Genomic introgressions from wild relatives in the wheat genome alter meiotic dynamics in inter‐varietal hybrids
Luxi Yan, Floriane Chéron, Isabelle Nadaud, Ming Hao, Dengcai Liu, Pierre Sourdille
TL;DR
Wild wheat relatives can alter meiosis in hybrids, affecting fertility and chromosome behavior, which is important for breeding resilient crops.
Contribution
A novel oligo-FISH approach reveals how introgressed wild segments affect meiosis in wheat hybrids.
Findings
Introgressions from Aegilops ventricosa reduce pollen viability and floret fertility in hybrids.
CSRe hybrids show increased rod bivalents and univalents, leading to reduced chiasma numbers and chromosome fragmentation.
2AS/2NS segments are often on rod bivalents/univalents, while 7DL/7DvL segments form ring bivalents.
Abstract
The use of wild relatives to introduce original diversity in the genome of bread wheat (Triticum aestivum L.) is an interesting approach to face the challenges of sustainable agriculture and the impact of climate change on wheat production. However, the influence of these wild‐species introgressions on meiosis in inter‐varietal wheat hybrids remains poorly understood. We analyzed the French wheat variety Renan (Re) carrying Aegilops ventricosa (Aev)‐derived 2AS/2NS and 7DL/7DvL introgressions, the reference cultivar Chinese Spring (CS), which lacks these introgressions, and their inter‐varietal hybrid Chinese Spring × Renan (CSRe). This analysis combined cytogenetic approaches with the assessment of reproductive performance. Furthermore, we generated a cytological atlas of meiosis in wild tetraploid Aev, quantifying bivalent configurations and chiasma frequency. We observed a reduced…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4|
| CS | Re | CSRe | |
|---|---|---|---|---|
|
| 111 | 119 | 127 | 126 |
|
| 28 | 42 | 42 | 42 |
|
| 27.06 ± 1.00 | 41.06 ± 1.06 | 41.37 ± 0.81 | 37.57 ± 1.84 |
|
| — | — | 0.3052 | 4.32 × 10−33 |
|
| — | 0.3052 | — | 1.16 × 10−43 |
|
| 0 | 0 | 0 | 0.78 ± 1.24 |
|
| 0.93 ± 1.00 | 0.94 ± 1.06 | 0.63 ± 0.81 | 3.65 ± 1.45 |
|
| — | — | 0.2435 | 2.36 × 10−30 |
|
| — | 0.2435 | — | 4.34 × 10−41 |
|
| 13.06 ± 1.00 | 20.06 ± 1.06 | 20.37 ± 0.81 | 16.96 ± 1.53 |
|
| — | — | 0.2850 | 1.60 × 10−32 |
|
| — | 0.2850 | — | 3.37 × 10−43 |
| Genotype | Chromosome 2A | Chromosome 7D | ||||
|---|---|---|---|---|---|---|
| CS | Re | CSRe | CS | Re | CSRe | |
| Cells with rings only | 15 | 19 | 0 | 17 | 21 | 0 |
| Cells with rods ≥1 | 35 | 31 | 50 | 33 | 29 | 50 |
| Type 1 | 34 | 31 | 16 | 33 | 29 | 48 |
| Type 2 | 1 | 0 | 13 | 0 | 0 | 2 |
| Type 3 | 0 | 0 | 10 | 0 | 0 | 0 |
| Type 4 | 0 | 0 | 11 | 0 | 0 | 0 |
- —Agence Nationale de la Recherche10.13039/501100001665
- —China Scholarship Council10.13039/501100004543
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsChromosomal and Genetic Variations · Wheat and Barley Genetics and Pathology · Genetic Mapping and Diversity in Plants and Animals
INTRODUCTION
The demographic increase projected over the coming five decades will induce a simultaneous growth in global cereal demand for both food and non‐food uses. In addition, the availability of agricultural land is expected to decrease, particularly in developing countries (United Nations, World Population Prospects, https://population.un.org/wpp/). Among staple crops, bread wheat (Triticum aestivum L.; 2n = 6x = 42; AABBDD) is the most extensively cultivated cereal worldwide, contributing approximately 20% of the total caloric and protein intake in the human diet (Food and Agriculture Organization of the United Nations, 2025, www.fao.org/faostat/en/#data/FBS; Ma et al., 2021). To meet future human needs by 2050, bread wheat production must increase by 40%, while considering the context of sustainable agriculture using less fertilizers, herbicides, pesticides, and water, as well as facing novel environmental conditions (drought and heat stresses) resulting from climate change.
One promising strategy to address this challenge is to revitalize the wheat gene pool by exploiting the wealth of genetic resources available in wheat and its wild relatives (Bailey‐Serres et al., 2019; Milner et al., 2019; Sansaloni et al., 2020). Despite their abundance, these resources remain largely underused, even though they harbor valuable genes and alleles underlying key agronomic traits, including resistance or tolerance to both biotic and abiotic stresses (Cox, 1997; Feuillet et al., 2008; Reif et al., 2005; Tanksley & McCouch, 1997; Wang et al., 2014). Harnessing this genetic diversity fundamentally relies on meiotic recombination, a highly conserved process at the core of the life cycle of all sexual eukaryotes, which ensures the faithful segregation of homologous chromosomes while generating genetic diversity among progeny (Mercier et al., 2015).
Because of its essential role in ensuring genome stability and fertility, meiotic recombination is tightly controlled in all organisms, including wheat (Sourdille et al., 2025). Although many recombination events are initiated by programmed DNA double‐strand breaks (DSBs), usually no more than one to three crossovers (COs), instances of reciprocal recombination between homologous chromosomes occur per pair of homologous chromosomes per meiosis (Lian et al., 2023; Saintenac et al., 2009). Moreover, COs are not randomly distributed along the chromosomes and in wheat, 80% of the COs occur only in the 20% distal regions of the chromosomes (Choulet et al., 2014; The International Wheat Genome Sequencing Consortium (IWGSC), 2014; The International Wheat Genome Sequencing Consortium (IWGSC), 2018), leaving 80% of the genome almost blind to any reshuffling. Finally, because wheat is an allohexaploid species deriving from two natural interspecific hybridizations involving three closely related diploid species (Triticum urartu, AA genome; a species related to Aegilops speltoides, SS genome related to BB genome; Aegilops tauschii, DD; Marcussen et al., 2014), there is a second layer of control, which prevents any recombination between related genomes (called ‘homoeologues’), avoiding introgressions of interesting genes or alleles from many wheat wild relatives.
