that progressed through karyogamy developed into six‐ to eight‐celled embryos at 3 d after fusion (DAF) (Fig. 2a). Although early embryos grew into globular‐like embryos consisting of c. 14–16 cells at 6 DAF, the GFP signal became undetectable. The cells of globular‐like embryos then became highly vacuolated, and developmental arrest was observed at 13 DAF. The rate at which ReWs hybrid zygotes developed was slower than that of IVF‐produced rice zygotes (Uchiumi et al., 2007). Among the 10 ReWs zygotes produced in this study, six developed into globular‐like embryos (Table 1), indicating that c. 60% (6/10) of ReWs hybrid zygotes successfully underwent early embryonic development through the possible activation of fused egg cells, although eventually ReWs embryos showed aberrant division and degenerated in the later stage (Fig. 2a).

Fig. 2

Developmental profiles of inter‐subfamily ReWs, 2ReWs and ReWsWe hybrid zygotes. (a) ReWs zygotes were produced fusing a rice egg (Re) expressing H2B‐GFP and a wheat sperm (Ws). ReWs zygotes underwent karyogamy at 1 d after fusion (DAF) and developed into early globular‐like embryos consisting of approximately six to eight cells (3 DAF). The GFP signal became undetectable at 6 DAF. Globular‐like embryos became highly vacuolated and arrested at 13 DAF. (b) 2ReWs zygotes were generated by first fusing two rice eggs (2Re) expressing H2B‐GFP, and then a wheat sperm (Ws). 2ReWs zygotes underwent karyogamy (1 DAF) and developed into two‐celled and globular‐like embryos at 3 and 5 DAF, respectively. The intensity of GFP signal decreased at c. 5 DAF and the signal was undetectable at 12 DAF. Then the development of the embryos arrested. (c) ReWsWe zygotes were generated by first fusing a rice egg (Re) expressing H2B‐GFP and a wheat sperm (Ws) and then finally a wheat egg (We). ReWsWe zygotes developed into two‐celled embryos (1 DAF) and early globular‐like embryos, in which a weak GFP‐derived signal was detected (2 DAF). Although globular‐like (4 DAF) and club‐shaped embryos (8 DAF) developed, no GFP signal was detected. White calli were developed from club‐shaped embryos (28 DAF), and regenerated shoots were formed in the presence of the calli on regeneration medium (45 DAF). Regenerated plantlets were obtained after 2 wk in the presence of regenerated shoots on rooting medium (59 DAF), and the regenerated plant flowered at 140 DAF. Bars, 50 µm (a, 1, 3, 6, 10, 13 DAF; b, 1, 3, 5, 12, 19 DAF; c, 1, 2, 4, 8 DAF); 5 mm (c, 28 and 45 DAF); 1 cm (c, 59 DAF).

Table 1

Developmental profiles of inter‐subfamily polyploid zygotes between wheat and rice.

Parental type	Gamete used for fusion		Abbreviated name of zygote	Fusion procedure	No. of produced zygotes	No. of zygotes developed into each growth stage
	Egg	Sperm				Karyogamy	Two‐celled embryo	Globular‐like embryo	Compact embryonic callus	Callus	Plantlet	Plant line
Isogenic	Rice	Rice	ReRs	Fig. 1(a)	97*	97*	89*	81*	81*	81*	81*	‐
Isogenic	Wheat	Wheat	WeWs	Fig. 1(b)	34**	34**	31**	30**	30**	30**	30**	‐
Inter‐subfamily	Rice	Wheat	ReWs	Fig. 1(c)	10	10	6	6	0	0	0	‐
Inter‐subfamily	Rice x2	Wheat	2ReWs	Fig. 1(d)	8	8	7	7	0	0	0	‐
Inter‐subfamily	Rice, wheat	Wheat	ReWsWe	Fig. 1(e)	5	5	4	4	4	4	3	1A, 1B, 2A, 2B, 3A, 3B
Inter‐subfamily	Wheat	Rice	WeRs	Fig. 1(f)	10	10	1	0	0	0	0	‐
Inter‐subfamily	Wheat	Rice(2n)	WeRs(2n)	Fig. 1(g)	10	10	9	9	9	9	3	1A, 1B, 2A, 2B, 3A
Inter‐subfamily	Rice, wheat	Rice, wheat	DZ	Fig. 1(h)	7	7	6	6	6	6	5	1A, 1B, 2A, 2B, 3A, 3B, 4A, 4B, 5A, 5B

Re, rice egg; Rs, rice sperm; We, wheat egg; Ws, wheat sperm; 2Re, two rice eggs; Rs(2n), diploid rice sperm; DZ, double zygote.

