Wheat (Triticum aestivum L.) is one of the most important cereals for human nutrition worldwide. Its grain is used to make bread, pasta, and various bakery products, while its straw is suitable for animal feed and the production of second-generation biofuel (Glithero et al., 2013). By 2050, crop yields will need to be significantly increased due to the world’s rapidly growing population (Ray et al., 2013) and climate change. However, the genetic variability of wheat germplasms has narrowed over thousands of years of cultivation and modern agricultural practices (Jiang et al., 1994; Roussel et al., 2005; van de Wouw et al., 2010), and finding optimal allele combinations within the pangenome of wheat for producing high-yielding cultivars with good nutritional quality and improved heat and drought tolerance, as well as resistance to new pathogen races, remains a major challenge (Moore, 2015).

Hybridizations between bread wheat and related species from primary, secondary, or tertiary gene pools provide a way to increase wheat genetic diversity through chromosome-mediated transfer of wild alleles not existing in cultivated wheat (Jauhar and Chibbar, 1999; Molnár-Láng et al., 2015).

Secale strictum (C. Presl) C. Presl (syn. S. montanum Guss.) (2n = 2x = 14, RR) is a wild perennial relative of cultivated rye (S. cereale ssp. cereale) possessing remarkable genetic diversity (Schreiber et al., 2019), high tillering ability even under suboptimal nutrient availability or water supply, and tolerance to frost (Wang, 1987), drought (Oram, 1996), and high soil aluminum and manganese content (Culvenor et al., 1986). This species is also resistant to rust diseases (Lukaszewski et al., 2001; Schneider et al., 2016) and has a high grain protein content (Kubiczek et al., 1981). Although S. strictum is not suitable for cultivation as a forage crop due to its small, scant leaves and “brittle ear” characteristics (Akgün and Tosun, 2004), its perennial habit has been useful in projects aimed at breeding perennial rye varieties with low demands for cultivation and combining the favorable traits of cultivated and wild rye (Reimann-Philipp, 1986).

To facilitate the use of wild genetic variation in breeding programs, two crosses between S. strictum and cultivated rye accessions were performed to produce S. cereanum varieties: a tetraploid variety (Permontra) in Germany (Reimann-Philipp, 1986) and a diploid variety (Black Mountain) in Australia (Oram, 1996). ‘Kriszta’, ‘Perenne’, and ‘Gergő’ are Hungarian S. cereanum cultivars descended from the late 1950s interspecific hybrid S. cereale cv. Várda × S. strictum ssp. anatolicum (Kotvics, 1970). ‘Kriszta’ is cold- and drought-tolerant and resistant to fungal diseases (rusts and powdery mildew) and has a high dietary fiber content (Schneider et al., 2016; Szakács et al., 2020). In order to transfer these traits into hexaploid wheat, cv. Kriszta was involved in a Hungarian introgression breeding program. The cross between the winter wheat line Martonvásári 9 kr1 (Mv9kr1) carrying the recessive crossability allele kr1 from the spring wheat variety ‘Chinese Spring’ (Molnár-Láng et al., 1996) and ‘Kriszta’ resulted in numerous backcrossed and selfed offspring. Among them, three addition lines disomic for chromosomes 1R, 4R, and 6R, respectively, and the T1BL.1RS translocation line ‘179’ have been identified and phenotyped for grain quality and resistance to foliar diseases (Schneider et al., 2016; Szakács et al., 2020).

The conventional method for visualizing alien chromatin in a host genome is genomic in-situ hybridization (GISH), in which labeled DNA probe libraries made from the sheared genomic DNA of multiple known genotypes are competitively hybridized to mitotic chromosomes and visualized (Le et al., 1989; Schwarzacher et al., 1989). GISH can therefore provide information on the number of introgressed chromosomes, the presence of interspecific translocations, and the relative length of the alien chromosome segments. Its combination with fluorescence in-situ hybridization (FISH), which relies upon observing chromosomal hybridization profiles of probes designed against specific DNA repeats (Rayburn and Gill, 1985), enables the identification of specific alien chromosomes or chromosome segments in the wheat genetic background (Schneider et al., 2016). In S. cereale, a combination of the DNA probe pSc119.2—a member of a 120-bp repeat family (Bedbrook et al., 1980; McIntyre et al., 1990)—and the microsatellite sequence (AAC)₅ (Cuadrado et al., 2000) is suitable for identifying the entire chromosome complement (Szakács and Molnár-Láng, 2008). However, the high level of polymorphism in the FISH hybridization pattern significantly hampers the identification of the chromosomes of outcrossing rye ‘Kriszta’ carrying both S. cereale and S. strictum chromatin (Szőke-Pázsi et al., 2022).