Recombination between homoeologous chromosomes in wheat is mainly controlled by two pairing homoeologous (Ph) genes. Ph1 was found to be located on the long arm of chromosome 5B (5BL; Riley & Chapman, 1958; Sears & Okamoto, 1958). Deletion of the Ph1 locus induces a significant increase in the number of chiasmata up to ~20 COs/cell in wheat‐Ae. speltoides interspecific hybrids (Maestra & Naranjo, 1998). It was recently demonstrated that Ph1 corresponds to TaZip4‐B2 (Rey et al., 2017), which plays a dual role by promoting accurate recognition and pairing of homologous chromosomes during early stage of meiotic prophase I, while simultaneously preventing early formed homoeologous D‐loops to become a CO (Martín et al., 2014). Ph2 is located on the short arm of chromosome 3D (3DS; Mello‐Sampayo, 1971). Positional cloning of Ph2 revealed that it corresponds to TaMSH7‐3D, a plant‐specific protein involved in the mismatch repair system (MMR). Like for Ph1, mutation of Ph2 induces an increase of homoeologous pairing up to 5.5‐fold in wheat ph2 mutants × Aegilops variabilis interspecific hybrids (Serra et al., 2021). Remarkably, combining ph1 and ph2 mutations does not increase homoeologous recombination compared to the mutation of ph1 alone (Ceoloni & Donini, 1993; Haquet et al., 2026), suggesting that both genes act in the same recombination pathway. More recently, another locus named ph‐KL has been identified on chromosome arm 3AL (Fan et al., 2019), which can also induce homoeologous recombination between wheat chromosomes and those of closely related species (Fan et al., 2023; Hao et al., 2025), further highlighting the complexity of the regulation of homoeologous recombination in wheat.
Inactivating the Ph‐system is an effective strategy to induce homoeologous exchanges (Sears, 1976, 1977, 1982; Serra et al., 2021; Sutton et al., 2003). Since the 1960s, numerous disease resistance genes have been extensively introgressed into the bread wheat genome from a wide range of wild relatives (Tian et al., 2025; Wulff & Moscou, 2014). One of the most well‐known examples of such introgressions is the introduction of a resistance gene cluster (Lr37/Yr17/Sr38/Cre5) onto wheat chromosome 2A (2AS/2NS translocation; Helguera et al., 2003; Tanguy et al., 2005), and the simultaneous introduction of an eyespot resistance gene (Pch1) onto chromosome 7D (7DL/7DvL translocation; Mena et al., 1992; Leonard et al., 2008), both derived from the tetraploid wild species Aegilops ventricosa Tausch (Aev) (2n = 4x = 28, DvDvNN). Unfortunately, the introduction of Pch1 is also associated with a yield and quality penalty in the absence of the disease (Naranjo, 2019). This unfavorable linkage has never been fragmented after more than 50 years, questioning the reasons of this clamping. The most likely explanation is the absence of CO events, which has been observed in both biparental populations and core collections (Danguy des Déserts et al., 2021). However, the reasons for this lack of recombination events remain elusive.
One hypothesis for the absence of COs in introgressions is defective homologous synapsis and/or chiasma resolution within the introgressed segments. Synapsis corresponds to the tight alignment of homologous chromosomes that depends on the formation of the synaptonemal complex (SC; Gordon & Rog, 2023). In plants, the SC comprises a lateral element involving the HORMA domain‐containing protein ASY1 (Armstrong et al., 2002) as well as ASY3 (Ferdous et al., 2012) and ASY4 (Chambon et al., 2018) core proteins, and a central region involving the transverse filament protein ZYP1 (Higgins et al., 2005) and the central element proteins SCEP1, SCEP2, and SCEP3 (Feng et al., 2025; Seear et al., 2025; Vrielynck et al., 2023).
To date, the effects of introgressions derived from wild species on meiosis in inter‐varietal wheat hybrids remain largely unexplored, with most existing studies confined to the cytogenetic characterization of either the materials carrying these introgressions or the introgressed segments themselves (Gao et al., 2021; Tanguy et al., 2005). In this study, we analyzed the French wheat variety Renan (Re), carrying two Aev translocated segments (2AS/2NS and 7DL/7DvL), the reference cultivar Chinese Spring (CS), and the hybrid derived from their cross (CSRe). We evaluated the functional impact of the introgressed segments on reproductive performance by measuring floret fertility and pollen viability. We further assessed meiotic stability and CO formation, and we generated the first cytological meiotic atlas of the wild tetraploid Aev. Finally, we applied oligonucleotide Fluorescent In Situ Hybridization (oligo‐FISH) using introgression‐specific probes to precisely localize alien chromosomal segments in inter‐varietal wheat hybrids. Our study shed light on the meiotic behavior of introgressed segments in inter‐varietal wheat hybrids and offers valuable insights for optimizing alien gene introgression in wheat breeding strategies.
RESULTS
Floret fertility and pollen viability are reduced in CSRe inter‐varietal hybrids
In a first step, floret fertility of the first and second florets was assessed in CS, Re, and CSRe genotypes (n = 3), using 150 spikelets (300 florets) per genotype. Both CS and Re exhibited complete fertility with no variation among replicates in the assessed florets (SD = 0; Figure 1a). In contrast, CSRe exhibited a slight reduction in fertility, with one or two florets per counted spikelet failing to develop into seeds, resulting in a mean of 97.33 ± 1.53% (Figure 1a; Figure S1a). Pairwise t‐tests with Bonferroni correction indicated that CSRe differed significantly from both parental lines (P < 0.05), whereas no significant difference was detected between CS and Re (Figure 1a).