*Total number of cells or embryos determined by Uchiumi et al. (2007).

**Total number of cells or embryos determined by Maryenti et al. (2019).

Next, an additional rice egg (Re) was fused to the ReWs zygote to increase the rice genome dosage, because the total length of genomic DNA and ploidy level differ significantly between rice (0.4 Mbp genome size, diploid) and wheat (1.7 Gbp, hexaploid). Eight 2ReWs zygotes were produced (Fig. 1d), and their developmental profiles were monitored. Karyogamy was detected at 1 DAF (Fig. 2b), and the hybrid zygotes divided into two‐celled and four‐ to eight‐celled embryos at 3 and 5 DAF, respectively. However, after 5 d of gamete fusion, the hybrid embryo stopped its development and degenerated. In addition to developmental arrest of the early 2ReWs embryos, the GFP‐derived signal decreased and disappeared at 5 DAF (Fig. 2b). Although 2ReWs zygotes showed a higher proportion of globular‐like embryo stages (87.5%, 7/8) than did ReWs zygotes (60%, 6/10) (Table 1), hybrid developmental arrest was not recovered by the fusion of an additional rice egg (Re).

Chromosome composition of ReWs embryos

Fluorescence in situ hybridization (FISH) analyses using the probes for rice centromere sequence and/or wheat genome were employed to determine chromosome composition of IVF‐produced embryos. Root cell nuclei from each species were used to confirm the specificity of the FISH probes for rice centromeres (red) and wheat genome (green) (Fig. 3a). The same probe sets were used to investigate the chromosome compositions of rice (control) and ReWs (experimental) IVF‐produced embryos. Rice centromeres were detectable in all nuclei of rice embryos; however, no wheat genomic signals were observed (Fig. 3b). In hybrid rice embryos, rice centromeres were detected at the periphery or outside nuclei, whereas signals of wheat‐specific probes were stable and uniformly detected in all hybrid nuclei (Fig. 3c–e). These data suggest that rice chromosomes are eliminated during embryogenesis, resulting in the formation of micronuclei. These findings are in line with reports on micronuclei formation in hybrid embryos after wide hybridization of two different species (Kasha & Kao, 1970; Gernand et al., 2005; Ishii et al., 2010). Moreover, the possible elimination of rice chromosomes in ReWs embryos is consistent with the disappearance of fluorescence signals derived from the H2B‐GFP transgene in rice genomes during the development of the hybrid embryos (Fig. 2a).

Fig. 3

Fluorescence in situ hybridization (FISH) analyses of rice centromeres and wheat genome in nuclei from ReWs hybrid embryos. FISH/genome in situ hybridization (GISH) was conducted on root cells from rice and wheat plants (a), cells of rice in vitro fertilization (IVF)‐produced embryos (b), cells of ReWs embryos (c–e). Red indicates FISH probes targeted to rice centromere‐specific repetitive sequences. Green indicates GISH probes targeted to wheat genomic DNA. Blue indicates DNA counterstaining with 4′,6‐diamidino‐2‐phenylindole (DAPI). Bars: (a–e) 5 μm.

Production and successful development of inter‐subfamily polyploid ReWsWe zygotes