The in-situ hybridization technique, however, has two significant drawbacks: first, because of its low sensitivity and spatial resolution (deJong et al., 1999), cryptic or micro-introgressions remain undetectable; second, it is laborious and time-consuming, limiting the extent of screening of large wheat-alien prebreeding populations for introgressions.

There are currently several high-throughput and cost-effective molecular techniques available to mitigate these critical limitations. The efficacy of single nucleotide polymorphism (SNP) markers in detecting Aegilops geniculata 5M^g chromatin introgressed into hexaploid wheat has been reported by Tiwari et al. (2015). The 820k exome-based Axiom SNP array (and its 35k subset) (Winfield et al., 2016; King et al., 2017) has been successfully used to detect introgressions from Amblyopyrum muticum (Aegilops mutica; King et al., 2017, 2019), T. monococcum (Cseh et al., 2019a), Th. intermedium (Cseh et al., 2019b), and T. timopheevii (Devi et al., 2019). A general disadvantage of fixed SNP arrays is their reliance on known DNA sequence information, and as a result, their use in exotic species is limited. Moreover, array platforms suffer from poor performance in the detection of rare alleles (Ganal et al., 2012; Negro et al., 2019).

DArTseq technology (Sansaloni et al., 2011) is a publicly available, sequence-independent, high-throughput genotyping-by-sequencing (GBS) platform that combines genome complexity reduction steps developed for the microarray hybridization-based DArT method (Jaccoud et al., 2001; Kilian et al., 2005) with next-generation sequencing (NGS) (Elshire et al., 2011). Since the highly methylated repetitive fraction of the DNA samples is eliminated by a methylation-sensitive restriction endonuclease used in the complexity reduction step of the protocol, the majority of high-quality molecular markers generated are low-copy sequences (gene-rich regions). DArTseq markers manifest as short sequences, which are clustered and classified as either dominant SilicoDArT markers (identical sequences that are either present or absent across genotypes) or co-dominant SNP-based markers (homologous sequences containing homo- or heterozygous SNP variants), allowing simultaneous typing of several thousand polymorphic loci spread over a genome in a single experiment. In the case of the genus Secale, the DArTseq platform has been successfully used for extending the high-density map of rye and mapping the Rfc1 male fertility restorer gene on the 4RL chromosome arm (Milczarski et al., 2016); identification of SNPs associated with α-amylase activity, preharvest sprouting, and leaf rust resistance (Rakoczy-Trojanowska et al., 2017); and localization of a gene encoding the transcription factor GAMYB on the 3R chromosome of rye (Bienias et al., 2020) and a new dw9 dwarfing gene in the 6RL arm (Grądzielewska et al., 2020), alongside the evaluation of genetic diversity in rye inbred lines (Targońska-Karasek et al., 2017) and taxonomic relationships within this genus (Al-Beyroutiová et al., 2016; Targońska-Karasek et al., 2018). Unlike SNP arrays, markers generated by GBS technology, and thus the DArTseq platform, may only be partially reproducible across two genotyping experiments. This issue may become more serious when DArTseq is used to screen different generations of wheat–rye BC populations. The recently published high-quality chromosome-scale genome assembly for rye ‘Lo7’ (Rabanus-Wallace et al., 2021) can overcome these problems by allowing the alignment of DArTseq markers on rye pseudomolecules.

Motivated by the need for a high-throughput and precise screening for the presence of “exotic” rye chromatin in wheat–S. cereanum prebreeding populations, the present study used the DArTseq platform in combination with molecular cytogenetic approaches to identify new introgressions. We demonstrated that the SilicoDArT and SNP markers assigned to rye chromosomes using the ‘Lo7’ pseudomolecules, combined with in-situ hybridization with DNA repeat and rye genomic probes, are suitable for the precise characterization of S. cereale and S. strictum chromatin in a wheat-hybrid BC₂F₈ generation.