*Comparative analysis of floret fertility (a) and pollen viability (b).Bars represent mean ± SD from multiple replicates. Pairwise comparisons were performed using t‐tests with Bonferroni correction; significant differences are indicated above the bars (*P ≤ 0.05; **P ≤ 0.001; ns, not significant).
Pollen viability was subsequently evaluated in CS, Re, and the CSRe hybrid, using approximately 9000 pollen grains per genotype. CS and Re both showed predominantly viable pollen grains, whereas CSRe exhibited a higher proportion of non‐viable pollen grains (Figure S1b). Pollen viability reached 99.13% ± 0.51 in CS and 99.24% ± 0.16 in Re, whereas it was significantly reduced to 86.37% ± 1.89 in the CSRe hybrid (Figure 1b). Pairwise t‐tests with Bonferroni correction demonstrated that pollen viability in CSRe was significantly lower than in both CS and Re (P < 0.001), while no significant difference was observed between CS and Re (P = 1, ns; Figure 1b). Since pollen grains develop from microspores produced during meiosis, the reduced pollen viability observed in CSRe hybrid may result from abnormalities during meiosis or from subsequent post‐meiotic processes.
Meiotic behavior is affected in CSRe inter‐varietal hybrids
To investigate the causes of reduced pollen viability in CSRe, we investigated the dynamic behavior of chromosomes during meiosis. We constructed a comprehensive meiotic atlas of CS, Re, CSRe, and the alien‐segment donor, the wild tetraploid Aev (Figure 2; Figure S2a). Both CS and Re displayed the same typical meiotic behavior throughout the entire process, with proper chromosome pairing and synapsis during prophase I (Figure S2a), resulting in the consistent formation of 21 bivalents at metaphase I, followed by balanced chromosome segregation during anaphase I and II. This ultimately led to the formation of tetrads containing four evenly balanced nuclei (Figure 2), in agreement with previous reports (Bazile et al., 2024; Benyahya et al., 2020). Physical connections between homologous chromosomes are established through one or several chiasmata that are visible at metaphase I and which represent the cytological evidence of reciprocal exchanges between homologous chromatids (CO; Mercier et al., 2015). Ring bivalents typically contain two chiasmata (at least one CO on each chromosome arm), whereas rod bivalents have a single chiasma reflecting the occurrence of CO on only one arm. In contrast, univalents fail to pair and recombine and do not form chiasmata. These three meiotic chromosome configurations (ring bivalents, rod bivalents, and univalents) are visually defined and classified in Figure S3 based on their morphology and number of chiasmata. At metaphase I, CS exhibited slightly more rod bivalents than Re (0.94 ± 1.06 versus 0.63 ± 0.81; Table 1), resulting in a slightly lower number of chiasmata per cell (41.06 ± 1.06 in CS versus 41.37 ± 0.81 in Re; Table 1; n = 119 and 127 cells analyzed for CS and Re, respectively). Both parental lines showed highly stable chromosome configurations, with CS and Re‐forming predominantly ring bivalents (respectively 20.06 ± 1.06 and 20.37 ± 0.81; Table 1). Furthermore, we investigated meiotic progression in Aev, where chromosome bridges were detected at both telophase I and telophase II (Figure 2). However, as expected, 14 bivalents (13.06 ± 1.00 rings and 0.93 ± 1.00 rods) were consistently observed leading to an average of 27.06 ± 1.00 chiasmata per cell (n = 111 cells; Table 1).
Progression of meiosis.Meiotic atlas showing the complete progression from leptotene to tetrad formation visualized by DAPI fluorescence (white). Magenta arrows indicate rod bivalents, blue circles indicate univalents, orange arrows indicate chromosome bridges, and orange rectangles highlight chromosome fragments. Scale bars = 10 μm.
In contrast, the CSRe hybrid displayed notable differences compared to its parents. CSRe had significantly less ring bivalents (16.96 ± 1.53), more rod bivalents (3.65 ± 1.45; Figure 3a; Table 1) and univalents (0.78 ± 1.24; Table 1), leading to a significantly lower average of chiasmata per cell (37.57 ± 1.84; P < 0.001; Figure 3b; Table 1; n = 126 cells analyzed). Chromosome fragmentation and chromosome bridges were consistently observed in CSRe from anaphase I to the tetrad stage, thereby disrupting normal cell division (Figure 2). At the tetrad stage, chromosome fragmentation was observed in 10 out of 36 (27.78%) and 11 out of 45 (24.44%) cells of CSRe‐3 and CSRe‐5 plants, respectively (Table S1). Triad cells were detected in two cells of CSRe‐3 and one cell of CSRe‐5, all of which also exhibited chromosome fragmentation. Quantitative analysis indicated that 26.11% ± 2.36 of the tetrads (Table S1), including three triad cases, displayed chromosome fragmentation (Figure S2b).
*Quantification of rod bivalents (a) and chiasmata (b) at metaphase I.Each violin plot illustrates the overall distribution of rod bivalents/chiasmata numbers, with embedded boxplots indicating the median and interquartile range, and individual dots represent observations for each meiocyte. The X‐axis shows the genotypes (with sample size indicated in brackets), and the Y‐axis represents the number of chiasmata/rod bivalents per meiocyte. Because the data did not follow a normal distribution, statistical comparisons were performed using the non‐parametric Kruskal–Wallis test, followed by Dunn's post hoc test with Bonferroni correction. Statistical significance is indicated as follows: ns, not significant; **P ≤ 0.001.
Accurate cytogenetic detection of 2AS/2NS and 7DL/7DvL introgressions
To further investigate the pairing behavior of chromosomes carrying introgressed segments (2AS/2NS and 7DL/7DvL translocations in Re) during meiosis, we performed oligo‐FISH analysis at metaphase I stage using oligonucleotide probes specifically targeting these two introgressions. To validate the specificity of the oligo‐FISH probes targeting chromosomes 2 and 7, their hybridization patterns were analyzed in CS, Re, and the CSRe hybrid (Figure S4). Three sets of fluorescently labeled oligoprobes were designed for each chromosome (see Material and Methods). For chromosome 2 (CHR‐2), the 2A‐specific probe, targeting a CS‐specific 29.30 Mb region, hybridized exclusively to CS, whereas the 2N‐specific probe, targeting a Re‐specific 34.10 Mb region, hybridized exclusively to Re. The 2C common probe, marking a 30.00 Mb conserved region between CS and Re, labeled the two parental genotypes, confirming the homology of the chromosomal regions between CS and Re (Table S2; Figure S4a).