The FISH analyses suggested that the elimination of rice nuclear genomes occurred during the development of ReWs zygotes, which concomitantly induces coexistence of remaining wheat genomes in nuclei and rice cytoplasmic genomes in the cells of ReWs embryos. This allogenic cytonuclear situation might explain the developmental arrest of ReWs globular‐like embryos, because dysfunctional interactions between a cytoplasmic genome (usually mitochondrial) from one species and a nuclear genome from the other often result in organelle dysfunction and, consequently, hybrid breakdown (Johnson, 2010; Burton et al., 2013). Therefore, an extra wheat egg (We) was fused with a ReWs zygote to supply wheat cytoplasm as well as a female nucleus to overcome the developmental arrest (Fig. 1e). Among five ReWsWe hybrid zygotes produced in this study, four hybrid zygotes developed into two‐celled embryos and globular‐like embryos at 2 and 4 DAF, respectively (Fig. 2c; Table 1). Notably, beyond the globular‐like stage, ReWsWe hybrid embryos developed into club‐shaped embryos (8 DAF), as in the case of the developmental process of wheat zygotes produced in vitro (Maryenti et al., 2019). Although a low‐intensity GFP signal was detectable in the nuclei of hybrid zygotes and early embryos at 1 and 2 DAF, respectively, the signal disappeared at the globular‐like embryo stage before the zygotes developed into club‐shaped embryos at 8 DAF. Club‐shaped embryos developed into calli at 28 DAF, and then shoots and plantlets were regenerated from the calli (Fig. 2c). The regenerated plantlets grew into mature plants, and spikes emerged from the regenerated plants at 140–150 DAF (Fig. 2c). Three of four ReWsWe calli demonstrated regeneration potential (Table 1), and two plantlets selected from the callus‐derived shoots/roots were grown into mature plants (n = 6 plants) (Table 1; Fig. S1a).

Characterization of ReWsWe hybrid plants: characteristics of phenotype, ploidy and chromosome number

All ReWsWe plants (1A, 1B, 2A, 2B, 3A, and 3B) were morphologically similar to wheat plants. The ReWsWe plants were fertile and viable seeds were successfully harvested (Figs S1a, S2b–d). The nuclear DNA content of ReWsWe plants (ReWsWe 1A, 2A, and 3A) was measured by flow cytometry. DNA peaks were detected in each of the three plant lines at a position corresponding to the 2C peak in control wheat plants (Figs 4c; S3). During the development of early ReWsWe embryos, GFP signals derived from transgenes on rice genomes disappeared (Fig. 2c) as in the case of ReWs embryos (Fig. 2a). In addition, 42 chromosomes, equivalent to chromosome numbers of hexaploid wheat, were detected in root cells from ReWsWe progeny (Fig. 4c). These suggested that elimination of rice chromosomes occurred in ReWsWe embryos as well as ReWs embryos (Fig. 3). Plant morphology, ploidy analysis and possible rice chromosome elimination suggest that the regenerated ReWsWe plants are equivalent to wheat plants.

Fig. 4

Ploidy level and chromosome number of plants regenerated from wheat–rice hybrids. Nuclei were extracted from leaves of wild‐type rice (a), wild‐type wheat (b), ReWsWe 3A (c), WeRs(2n) 2A and 2B (d,e) and double zygote (DZ) 5B (f) plants, and the DNA content per nucleus was measured by flow cytometry. Nuclei extracted from parsley was used as internal control. Right panels show chromosomes in root cells of seeds obtained from wild‐type wheat (b), ReWsWe 3A (c), WeRs(2n) 2B (e), DZ 5B (f) plants. Forty‐two chromosomes were detected in all hybrid root cells (c, e–f) as well as control wheat root cells (b). Bars: (b–c, e–f) 10 μm. Re, rice egg; Rs, rice sperm; We, wheat egg; Ws, wheat sperm.

Production and successful development of inter‐subfamily polyploid WeRs and WeRs(2n) zygotes

Since the fusion products of ReWs combination ceased the development at the multicellular embryo‐like structure stage, a reciprocal fusion combination between wheat egg (We) and rice sperm (Rs) was examined (WeRs; Fig. 1f). A rice sperm cell expressing H2B‐GFP was used to monitor developmental profiles of WeRs hybrid zygotes. At 1 DAF, zygotic nuclei labeled with H2B‐GFP were visible, suggesting that karyogamy had occurred (Fig. 5a). Zygotes at the one‐celled stage were still observed at 3 DAF, and the hybrid zygotes degenerated without division (Fig. 5a; Table 1). These findings suggest that a haploid rice sperm cell does not contain enough cytoplasm and/or genome dosage to activate/initiate division and developmental machineries in hybrid zygotes (fused wheat cells). These also provide the possibility that doubling of cytoplasm and/or genome dosage in a rice sperm cell triggers the activation and subsequent development of hybrid zygotes, because the sperm cell delivers cytoplasmic and/or genetic male factors into the egg cell that initiates egg cell activation and subsequent early development (Denninger et al., 2014; Hamamura et al., 2014; Ohnishi et al., 2019).