The development of high-throughput genotyping-by-sequencing has greatly increased the sensitivity of introgression detection in crop plants (Scheben et al., 2017). A great advantage of the DArTseq genome profiling platform over the hybridization-based DArT method (e.g., microarray-based DArT) is its flexibility, as it does not require prior sequence information. The present study further demonstrated the efficiency of DArTseq technology for screening a wheat introgression breeding population for the presence of exotic rye (a hybrid of S. cereale and S. strictum) chromatin.

By mapping reads generated by high-throughput genotyping systems onto publicly accessible high-quality genome assemblies and chromosome-scale pseudomolecules for Hordeum vulgare (Mascher et al., 2017), Aegilops tauschii (Luo et al., 2017), and several Triticum species (Avni et al., 2017; Appels et al., 2018; Ling et al., 2018; Maccaferri et al., 2019; Zhu et al., 2021), it was possible to identify chromosome rearrangements between and within species.

High-quality chromosome-scale reference genome assemblies for S. cereale have also become available recently for the inbred line ‘Lo7’ (Rabanus-Wallace et al., 2021) and another for the elite Chinese cultivar ‘Weining’ (Li et al., 2021). These genomic resources shed light on the evolution of cultivated rye and the mechanisms of genome structure modifications (Rabanus-Wallace et al., 2021), as well as the physical organization of complex loci and gene expression features in the rye (Li et al., 2021). The availability of Lo7 pseudomolecules also allowed us to determine the genomic positions of DArT marker sequences obtained in the parental and control rye genotypes. In the present study, ~30% of the SilicoDArT markers were located in coding regions, while ~70% of the SNP markers were gene-specific. This ratio was higher than those obtained in common bean, where 43.3% of the SNP DArT markers were gene-specific (Valdisser et al., 2017), or the ratio (~60%) obtained in the genus Psidium (Grossi et al., 2021). In connection with this, we also found a general trend of increasing DArT marker density from centromeric–pericentromeric regions to telomeres. This phenomenon corresponds to the expectation that marker coverage will be higher in distal regions of the chromosomes with active gene expression (Akhunov et al., 2003). However, Bauer et al. (2017) discovered that rye chromosomes contain numerous genes in their centromeric and pericentromeric regions. Given that the methylation-sensitive PstI and frequently cutting TaqI restriction endonucleases used for genome complexity reduction generate low-copy (gene-rich) sequences, markers should theoretically be present in these chromosome regions. The low abundance of the markers in the pericentromeric regions can most likely be explained by the small number of low-copy sequences compared to distal regions. An alternative explanation could be the high methylation level of the proximal pericentromeric regions, which are known to be strongly heterochromatic (Kalinka and Achrem, 2020; Schreiber et al., 2022). The high methylation level in these regions can inhibit the methylation-sensitive PstI function, resulting in a relative lack of cleavage sites and, consequently, a decrease in marker density.

Differences in marker density between the proximal and distal parts of chromosome arms, on the other hand, did not limit the detection of individual rye chromosomes or the identification of small intra- or interchromosomal rearrangements that could not previously be detected by PCR-based markers or in-situ hybridization (Schneider et al., 2016).

Recently, Keilwagen et al. (2019) examined diverse barley and wheat collections to predict the size, frequency, and identity of introgressed chromosome regions, using the bioinformatics approach GBS coverage analysis. From the sequencing coverage profiles, which depict patterns of increased and decreased sequence coverage, the authors identified chromosomal alterations on a genome-wide scale, including introgressions, copy number variations, and modifications in DNA methylation. The same sequence coverage method was also applicable to identify alien chromatin from wild Triticum and Aegilops species in wheat chromosomes 2A, 2B, 2D, 3D, and 4A and to predict putative donor species (Keilwagen et al., 2022).