Similarly, for chromosome 7 (CHR‐7), the 7D‐specific probe, targeting a CS‐specific 28.90 Mb region, labeled CS only, while the 7Dv‐specific probe, targeting a Re‐specific 31.20 Mb region, labeled Re only. The 7C common probe, which targets ~33 Mb conserved regions shared between CS and Re, labeled the two parental genotypes as well, confirming the 7D‐chromosomal specificity of the probes and the existence of homologous regions between CS and Re (Table S2; Figure S4b).
We therefore had appropriate tools to follow the behavior of chromosomes with introgressions during meiosis in CS, Re, and CSRe.
Rod bivalents are not associated with the presence of introgressions in CS and Re
To determine whether chromosomes carrying introgressions formed rod bivalents more frequently, we analyzed 50 meiotic cells at metaphase I stage from CS and Re parents, and we selected those that contained at least one pair of rod bivalents (Table 2).
In CS, 15 cells (30%) displayed only ring bivalents and 35 cells (70%) contained between one and three rod bivalents (Figure 4a). In these 35 cells, 97.14% (34/35) showed the 2A‐specific and 2C‐common probes on ring bivalents (CS type 1; Figure 4a), whereas only one cell (2.86%, 1/35) showed these probes at both ends of the rod bivalent. This configuration was classified as CS type 2 (Figure 4a). In Re, 19 cells (38%) exhibited ring bivalents only while 31 cells (62%) contained between one and three rod bivalents (Figure 4a). All 31 cells containing at least one rod bivalent showed both the 2N‐specific and 2C common probes on ring bivalents (Re type 1; Figure 4a), with none detected on rod bivalents.
*Oligo‐FISH‐based cytogenetic characterization of chromosome 2 (CHR‐2) and chromosome 7 (CHR‐7) at metaphase I.Fluorescence oligo‐FISH showing CHR‐2–specific (a) and CHR‐7–specific (b) signals at metaphase I meiocytes of CS, Re, and CSRe. Each meiocyte is shown with merged fluorescence signals in paired images: the left panel shows the overall view of all chromosomes, and the right panel shows a magnified view of CHR‐2 and CHR‐7 regions. Yellow boxes indicate chromosome regions carrying CHR‐2/CHR‐7 signals. Magenta arrows indicate rod bivalents. Chromosomes were stained with DAPI (white); probe colors correspond to genomic origins: orange, CS‐specific probes (2A, 7D); cyan, Re‐specific probes (2N, 7Dv); magenta, probes targeting regions shared between CS and Re (2C, 7C). CHR‐2/CHR‐7 probe localization is observed on ring bivalents, rod bivalents, and univalents. Scale bar = 10 μm.(c) Distribution of CHR‐2‐specific probe signals on rod bivalents and/or univalents at metaphase I meiocytes. The bar plot shows the percentage of metaphase I meiocytes containing at least one rod bivalent with CHR‐2 probe signals detected on rod bivalents and/or univalents in CS and CSRe. The Re genotype is included at 0% and is shown as a line without filled color. Pairwise comparisons were performed using Fisher's exact test, with significance indicated by asterisks (ns: not significant; **P ≤ 0.001). The X‐axis represents genotypes, and the Y‐axis indicates the percentage of meiocytes with CHR‐2 signals on rod bivalents and/or univalents.
We applied the same approach for 7DL/7DvL introgression. Similarly, further analysis revealed that in cells containing at least one rod bivalent among the 21 bivalents (CS, 33 cells; Re, 29 cells; Figure 4b; Table 2), the 7D/7Dv/7C segments were consistently located within ring bivalents (CS type 1 and Re type 1; Figure 4b), and no localization on rod bivalents was observed, which is similar to the pattern observed for 2A/2N/2C‐specific probes (Figure 4a; Table 2). This indicates that in the Re parental line, even though the chromosomes 2A and 7D carry ~30 Mb alien segments from Aev, these two chromosomes do not give rod bivalent more frequently than other chromosomes and the same holds true for CS without any introgression.
Differential effects of 2AS/2NS and 7DL/7Dv introgessions on rod bivalent frequency in CSRe inter‐varietal hybrids
In CSRe, 50 meiotic cells were analyzed, all of which contained between one and 10 rod bivalents, which was more than three times higher than that observed in the two parents (Figure 4a,b; Table 2). Oligo‐FISH observations focused on these cells to assess the positioning of specific probes in the presence of rod bivalents. Interestingly, the 2A, 2N, and 2C labeled regions exhibited four distinct localization patterns among the 50 cells. First, in 16 cells (32%), 2A, 2N, and 2C probes were closely positioned and located on ring bivalents, a configuration we refer to as CSRe type 1. Second, in 13 cells (26%), the 2A, 2N, and 2C probes were positioned on rod bivalents, with 2A/2N located at each opposite end. We refer to this configuration as CSRe type 2. Third, in 10 cells (20%), 2A, 2N, and 2C probes were also located on rod bivalents, but in this case, they were closely positioned at the central region of the rod, corresponding to the chiasma. This configuration was termed CSRe type 3. Finally, in 11 cells (22%), the 2A chromosome was unpaired, with 2A/2C and 2N/2C located on two separate univalents, termed CSRe type 4. Pairwise Fisher's exact tests revealed a significantly higher frequency of 2A/2N/2C signal localization on rod bivalents and/or univalents in CSRe compared with CS and Re (P < 0.001 for both comparisons) (Figure 4c; Table 2).