Fig. 5

Developmental profiles of inter‐subfamily WeRs and WeRs(2n) zygotes and characterization of WeRs(2n) embryos and plants. (a) Developmental profile of the inter‐subfamily WeRs zygotes. WeRs zygotes were produced by fusing a wheat egg (We) and a rice sperm (Rs) expressing H2B‐GFP. Although WeRs zygotes underwent karyogamy at 1 d after fusion (DAF), no division or further development was detected, and the zygotes degenerated. (b) Isolated haploid and diploid sperm cells. Haploid (n) and diploid (2n) sperm cells (upper‐right panels) were isolated from diploid (2×) and tetraploid (4×) rice plants (left panels), respectively. Diameters of haploid and diploid sperm cells are presented in the lower‐right panel. The data are mean ± SD of diameters of haploid (n = 20) and diploid (n = 20) sperm cells. Statistical significance is determined by Student’s t‐test (***, P < 0.001). (c) Developmental profile of the inter‐subfamily WeRs(2n) zygote. WeRs(2n) zygotes were produced by fusing a wheat egg (We) and a diploid rice sperm (Rs(2n)) expressing H2B‐GFP. WeRs(2n) zygotes divided into two‐celled and four‐ to eight‐celled embryos at 1 and 2 DAF, respectively. The loss of GFP signal was observed in globular‐like embryos at 5 DAF. The WeRs(2n) hybrid embryos grew into calli (34 DAF), which formed green spots (44 DAF) and small shoots (61 DAF). The small shoots developed into the regenerated shoots and plantlets at 81 and 176 DAF, respectively. The regenerated plant flowered at 269 DAF. (d) Reproductive organs and tissues of WeRs(2n) plants. The spikes, spikelets, pistils, anthers and pollen grains from WeRs(2n) and wild‐type (WT) wheat plants are presented in the upper‐left, upper‐middle, upper‐right, lower‐left and lower‐right panels, respectively. (e) Rice centromeres in the nuclei of WeRs(2n) cells were visualized by fluorescence in situ hybridization (FISH). Red indicates FISH probes targeting rice centromere‐specific repetitive sequences. Blue indicates DNA counterstaining with 4′,6‐diamidino‐2‐phenylindole (DAPI). Bars: (e) 5 μm; (b, n sperm, 2n sperm) 20 μm; (a, 1, 3, 6 DAF; c, 0, 1, 2,5 DAF; and d, pollen grains) 50 μm; (c, 34 DAF) 1 mm; (d, spikelets, pistils, anthers) 2 mm; (c, 44, 61 and 81 DAF) 5 mm; (b, panicles; c, 176 DAF) 1 cm; (d, spikes) 2 cm.

To estimate the doses of male cytoplasm and genome, the sizes of haploid and diploid sperm cells isolated from diploid and tetraploid rice plants, respectively, were compared. The results indicate that the diameter of a diploid rice sperm (Rs(2n)) is approximately 1.4‐fold larger than that of a haploid rice sperm (Rs(n)) (Fig. 5b), and it was estimated that the doses of male cytoplasm and genome, which are delivered into the fused wheat egg cell, are increased in the diploid rice sperm (Rs(2n)). Ten WeRs(2n) hybrid zygotes were produced by fusing a wheat egg (We) with a diploid rice sperm (Rs(2n)) (Fig. 1g). Nine zygotes underwent karyogamy and developed into two‐celled and globular‐like embryos by 2 and 5 DAF, respectively (Table 1; Fig. 5c). Although the GFP signal was visible in the nuclei of the hybrid zygotes and early embryos at 2 DAF, the fluorescence signal progressively decreased as the zygotes became globular‐like embryos at 5 DAF. These nine WeRs(2n) hybrid embryos grew into calli at c. 34 DAF and the green spots on the calli became visible around 10 d after subculture of calli onto the regeneration medium (44 DAF). Although the regeneration of the WeRs(2n) shoots was notably slower than IVF‐produced wheat calli, small shoots emerged from six of the nine WeRs(2n) calli after being in the presence of regeneration medium (61 DAF). The development of regenerated shoots (81 DAF) into plantlets (176 DAF) was observed for three of the nine regenerated calli (Fig. 5c). However, no regenerated plantlets were obtained from the remaining six calli; instead, the green‐spotted calli and emerging shoots withered and turned brown at c. 80 DAF and stopped growing several days after being in the presence of regeneration medium (Fig. S4). Five plantlets (1A, 1B, 2A, 2B and 3A) were selected from the three callus‐derived shoots/roots and grew into mature plants (Figs 5c; S1b; Table 1). It took approximately 270 d for WeRs(2n) zygotes to develop into mature plants (Figs 5c; S1b), which is approximately two‐fold longer than wild‐type wheat plants (Maryenti et al., 2019) and ReWsWe plants (Fig. 2c).