In the present study, we used heatmap visualization of the marker density to detect chromatin from the S. cereanum chromosomes 1R, 3RS, 4R, 5RL, 6RL, and 7RL introgressed into bread wheat. The resolution of rye-specific DArT markers whose chromosomal position was determined under a less stringent condition (70% identity percent) was sufficient to identify rye chromatin transferred into the Mv9kr1 wheat background. However, some discrepancies were found between the cytogenetic and DArTseq identification of S. cereanum chromosomes in some BC₃F₈ lines. We believe that the main reason for these discrepancies was not the low cutoff used to determine the genomic position of the markers and, thus, the poor determination of the marker positions. This opinion is supported by the results of 16 Chinese Spring (CS)–Imperial (I) disomic and ditelosomic addition lines (CS-I 1–7) ( Figures 1 , 2 ), genotyped together with the BC₃F₈ population and representing all of the rye chromosomes (1R–7R) and several chromosome arms (1RS, 1RL, 2RL, 3RS, 4RS, 4RL, 5RL, 6RL, 7RS) ( Table 3 ). These reference wheat–rye cytogenetic stocks, similar to many genetic studies aimed to identify molecular markers on rye chromosomes (Qiu et al., 2016; Ma et al., 2020; Li et al., 2022), were used to validate if rye-specific marker positions were well determined on the Lo7 reference sequence. Figures 1 , 2 demonstrate that the corresponding rye chromosomes or chromosome arms could be detected perfectly in 15 of the 16 reference CS-Imperial cytogenetic stocks. A 3RS-3RL rearranged structure was only identified for the 3RS ditelosomic line, but this is because of the rearranged structure of the chromosome arm in this specific genotype. (If the wrong map position of the markers was the reason, the distal half of 3RS would be missing from all 3RS-carrying genotypes.) We have to note that similar results were obtained with DArTseq markers mapped under a more stringent condition (93% IDT), but the marker coverage of the chromosomes was so low that visualizing the entire length of rye chromosomes and chromosome arms became problematic. If the genomic position of the rye-specific markers was incorrectly identified, the identified rye chromatin in the CS-Imperial lines would differ from the “official” rye chromosome content.

In our opinion, the most likely cause of discrepancies between cytogenetic and DArTseq identification of the rye chromosome in the BC₃F₈ lines is cytogenetic misidentification. In the present study, we used the pSc119.2 repeat as a FISH probe to identify rye chromosomes based on their hybridization pattern, which is mainly composed of telomeric and subtelomeric bands. However, our previous study revealed a high level of pSc119.2 hybridization pattern polymorphism on the chromosomes of S. cereanum ‘Kriszta’ (Szőke-Pázsi et al., 2022). The inability of GISH to detect intragenomic rearrangements in rye, combined with the existence of similar FISH patterns on different chromosome arms and FISH polymorphism on homologs, leads to an incorrect cytogenetic identification of rye chromosomes. These results highlight one of the advantages of using high-resolution marker systems together with the rye reference assembly; when combined with GISH, they provide a more detailed insight into the structure of transferred rye chromatin and are suitable for detecting intragenomic structural changes of rye chromosomes.

We also speculated on when the intragenomic rye chromosome rearrangements took place. We think that these rearrangements (like 1RS.3RS, 1RS.4RS) formed most likely after the first S. cereale × S. strictum crosses and/or during the breeding of the ‘Kriszta’ variety. This idea seems to be supported by the polymorphic karyotype of ‘Kriszta’ (Szőke-Pázsi et al., 2022). However, the planned development of a chromosome-scale reference genome sequence for ‘Kriszta’ will provide a more detailed understanding of the genome structure of this exotic S. cereale × S. strictum hybrid, as well as its genome relationship with domesticated S. cereale. Because ‘Kriszta’ is a source of several agronomically important traits (resistance to yellow rust and leaf rust, yield-increasing effect, grain quality) (Schneider et al., 2016; Szakács et al., 2020), determining the gene content of the introgressed QTLs would be highly beneficial. If the size of wheat–rye translocation chromosomes has changed significantly relative to the normal wheat chromosomes, and the distribution of FISH microsatellite probes on them is appropriate, the application of MutChromSeq pipeline using bivariate chromosome flow-sorting in combination with EMS mutagenesis (Sánchez-Martín et al., 2016) could be an effective option to identify agronomically important genes on the transferred alien chromosomes as demonstrated for cloning disease resistance genes (Holden et al., 2022; Yu et al., 2022).