To also examine the behavior of the 7DL/7DvL translocations in CSRe, we applied the same oligo‐FISH strategy. In CSRe, all 50 cells contained at least one rod bivalent (100%; Figure 4b; Table 2). The pattern was markedly different from that of 2A/2N/2C specific probes, with 48 among the 50 cells (96%) showing 7D/7Dv/7C segments within ring bivalents (CSRe type1), while only two cells (4%) displayed 7D/7C and 7Dv/7C at the ends of rod bivalents (CSRe type2). No other localization on CSRe type3 or CSRe type4 was observed (Table 2). These results indicate that, in CSRe, the 7D/7Dv/7C segments are predominantly located within ring bivalents, whereas only a few cells show 7D/7Dv/7C at the ends of rod bivalents and this does not differ from random.
Altogether, our results showed that meiosis in the CSRe hybrids between Re with 2AS/2NS and 7DL/7DvL and CS lacking these two segments is disturbed with more univalents and more rod bivalents at the meiotic stage in hybrids. However, among the two chromosomes (2A and 7D), only 2A is more often affected by irregular pairing behavior (rod bivalent or univalent), while 7D appears not affected.
DISCUSSION
Introduction of diversity from wild relatives into the wheat genomes has been widely used in the past decades (Tian et al., 2025; Wulff & Moscou, 2014). However, once introgressed, the fragments are fixed and the occurrence of CO events appears impossible (Danguy des Déserts et al., 2021; Naranjo, 2019). This is problematic because of linkage drag that may introduce deleterious traits together with the favorable ones. In our study, we focused on two well‐known introgressions from Aev (2AS/2NS and 7DL/7DvL translocations) and we studied their effect and behavior during meiosis in hybrids derived from a cross between wheat varieties with or without these two introgressions.
Aev exhibits a diploid‐like meiosis
As an allopolyploid species, we evaluated the meiotic behavior of Aev. We observed regular meiosis, showing 14 bivalents (13.03 ± 0.94 ring and 0.97 ± 0.94 rod bivalents), and no univalents were observed confirming previous observations (Baik et al., 2017; Pignone et al., 1994). The proportion of rod bivalents is similar to that observed in bread wheat CS (21 bivalents, 0.94 ± 1.06 rods) and slightly higher than in Re (21 bivalents, 0.63 ± 0.81 rods). We constructed the first comprehensive cytological atlas of meiosis in Aev (and indeed in any Aegilops species) which allowed us to clearly observe chromosome bridges at both telophase I and telophase II (Figure 2). Such bridges were not observed in CS or Re. This suggests that the regulation of meiosis in Aev may be less tightly controlled than in bread wheat and this could concern homoeologous recombination. In bread wheat, homoeologous recombination is mainly controlled by the Ph1 (TaZip4‐B2) and Ph2 (TaMsh7‐3D) loci (for a review, see Sourdille et al., 2025). A similar control in Aev is likely since a diploid‐like behavior is observed but the factors involved remain unknown. However, such control may be weaker, potentially leading to reduced fidelity of homologous recognition during meiosis. For instance, Ph1 is suggested to favor homologous synapsis rather than actively preventing homoeologous synapsis (Martín et al., 2021). Because Ph1 is likely absent in Aev, homologous recombination may be less strongly promoted, resulting in reduced synapsis fidelity. At least MSH7 genes also exist in Aev genome and the resulting proteins may constitute a first player in the process (Serra et al., 2021). Our work fills a gap in meiotic research on this species but it also provides a solid foundation for a deeper understanding of its chromosome behavior during meiosis. In addition, our work will facilitate future efforts to explore and to utilize Aev as an interesting source of diversity to improve agronomical traits for wheat.
Hemizygous Aev introgressions disturb hybrids meiosis
The primary genomic difference between CS and Re varieties lies in the presence of two Aev chromosomal introgressions in the Re genome on chromosomes 2A and 7D (The International Wheat Genome Sequencing Consortium (IWGSC), 2018; Aury et al., 2022). However, both lines exhibit normal meiotic behavior with 21 bivalents at metaphase I and balanced chromosome segregation throughout the subsequent stages of meiosis. The resulting meiotic products comprised four nuclei (tetrads) with balanced genetic material consistent with the typical meiotic characteristics of bread wheat, and fertility is normal in both lines. This suggests that when the introgressions are present at a homozygous state, this disturbs neither the stability of the meiotic process nor the fertility of the varieties compared to accessions without these two introgressions.
On the contrary, in hybrids between CS without introgressions and Re with the two introgressions, meiosis was severely hampered with significantly higher numbers of rod bivalents as well as univalents leading to a lower number of chiasmata (Figures 2 and 3). Occurrence of univalents is the most problematic aspect, as it is likely to lead to aneuploid gametes. This could explain the reduced fertility observed in the hybrid compared to parental lines, as well as the presence of non‐viable pollen grains (Figure 1; Figure S1). These results suggest that the introgression from wild species may have a subtle but detectable impact on the reproductive performance of inter‐varietal hybrids. Several hypotheses can be proposed to explain this phenomenon:
- the combination of genes (or alleles) involved in the meiotic process coming from these two varieties is inappropriate leading to a misfunction of the whole process and to a disturbed meiosis. Some of these genes could be located on introgressions. It is well known in plants that a variation of recombination rate exists between varieties (Bauer et al., 2013; Dluzewska et al., 2023; Jordan et al., 2018; López et al., 2012; Mikhailov et al., 2025) and that heterozygosity induced by hybridization can affect recombination as well (Ziolkowski et al., 2015). This variation of recombination rate can be due either to a variation in alleles of genes controlling meiosis or to various genomic features (including introgressions) or chromatin structure such as epigenetic landmarks (Akhunov et al., 2003; Liu et al., 2009; Melamed‐Bessudo et al., 2016; Rodgers‐Melnick et al., 2015; Shilo et al., 2015; Wijnker et al., 2013).
- the presence of gene(s) on at least one of the two introgressions that affect(s) meiosis through a trans‐effect. Such a trans‐effect has already been described in maize (Yandeau‐Nelson et al., 2006). Using the same line (A1‐LC Sh2) crossed with three different accessions (A632, Oh43, W64A) that were recessive at the Shrunken locus (a1::rdt sh2), they observed that genetic distances across a1–sh2 varied almost twofold between Oh43/LC (0.081 cM) and the other two varieties (A632/LC 0.143 cM and W64A/LC 0.126 cM). They conclude that some modifiers of the meiotic recombination process can act in trans, depending on the genetic background. However, if this is the case, the fact that it affects 2A/2N introgression only and not the two introgressions remains to be elucidated.