Characterization of WeRs(2n) hybrid plants: morphological characteristics, chromosome composition, ploidy and chromosome number

The WeRs(2n) plants were morphologically equivalent to wheat plants (Fig. S1b), although two plants (1A and 3A) exhibited dwarf morphology (Fig. S1b). WeRs(2n) plants had smaller spikes and spikelets than wild‐type wheat plants, and the sizes of the pistils and anthers were approximately half of those of wild‐type plants (Fig. 5d). In addition, the pollen grains were poorly developed and looked empty, as shown in Fig. 5(d). WeRs(2n) plants were infertile, except for one plant (WeRs(2n) 2B in Fig. S1b; Table 1) that produced seeds (Fig. S2e).

The loss of GFP signal in WeRs(2n) embryos was similar to that observed in ReWs and ReWsWe embryos (Figs 2a,c, 5c). The FISH analyses of WeRs(2n) embryos were also conducted and the result showed that a few of rice centromere signals were detected and tended to localize at the periphery of the nuclei (Fig. 5e). These findings suggest that the elimination of rice chromosomes occurred during the early development of WeRs(2n) embryos.

The ploidy of five WeRs(2n) plants was analyzed by flow cytometry. In four plants (1A, 1B, 2A and 3A), the peak was detected at a position associated with half the DNA content of wild‐type wheat plants (Figs 4d; S1b, S5c–e,g), suggesting that the WeRs(2n) plants were haploid wheat plants. These are consistent with the phenotype of these WeRs(2n) plants: with small flower organ size and infertility which are the typical characteristics of haploid plants. Meanwhile, the fertile WeRs(2n) 2B plant (Figs 4e; S1b, S2e, S5f) possessed a nuclear DNA content equivalent to that of wild‐type wheat plants, and 42 chromosomes were observed in root cells from the seeds of WeRs(2n) 2B plant (Fig. 4e). These indicate that chromosome doubling of the haploid plant occurred at some point during development.

Production and successful development of inter‐subfamily wheat and rice double zygote (DZ)

A zygote is a totipotent cell with a complete genome and cytoplasm equipped for proliferation. To investigate the effects of hybridizing the totipotent cellular component between wheat and rice, the fused gametes consisting of a rice egg (Re), a rice sperm (Rs), a wheat egg (We) and a wheat sperm (Ws) were produced (Fig. 1h), and termed as wheat–rice double zygote (DZ), in which complete genome and cytoplasm sets from wheat and rice coexist. A set of rice gametes, egg and sperm cells fluorescently labeled with H2B‐GFP, was combined with a pair of wheat gametes, resulting in DZ production (Fig. 1h). Seven DZ zygotes were produced, and six divided into two‐celled embryos and globular‐like embryos by 2 and 4 DAF, respectively (Table 1; Fig. 6a). Although the GFP signal was visible in the nuclei of early DZ embryos until 4 DAF, it disappeared when the zygotes developed into globular‐like embryos at 5 DAF. Calli were obtained after subculturing the compact embryonic calli on callus induction medium at 29 DAF, and five out of the six calli regenerated into plantlets at c. 59 DAF. The plantlets grew into mature plants that flowered at 130–160 DAF (Figs 6a; S1c), and seeds were harvested from several regenerated plants (Fig. S2f–h). Ten plantlets (DZ 1A, 1B, 2A, 2B, 3A, 3B, 4A, 4B, 5A and 5B) were selected from five callus‐derived shoots/roots (Table 1; Fig. S1c). The time course from a two‐celled DZ embryo to a callus was equivalent to that of an in vitro‐produced wheat plant (Maryenti et al., 2019); however, the flowering time was 2 wk slower.