We have to note, however, that molecular cytogenetic analysis cannot be eliminated from the characterization of wheat-alien hybrid progenies during the alien gene transfer. Our results indicate that the parallel analysis using cytogenetic methods and high-resolution marker systems could be the most optimal for the precise detection of intergenomic rearrangements. The 7RL-6RL chromosome arm is present in two kinds of translocation: 3RS.7RL-6RL and 7RS.7RL-6RL. As these rearranged rye chromosomes can be found as additions, the translocation between the arms 6RL and 7RL arose directly after the cross between S. cereale and S. strictum ssp. anatolicum or during the development of the cultivar Kriszta, rather than after the cross between Mv9kr1 and ‘Kriszta.’ This 6R/7R translocation is not unique. The studies by Benito et al. (1991) on the location of the endopeptidase-1 (Ep-1) gene on S. cereale and S. montanum (syn. S. strictum) chromosomes provided biochemical evidence that one of the three translocations evolutionary separating S. cereale from S. montanum (Singh and Röbbelen, 1977) had occurred between 6RL and 7RL. This suggests that the translocation in the genotypes 3RS.7RL-6RL and 7RS.7RL-6RL between the 6RL and 7RL arms may be of compensating type. The seed set data of these addition lines suggest that the long arm 7RL-6RL does not affect this trait but the short arm of the translocation chromosome does. It seems that 7RS has a strong yield-decreasing effect as the grain number per ear values for the 7RS.7RL-6RL disomic addition, as well as the 7RS ditelosomic addition, are extremely low (2 and 5 on average, respectively), while those for the 3RS.7RL-6RL addition line are the highest (37 grains on average) among the studied genotypes.

The line disomic for the 3RS isochromosome (carrying four copies of the 3RS chromosome arm) also shows good fertility, with an average of 34 grains per ear. Based on the grain number per ear value (24 on average), which is higher than that of the non-compensating translocation line T1BL.5RL (19 on average) but lower than that of the addition line 3RS.7RL-6RL, it is difficult to predict whether the translocation composed of a 3RS arm and a shorter wheat chromosome arm (segment) is compensating or not. The weak, terminal pSc119.2 signal suggests that the wheat arm is either 3DS or 2BS. Nevertheless, rye chromosomes pair more frequently with B-genome chromosomes than with those from the A and D genomes (Naranjo and Fernández-Rueda, 1996).

The existence of intergenomic relationships between the group-4 chromosomes of wheat and 7R, as well as group-7 chromosomes of wheat and 4R, has long been known (Zeller and Koller, 1981; Devos et al., 1993). So far, only two fertile translocation lines involving these chromosomes have been published. One of them is composed of 4A (or segment) and 7RS arms (Zeller and Koller, 1981), and the other one carries 7DL and S. montanum 4RL chromosome arms (Lukaszewski et al., 2001).

The present study focused on the detection of wheat–S. cereanum introgressions and the precise characterization of their genome structure. Because of the limited seed set, a detailed evaluation of agronomic performance, abiotic and biotic stress tolerance, and grain quality traits is not possible. However, several valuable traits have previously been mapped on chromosome 4R and chromosome arms 3RS, 5RL, 6RL, and 7RS, implying that the new wheat–S. cereanum introgression lines could be important gene sources for wheat improvement.

3RS harbors the stem rust resistance gene Sr27 (Rao, 1979). Although translocation lines carrying 3RS have been developed (Marais and Marais, 1994), they have not been used in wheat breeding because of their “detrimental” effect on grain yield (Marais, 2001). In contrast, the Mv9kr1-’Kriszta’ addition line disomic for the 3RS isochromosome yields 34 ± 9.6 grains per ear, comparable to cultivated wheat. It might therefore be a realistic plan for the future to produce translocations with high yields. One of the most efficient aluminum tolerance genes found in the cultivated species of Triticeae, Alt2 (Aniol and Gustafson, 1984), is also located on 3RS (Aniol, 2004).