- the sequence divergence between the introgression is too important or there are chromosomal rearrangements between the two regions and either of these prevents recombination to occur. This was illustrated in Arabidopsis where there is a heterochromatic knob (hk4S) on chromosome 4 with an inversion in Columbia‐0 (Col‐0). This inversion does not exist in Landsberg‐1 (Ler‐1) and recombination is prevented in Col‐0 × Ler‐1 crosses but restored when the knob is reversed in Col‐0 (Schmidt et al., 2020).
Most importantly is the occurrence of univalents which can result in the production of poorly fertile aneuploid lines. Whether this relies on the presence of the introgression remains to be elucidated. It would be interesting to develop more crosses between Re and other varieties without the same introgression including its parents (Courtot, Mironovskaia, Maris Huntsman; Cadalen et al., 1997; Boeuf et al., 2003) to minimize the effect of the rest of the genetic background and to look at the fertility of the hybrids as well as their meiotic behavior. If univalents are frequently observed, then we could suspect that this is due to the presence of the introgressions and maybe to some genes it could carry.
Meiotic structure of chromosome with 2AS/2NS introgression is more often affected than chromosome with 7DL/7DvL introgression in CSRe hybrids
CS and Re parental lines gave a classical metaphase I plate with 21 bivalents, and these are mostly ring bivalents. Using the oligo‐FISH approach with sets of fluorescent probes designed against the regions of the translocations, we did not detect any bias toward chromosomes 2A and 7D bearing the introgressions in Re. These two chromosomes occur as classical ring bivalents in both lines (Figure 4; Table 2). However, the results were significantly different in CSRe. 2AS/2NS introgression showed unexpected structures (rod bivalents or univalents) more frequently than expected by chance (68% in hybrids while 1/21 ~5% expected by chance), while this was not the case for the chromosome with 7DL/7DvL introgression (4% in hybrids; Figure 4; Table 2). This can be explained by the highest similarity that exists between chromosomes from the D genome. The D genome is known for long as less polymorphic than the A and B genomes leading to difficulties in elaborating inter‐varietal genetic maps for the D chromosomes (Cadalen et al., 1997). Because of the better similarity between the chromosomes 7D of CS and Re, it is likely that synapsis is easier and more efficient resulting in classical ring bivalents for this chromosome.
On the contrary, because of a likely lower similarity together with the presence of the introgression on the distal region of chromosome 2A in Re, probably synapsis is more difficult in hybrids. Usually, distal regions are associated with high CO frequency (Choulet et al., 2014; The International Wheat Genome Sequencing Consortium (IWGSC), 2014; The International Wheat Genome Sequencing Consortium (IWGSC), 2018). Our results suggest that even within recombination‐prone distal regions, the alien segment on 2A may undergo CO failure in the inter‐varietal hybrid background. In wheat, telomeres form a bouquet‐structure in early steps of meiosis and synapsis is initially polarized to the distal regions of the chromosomes (Osman et al., 2021). Presynaptic ZYP1 foci appear just after bouquet formation and synapsis starts in the distal regions of the chromosomes at early zygotene. This results in mostly distal CO events (Saintenac et al., 2009; The International Wheat Genome Sequencing Consortium (IWGSC), 2014; The International Wheat Genome Sequencing Consortium (IWGSC), 2018). Because of the presence of the 2AS/2NS introgression coming from a wild species and because of the presence of the dominant alleles at Ph1 and Ph2 loci controlling homoeologous recombination (Martín et al., 2021; Serra et al., 2021), maybe distal initiation of the synapsis becomes less efficient in hybrids resulting in less ring bivalents. However, albeit synapsis initiates in sub‐telomeric regions, small foci and short stretches of ZYP1 also appear later in the interstitial regions (Osman et al., 2021). This may explain the fact that we observed almost the same proportion of rod bivalents with a chiasma at the level of the introgression (20% of the cells) or on the opposite long arm (26% of the cells; Table 2). Moreover, chiasmata just reflect the occurrence of a CO within the region. But because of the low resolution we have, it is not possible to see whether the CO is located within or outside the introgression. Based on our genetic map derived from the cross between CS and Re (Danguy des Déserts et al., 2021; The International Wheat Genome Sequencing Consortium (IWGSC), 2014; The International Wheat Genome Sequencing Consortium (IWGSC), 2018), it is likely that there is no CO within the introgression and that the chiasmata we observed in the distal region of the short arm of chromosome 2A bearing the introgression in Re rather reflect a CO in its vicinity. Combining oligo‐FISH with MLH1 or HEI10 immuno‐FISH could help to solve this question.
To conclude, even if introducing new diversity from wild species remains an interesting approach to revitalize the wheat gene pool, introgressions may still have deleterious effects on both agronomical traits (because of linkage drag) but also on fertility and meiotic behavior. Our finding provides important insights for wheat breeding practices, underscoring the need to closely monitor the heterozygous background of alien chromosomal segments and their potential impact on meiotic stability, in order to avoid genetic barriers and the loss of valuable genetic resources caused by chromosomal abnormalities. Moreover, it lays a solid cytogenetic foundation for advancing fundamental research on the integration and stability of alien chromosomes in crops, thereby promoting both theoretical investigations and applied explorations in related fields. There is therefore a necessity to continue to develop research to understand this tricky phenomenon and the behavior of introgressions once introduced to help develop new and productive wheat varieties facing the challenges of sustainable agriculture and of climate change.