Fig. 6

Developmental profiles of inter‐subfamily double zygote (DZ) and characterization of DZ embryos. (a) DZ zygotes were generated in three steps: first by fusing a rice egg (Re) and a rice sperm (Rs) expressing H2B‐GFP, and then a wheat sperm (Ws), and finally a wheat egg (We). DZ zygotes divided into four‐ to eight‐celled embryos and globular‐like embryos by 2 and 4 d after fusion (DAF), respectively. A weak GFP signal was detected at 4 DAF and disappeared at 5 DAF. DZ hybrid embryos grew into calli (29 DAF), regenerated shoots (43 DAF) and regenerated plantlets (59 DAF). Regenerated plants flowered at 130 DAF. (b) Rice centromeres in the nuclei of DZ cells were visualized by fluorescence in situ hybridization (FISH). Red indicates FISH probes targeting rice centromere‐specific repetitive sequences. Blue indicates DNA counterstaining with 4′,6‐diamidino‐2‐phenylindole (DAPI). Bars: (b) 5 μm; (a, 0, 2, 4, 5 DAF) 50 μm; (a, 29 DAF) 1 mm; (a, 43 DAF) 5 mm; (a, 59 DAF) 1 cm.

Characterization of DZ hybrid plants: morphological characteristics, chromosome composition, ploidy and chromosome number

Of the 10 regenerated DZ plants, six (DZ 1A, 1B, 3A, 3B, 5A, and 5B) were morphologically identical to wheat plants, but the four remaining plants (DZ 2A, 2B, 4A and 4B) were phenotypically different, exhibiting dwarf (DZ 2A, 2B and 4B) and infertile phenotypes (DZ 2A, 2B, 4A and 4B) (Fig. S1c). Rice genome‐derived GFP signals disappeared during the development of DZ embryos, as they did in WeRs(2n) and ReWsWe embryos, suggesting that rice chromosomes were eliminated during the development of DZ embryos that originally possessed complete rice and wheat genomes. Therefore, FISH was performed to confirm that chromosome elimination was occurring in DZ embryos. The FISH results indicated that rice chromosomes were eliminated, despite the presence of a pair of rice gametes in a DZ zygote (Fig. 6b).

The nuclear DNA content of DZ plants was also examined by flow cytometry. The positions of the DNA peaks from the four DZ plants tested, including DZ plants with normal and dwarf morphology (Fig. S1c), were equivalent to the 2C DNA peak of a control wheat plant (Figs 4f; S6), which is consistent with the fertility observed in these DZ plants (Fig. S2f–h). In addition, 42 chromosomes were detected in root cells of DZ seedlings as well as control wheat seedlings (Fig. 4b,f).

Genomic PCR for rice SSR markers and a rice centromere‐specific repeat

Genomic PCR for rice SSR markers and a rice centromere‐specific (CentO satellite) repeat were conducted using ReWsWe 3A, WeRs(2n) 2B and DZ 5B hybrid plants, as it was estimated that rice chromosomes were eliminated during the development of hybrid zygotes/embryos. Simple sequence repeat markers on short or long arms of each of 12 rice chromosomes (total 24 SSR marker sets) were selected from the Gramene marker database, and utilized for genomic DNA PCR (Table S1). When genomic DNAs from hybrid plants and wheat plants were used as templates, no SSR‐derived band was amplified in all primer sets for 24 SSR markers although specific amplification for SSR loci was detected in rice genome templates (Fig. S7). In addition, genomic PCR for a rice centromere‐specific repeat resulted in specific amplification of PCR band only in the genomic DNA from rice plants (Fig. S7). These results supported the elimination of rice chromosomes in wheat–rice hybrid combinations.

Discussion

The multiple combinations of fusion between rice and wheat gametes required more intricate conditions than the fusion of gametes from a single species because wheat and rice gametes possess distinct cellular characteristics. When using a rice egg cell for fusion, only a single DC pulse is necessary to produce the zygote (Uchiumi et al., 2007), which explains why polyploid rice zygotes have been successfully produced via IVF and used to study the functions of parental genomes and male/female gametic factors during zygotic development (Toda et al., 2018; Ohnishi et al., 2019). By contrast, the plasma membrane of a wheat egg cell is rigid (Maryenti et al., 2019), making it difficult to fuse with other gametes. Therefore, multiple DC pulses were necessary to fuse wheat egg cell and other gametes. Furthermore, the quality of the gametes affects the fruitfulness of fusion. Although freshly isolated gametes were more successful in inter‐subfamily fusions, storing egg cells for 1 d at 4°C was possible with a slight decrease in fusion efficiency. However, as the sperm cells needed to be fused within 15 min of isolation, a repeated isolation of sperm cells was required to increase the probability of success in inter‐subfamily gamete fusions. Therefore, the production of wheat–rice hybrid zygotes requires well‐organized processes for gamete isolation and fusion, and the steps for producing allopolyploid zygotes from different combinations of rice and wheat gametes (Fig. 1c–h), as described in the Materials and Methods section, are optimized for each combination.