The disomic addition of chromosome 4R increases the number of grains per spike (An et al., 2019), total protein and arabinoxylan content, and stripe rust resistance in wheat (Szakács et al., 2016). The 4RL chromosome arm carries a dominant locus (Alt3) for aluminum tolerance (Miftahudin et al., 2002) and resistance genes for powdery mildew (Duan et al., 2017; Ma et al., 2020).

A good copper efficiency in rye has been associated with 5RL (Schlegel et al., 1991). Dwarfing genes ct2 and Ddw1 (Plaschke et al., 1993; Korzun et al., 1996) and a stripe rust resistance gene (Xi et al., 2019) have also been mapped to this rye chromosome arm.

7RS carries the aluminum tolerance gene Alt4 (Matos et al., 2005) and genes for zinc efficiency (Schlegel et al., 1999). Using CS-’Imperial’ disomic addition lines, it was shown that chromosome 7R increases the drought tolerance in wheat (Mohammadi et al., 2003), but the location on chromosome arms of the responsible genes is not known.

Several studies have found that chromosome arm 6RL is a potential gene source for wheat resistance breeding. On the 6RL arm of the rye cv. Prolific, a powdery mildew resistance gene (Pm20) was physically mapped by Friebe et al. (1994) to a distal position. Additional loci for resistance to powdery mildew on 6RL have been found in the rye cultivars Jingzhouheimai (Wang et al., 2010) and Kustro (Du et al., 2018). On this chromosome arm of rye cv. Merced, a novel stripe rust resistance gene (Yr83) has been reported recently (Li et al., 2020) and physically located at the fraction length 0.87–1.00, which corresponds to the 806–881-Mb region on the ‘Lo7’ reference genome (Li et al., 2022). Given that the small-sized 6RL fragment of the Mv9kr1-’Kriszta’ 3RS.7RL-6RL translocation chromosome is located terminally and the disomic addition line carrying this chromosome has good fertility, we consider this genetic material as a candidate for wheat resistance breeding. This is supported by field observations showing adult plant resistance of this genotype to naturally occurring stripe rust infections in years when the pathogen was present in the breeding nursery. It should be noted that detailed resistance tests in seedling and adult plant stages with monosporic isolates are needed to confirm the presence of an effective Yr resistance gene on this 6RL chromosome fragment. In addition, a translocation line containing only the short 6RL segment must be developed in order to capture the gene(s) of interest without the undesirable linkage drag caused by the 3RS.7RL chromosome part. Türkösi et al. (2022) introduced the ph1b mutation from Chinese Spring ph1b mutants into the winter wheat line Mv9kr1, which is the wheat genetic background of the BC₂F₈ generation. The Mv9kr1ph1b mutant line presents a promising tool for inducing the desired translocation. We also plan to convert the 1R–7R chromosome-specific SNPs to Kompetitive Allele-Specific PCR (KASP) markers for faster and less expensive screening of wheat lines for the presence of introgressed chromatin as it was demonstrated for Ae. mutica (King et al., 2017). This will enable us to identify additional introgression lines that were previously unidentified, such as those that carry 2R chromatin.

Great efforts are being made to increase the genetic diversity of wheat in order to improve its yield, stress tolerance, and disease resistance. Wheat–rye introgression lines offer promising genetic materials for achieving this goal, as cultivated and wild rye chromosomes harbor many useful genes hitherto unexploited in wheat breeding as it was systematically reviewed by Crespo-Herrera et al. (2017). The present study demonstrates that high-throughput DArTseq genotyping accurately identifies the “exotic” rye chromatin of S. cereanum in the wheat genetic background. However, molecular cytogenetic analyses (FISH, GISH) are still required to determine the precise chromosomal composition (monosomic, disomic, etc.) of potentially agronomically useful prebreeding populations. The genotypes from the wheat–S. cereanum BC₂F₈ generation, particularly the translocation (T1BL.5RL, Twheat.3RS) and addition (3RS isochomosomic, 4R, and 3RS.7RL-6RL) lines with good seed set, can be potentially valuable resources for breeding stress-tolerant or disease-resistant wheat cultivars. However, further experiments will be needed to prove if the rye chromosome arms introgressed from S. cereanum into wheat carry the effective alleles of the aforementioned genes. Future research will address this issue.