EXPERIMENTAL PROCEDURES
Plant material and growth conditions
Two cultivated hexaploid wheat were used in this study: (1) Re, a French winter wheat cultivar carrying two well‐characterized introgressed segments derived from Aev, including a 34 Mb translocated segment from chromosome 2N on chromosome 2A (2AS/2NS translocation) harboring disease resistance genes Yr17, Lr37, Sr38, and Cre5, and a 30 Mb segment on chromosome 7D (7DL/7D_V_L translocation) carrying the Pch1 gene. A fully assembled and annotated genome sequence of Re has been recently produced (Aury et al., 2022), enabling detailed molecular characterization of these introgressions. (2) CS, the international reference genotype widely used in genetic and genomic studies, serves as the wheat reference line. A high‐quality reference genome sequence is publicly available for this cultivar (The International Wheat Genome Sequencing Consortium (IWGSC), 2018; Zhu et al., 2021). (3) F_1_ hybrids CSRe were generated through a cross between CS and Re to investigate the meiotic behavior of introgressed segments in a heterozygous background. Additionally, (4) the wild tetraploid species Aev accession n°10, the exact donor of the alien chromatin segments introgressed into Renan, was included as a reference for probe design and cytogenetic analyses.
Seeds were germinated in a growth chamber at 20°C. At the three‐leaf stage, seedlings underwent vernalization at 6°C ± 1°C for 6–8 weeks under an 8 h light/16 h dark photoperiod. After vernalization, plantlets were transplanted into 4‐L pots containing Nutricote (Fertil), a commercial slow‐release fertilizer, and grown in a greenhouse under a 16 h light/8 h dark photoperiod at 22°C ± 1°C during the day and 16°C ± 1°C at night.
Cytological analysis
Cytological analyses of developing meiocytes were performed as previously described (Bazile et al., 2024; Benyahya et al., 2020; Serra et al., 2021). Briefly, meiotic stages were determined by acetocarmine staining (10 g L^−1^ carmine, 45% acetic acid). Synchronized anthers were fixed in Carnoy's solution (100% ethanol and glacial acetic acid, 3:1, v/v), then transferred to 70% ethanol and stored at 4°C for subsequent analyses. For meiotic atlas generation, anthers at selected developmental stages were dissected on poly‐L‐lysine‐coated slides in acetocarmine staining solution. After removing the stain with 45% acetic acid, slides were flash‐frozen in liquid nitrogen, air‐dried, and mounted with Vectashield‐DAPI (Eurobio Ingen, Les Ulis, France). Images were captured using a ZEISS Axio Observer Z1 fluorescence microscope with Zen software (Carl Zeiss Microscopy, Jena, Germany). For cytological analysis at metaphase I, chromosome spreads were prepared following established protocols (Jahier & Tanguy, 1992; Chen et al., 1995, 2001). Pollen mother cells (PMCs) were released by gently squashing anthers in acetocarmine staining solution. After removing debris, slides were briefly heated on a hotplate (90°C) to facilitate chromosome separation, then treated with 45% acetic acid and firmly pressed to spread chromosomes. Chromosome configurations were observed under brightfield using a ZEISS Axio Observer Z1 microscope, and chiasma frequencies were scored from at least 50 PMCs per sample for statistical analysis. Each genotype was analyzed with two independent biological replicates. If no significant differences were observed between the replicates, the data were combined for the final statistical analysis. Images of CS, Re, and CSRe genotypes were anonymized using the Blind Analysis Tools plugin for ImageJ to avoid bias during image analysis.
Pollen viability and fertility
Pollen viability and fertility were assessed following well‐established protocols (Jahier & Tanguy, 1992; Bazile et al., 2024; Benyahya et al., 2020). Fresh pollen from 2 to 3 anthers was stained with Alexander reagent and examined with a Zeiss Axio Observer microscope. Pollen viability was quantified using a semi‐automated ImageJ protocol, with viable grains identified based on grayscale intensity (0–127), size (≥50 pixels), and circularity (0.8–1.0). These standardized digital criteria minimized observer bias, and non‐viable grains were manually cross‐verified. Approximately 1500 grains per replicate were scored across three biological and two technical replicates per genotype, totaling approximately 9000 grains. To assess floret fertility in each genotype, at least three plants were analyzed. For each plant, 5 primary spikes were selected, and 10 central spikelets were chosen from each spike. Only the first and second florets of each spikelet were scored. The number of florets setting seeds was recorded, and the total number of florets analyzed was calculated. Floret fertility was expressed as the percentage of florets setting seeds out of the total number of florets analyzed.
Design of oligonucleotide (oligo) probes
To refine the identification of introgressed chromosomal segments, three sets of oligo probes (45 nucleotides long) were designed and synthesized by Arbor Biosciences (Ann Arbor, MI, USA). The probes were developed based on the IWGSC RefSeq v2.1 assembly for CS (The International Wheat Genome Sequencing Consortium (IWGSC), 2018) and the Re draft genome (Aury et al., 2022), targeting specific or shared genomic regions between the two genotypes. For 2AS/2NS translocation, three probe sets were designed: (i) 2A‐specific probes (CS, 0–29.30 Mb; 13 000 oligos), (ii) 2N‐specific probes (Re, 0–34.10 Mb; 19 301 oligos), and (iii) 2C‐common probes corresponding to the regions shared between CS (29.50–59.50 Mb) and Re (34.30–64.40 Mb; 9096 oligos). For 7DL/7DvL translocation, three additional probe sets were designed: (i) 7D‐specific probes (CS, 614.00–642.90 Mb; 6788 oligos), (ii) 7Dv‐specific probes (Re, 617.50–648.70 Mb; 9128 oligos), and (iii) 7C‐common probes corresponding to the conserved regions between CS (580.00–613.00 Mb) and Re (583.40–616.90 Mb; 11 720 oligos). Each probe set was labeled with a distinct fluorophore: Alexa Fluor 488 (green) for Re‐specific probes (2N and 7Dv), ATTO 550 (red) for CS‐specific probes (2A and 7D), and ATTO 647 (far‐red) for common probes (2C and 7C).