The WeRs hybrid zygotes remained in the single‐cell stage and did not divide (Fig. 5a). The developmental arrest of WeRs zygotes was overcome by using a diploid rice sperm cell isolated from tetraploid plants (Fig. 5b), containing a double dose of male cytoplasm and genome. The entry of a sperm cell into an egg cell transiently increases the intracellular Ca2+ concentration in the fused egg cell (Antoine et al., 2000; Denninger et al., 2014; Hamamura et al., 2014; Ohnishi & Okamoto, 2017). This increase in Ca2+ concentration induces cell wall formation and accelerates karyogamy (Denninger et al., 2014; Ohnishi et al., 2019). In addition, transcripts of SHORT SUSPENSOR (SSP), a member of the interleukin‐1 receptor‐associated kinase/pelle‐like kinase family, are delivered from the sperm cells upon gamete fusion and help to establish the cell fate and body plan of early embryos in Arabidopsis (Bayer et al., 2009). Moreover, genes preferentially expressing from paternal alleles have been identified, and a paternally derived AP2‐type transcription factor, termed OsASGR‐BBML1 or OsBBML1, was shown to be a possible inducer of zygotic development in rice (Khanday et al., 2019; Rahman et al., 2019). The activation and subsequent development of a wheat egg (We) fused with a diploid rice sperm (Rs(2n)) can be explained by the fact that a double dose (relative to a haploid rice sperm cell) of such male‐derived activation factors was delivered, activating the WeRs(2n) zygote and initiating its developmental machineries.

As WeRs(2n) zygotes developed, rice chromosomes (male‐derived) appeared to be eliminated, and the regenerated plants from WeRs(2n) zygotes were nearly equivalent to haploid wheat plants. Selective chromosome elimination (i.e. a complete loss of paternal chromosomes) has been reported for a broad range of plant hybridizations (Ishii et al., 2016). Examples include crosses between wheat (female) and Hordeum bulbosum (male) or more distantly related species, such as maize (Zea mays), pearl millet (Pennisetum glaucum), sorghum (Sorghum bicolor), Coix lacryma‐jobi and Imperata cylindrica (Barclay, 1975; Laurie & Bennett, 1986, 1988, 1989; Laurie, 1989; Inagaki & Mujeeb‐Kazi, 1995; Mochida & Tsujimoto, 2001; Komeda et al., 2007). The reason for inadequate information on the dynamics of parental genomes in hybrids between rice and wheat is probably a result of unsuccessful hand pollination. The present study clearly showed that selective elimination of rice chromosomes from wheat–rice hybrids occurred in each combination, regardless of whether a complete set of rice gametes was present, as in the DZ combination.

In contrast to WeRs hybrid zygotes, the ReWs hybrid zygotes, the fusion products between rice egg (Re) and wheat sperm (Ws), progressed through several rounds of division; however, development was arrested at the early embryo stage. Notably, female gamete‐derived rice chromosomes were selectively eliminated during early embryonic development in ReWs zygotes. This finding indicates that selective rice chromosome elimination occurs in a hybrid zygote that is reciprocal to WeRs(2n). Moreover, selective rice chromosome elimination in ReWs embryos resulted in cells possessing the cytonuclear environment of wheat nucleus and rice cytoplasm. This allogenic cytonuclear environment was presumed to be the reason that ReWs embryos arrested at the globular‐like embryo stage, because dysfunctional interactions between a cytoplasmic genome from one species and a nuclear genome from another often result in organelle dysfunction and, consequently, hybrid breakdown (Johnson, 2010; Burton et al., 2013). The arrest of hybrid embryos was overcome by the addition of wheat egg (We) to ReWs zygotes (ReWsWe zygotes), which delivered additional wheat cytoplasm into ReWs zygotes/embryos. Krantz et al. (1995) showed that IVF‐produced hybrid zygotes between maize egg cell and wheat sperm cell developed into multicellular embryo‐like structures; however, the embryos ceased to develop. This embryonic arrest will also be a result of cytonuclear dysfunction between the wheat nucleus and maize cytoplasm, because the elimination of the maize chromosomes occurred during the development of the hybrid zygote with the wheat egg–maize sperm combination (Laurie & Bennet, 1988; Zhang et al., 1996).