Oligo‐FISH procedure
Oligo‐FISH experiments were performed on chromosome spreads prepared as described above, following the method of Hao et al. (2011) with specific modifications. Prior to hybridization, slides were pre‐dried and then subjected to fixation, washing, dehydration, and DNA denaturation according to standard procedures. The hybridization mixture (50 μl per slide), containing 100% ultrapure formamide, 20 × SSC, 50% dextran sulfate, 20% SDS, 10 pmol of oligo probe, and sonicated Salmon DNA (<2 kb, 10 μg μl^−1^), was denatured and immediately cooled before application to the slides. Hybridization was carried out in a humidified chamber at 37°C for 12–16 h, followed by post‐hybridization washes to remove unbound probes. Slides were then mounted with Vectashield‐DAPI (Eurobio Ingen). Each genotype was analyzed using two to three independent biological replicates, and the data were pooled for final statistical analysis.
Confocal microscopy and image analysis
Fluorescence images of meiocytes were acquired using a Carl ZEISS LSM 800 confocal microscope with the high‐resolution AIRYSCAN module across four channels (358 488, 568, and 647 nm). Image acquisition and analysis were performed with Zeiss ZEN2 software using optimized intensity settings for each channel. Final image processing was conducted on the OMERO platform (https://omero.mesocentre.uca.fr/webclient/), and fluorescence channels were adjusted to improve visibility for individuals with color vision deficiencies: Alexa Fluor 488 (green) was displayed as cyan, ATTO 550 (red) as orange, and ATTO 647 (far‐red) as magenta.
Statistical analysis
All statistical analyses were performed in R software (version 4.4.1). Descriptive statistics were calculated to summarize the data, and for each group, the mean and standard deviation (SD) were computed. Normality of continuous data was assessed using the Shapiro–Wilk test, and homogeneity of variances was evaluated with Levene's test (median‐centered) in the car package. The consistency between the two biological replicates was assessed using the Wilcoxon test for each genotype. For datasets meeting these assumptions, group comparisons were conducted using Student's t‐test (α = 0.05). For datasets that violated normality or variance homogeneity, overall differences were first tested using the Kruskal–Wallis rank sum test, and significant results (P ≤ α) were followed by Dunn's post hoc test with multiple testing correction. Fisher's exact test was used to evaluate associations between categorical variables, as some groups had very low or zero counts. Significance was defined as P ≤ α.
AUTHOR CONTRIBUTIONS
PS conceived the experiments, acquired funding, and supervised the study. PS, IN, LY, and FC participated in the experimental design. PS and IN contributed to the oligo‐probe design strategy. LY performed all experiments, analyzed the data, drafted the figures, tables, and the manuscript. MH elaborated the oligo‐FISH procedure; IN and FC assisted with experiments, data analysis, and figure preparation. PS wrote the core of the manuscript, and IN, FC, MH, and DL assisted in writing the manuscript. All authors contributed to the final reading and approved the submitted version.
CONFLICT OF INTEREST
The authors declare no competing interests.
Supporting information
Figure S1. Visual comparison of spike morphology and pollen viability in CS, Re and CSRe.
Figure S2. Meiotic chromosome dynamics during prophase I and at the tetrad stage.
Figure S3. Meiotic chromosome configurations in wheat genotypes.
Figure S4. Validation of the specificity of oligo‐probes for chromosomes 2A and 7D.
Table S1. Frequency of chromosome fragmentation at the tetrad stage in CSRe.
Table S2. Chromosome‐specific and common fragments identified in chromosome 2A (CHR‐2) and chromosome 7D (CHR‐7).
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1Akhunov, E. , Goodyear, A. , Geng, S. , Qi, L.L. , Echalier, B. , Gill, B.S. et al. (2003) The organization and rate of evolution of wheat genomes are correlated with recombination rates along chromosome arms. Genome Research, 13, 753–763.12695326 10.1101/gr.808603 PMC 430889 · doi ↗ · pubmed ↗
- 2Armstrong, S. , Caryl, A. , Jones, G. & Franklin, F.C.H. (2002) Asy 1, a protein required for meiotic chromosome synapsis, localizes to axis‐associated chromatin in Arabidopsis and Brassica . Journal of Cell Science, 115, 3645–3655.12186950 10.1242/jcs.00048 · doi ↗ · pubmed ↗
- 3Aury, J.M. , Engelen, S. , Istace, B. , Monat, C. , Lasserre‐Zuber, P. , Belser, C. et al. (2022) Long‐read and chromosome‐scale assembly of the hexaploid wheat genome achieves high resolution for research and breeding. Giga Science, 11, giac 034.35482491 10.1093/gigascience/giac 034PMC 9049114 · doi ↗ · pubmed ↗
- 4Baik, N. , Maamri, F. & Bandou, H. (2017) Karyological study and meiotic analysis of four species of Aegilops (Poaceae) in Algeria. Caryologia, 70(4), 1.
- 5Bailey‐Serres, J. , Parker, J.E. , Ainsworth, E.A. , Oldroyd, G.E.D. & Schroeder, J.I. (2019) Genetic strategies for improving crop yields. Nature, 575, 109–118.31695205 10.1038/s 41586-019-1679-0PMC 7024682 · doi ↗ · pubmed ↗
- 6Bauer, E. , Falque, M. , Walter, H. , Bauland, C. , Camisan, C. , Campo, L. et al. (2013) Intraspecific variation of recombination rate in maize. Genome Biology, 14, R 103.24050704 10.1186/gb-2013-14-9-r 103PMC 4053771 · doi ↗ · pubmed ↗
- 7Bazile, J. , Nadaud, I. , Lasserre‐Zuber, P. , Kitt, J. , De Oliveira, R. , Choulet, F. et al. (2024) Ta RECQ 4 contributes to maintain both homologous and homoeologous recombination during wheat meiosis. Frontiers in Plant Science, 14, 1342976.38348162 10.3389/fpls.2023.1342976 PMC 10859459 · doi ↗ · pubmed ↗
- 8Benyahya, F. , Nadaud, I. , Da Ines, O. , Rimbert, H. , White, C. & Sourdille, P. (2020) SPO 11.2 is essential for programmed double‐strand break formation during meiosis in bread wheat (Triticum aestivum L.). The Plant Journal, 104, 30–43.32603485 10.1111/tpj.14903 · doi ↗ · pubmed ↗