DZ zygotes, consisting of a rice egg (Re), a rice sperm (Rs), a wheat egg (We) and a wheat sperm (Ws), also experienced rice chromosome elimination, yet developed and regenerated into hybrid plants. This finding indicates that the selective elimination of rice chromosomes progresses even when complete sets of wheat and rice nuclei and cytoplasms are present in hybrid zygotes. In all three types of hybrid zygotes/embryos, elimination of rice chromosomes was observed during early embryonic development. GFP signals derived from the rice genome started to disappear at 3–4 DAF, and FISH analyses confirmed that uniparental elimination of rice chromosomes occurred in the early embryos at 3–4 DAF. In addition, SSR markers for all 12 rice chromosomes and a rice centromere‐specific repeat marker were undetectable in the genomic DNA of hybrid plants. The developmental timing of massive elimination of rice chromosomes in wheat–rice hybrid embryos is similar to that of embryos produced via inter‐subfamily crossing in plants. Chromosome elimination occurred during early divisions in wheat × maize (Laurie & Bennett, 1989; Mochida et al., 2004), wheat × I. cylindrica (Komeda et al., 2007) and wheat × pearl millet crosses (Gernand et al., 2005; Ishii et al., 2010) at 3–4 post‐pollination, and more gradual chromosome elimination occurred at 7 d after pollination. In addition to elimination of uniparental genome, partial and true hybrid plants have been obtained via inter‐subfamily crossings of oat × maize and oat × pearl millet, respectively (Kynast et al., 2001; Ishii et al., 2013, 2015). These data suggest that the mechanisms involved in regulating the stability of inter‐subspecific genomes in the hybrid nucleus are highly complex. Because the combination of fused gametes can be artificially controlled with IVF, IVF‐produced wheat–rice hybrid zygotes and embryos are suitable materials for determining the mechanisms underlying genome stability/elimination.

In this study, rice chromosomes were progressively eliminated in all the wheat–rice hybrid embryos that overcame developmental barriers, and the hybrid plants displayed primarily wheat plant morphology. However, five out of 21 plants regenerated from hybrid zygotes exhibited abnormal morphology, such as dwarf, bushy or immature wheat plants, suggesting that the proportion of regenerated plants with abnormal morphology is c. 20%. This is not a result of the conditions of culture and regeneration procedures of hybrid zygotes/embryos/calli because wheat plants with abnormal morphology are rarely observed when IVF‐produced wheat zygotes were cultured and regenerated into wheat plants (Maryenti et al., 2019). To make clear the reason for these morphological variations, resequencing of the genomic DNA from the hybrid plants regenerated from wheat–rice hybrid zygotes is under way in our laboratories.

Author contributions

TM, TI and TO designed the experiments; TM performed most of the experiments; TI performed FISH/genome in situ hybridization (GISH) and chromosome counting experiments; TI and TO supervised the project; and TM and TO conceived the project and wrote the article.

Fig. S1 Regenerated plants of ReWsWe, WeRs(2n) and double zygote (DZ).

Fig. S2 Seeds of wild‐type wheat plants and inter‐subfamily regenerated plants of ReWsWe, WeRs(2n) and double zygote (DZ).

Fig. S3 Ploidy level of plants regenerated from ReWsWe zygotes.

Fig. S4 Arrested development in the early regenerated shoot stage of WeRs(2n).

Fig. S5 Ploidy of plants regenerated from WeRs(2n) zygotes.

Fig. S6 Ploidy level of plants regenerated from double zygote (DZ) zygotes.

Fig. S7 Genomic PCR for rice simple sequence repeat (SSR) and rice centromere‐specific (CentO) repeat markers.

Table S1 List of rice simple sequence repeat (SSR) primer sets.

Please note: Wiley Blackwell are not responsible for the content or functionality of any Supporting Information supplied by the authors. Any queries (other than missing material) should be directed to the New Phytologist Central Office.

Click here for additional data file.