Characterization of a wheat–tetraploid Thinopyrum elongatum 1E(1D) substitution line K17–841-1 by cytological and phenotypic analysis and developed molecular markers

Background Tetraploid Thinopyrum elongatum (2n = 4x = 28) is a promising source of useful genes, including those related to adaptability and resistance to diverse biotic (Fusarium head blight, rust, powdery mildew, and yellow dwarf virus) and abiotic (cold, drought, and salt) stresses. However, gene transfer rates are low for this species and relatively few species-specific molecular markers are available. Results The wheat-tetraploid Th. elongatum line K17–841-1 derived from a cross between a hexaploid Trititrigia and Sichuan wheat cultivars was characterized based on sequential genomic and fluorescence in situ hybridizations and simple sequence repeat markers. We revealed that K17–841-1 is a 1E (1D) chromosomal substitution line that is highly resistant to stripe rust pathogen strains prevalent in China. By comparing the sequences generated during genotyping-by-sequencing (GBS), we obtained 597 specific fragments on the 1E chromosome of tetraploid Th. elongatum. A total of 235 primers were designed and 165 new Th. elongatum-specific markers were developed, with an efficiency of up to 70%. Marker validation analyses indicated that 25 specific markers can discriminate between the tetraploid Th. elongatum chromosomes and the chromosomes of other wheat-related species. An evaluation of the utility of these markers in a F2 breeding population suggested these markers are linked to the stripe rust resistance gene on chromosome 1E. Furthermore, 28 markers are unique to diploid Th. elongatum, tetraploid Th. elongatum, or decaploid Thinopyrum ponticum, which carry the E genome. Finally, 48 and 74 markers revealed polymorphisms between Thinopyrum E-genome- containing species and Thinopyrum bessarabicum (Eb) and Pseudoroegneria libanotica (St), respectively. Conclusions This new substitution line provide appropriate bridge–breeding–materials for alien gene introgression to improve wheat stripe rust resistance. The markers developed using GBS technology in this study may be useful for the high-throughput and accurate detection of tetraploid Th. elongatum DNA in diverse materials. They may also be relevant for investigating the genetic differences and phylogenetic relationships among E, Eb, St, and other closely-related genomes and for further characterizing these complex species.

Marker-assisted selection is useful for detecting genes associated with a trait of interest based on a linked marker, with implications for breeding involving multiple traits [11]. Developing species-specific molecular markers that facilitate the identification of alien chromosomes or segments is very important for wheat breeding programs [19,20]. Diverse molecular markers specific to diploid Th. elongatum and Th. ponticum have been reported, including random-amplified polymorphic DNAs (RAPDs), simple sequence repeats (SSRs), expressed sequence tags, cleaved amplified polymorphic sequences, sequence-tagged sites, sequence-characterized amplified regions, amplified fragment length polymorphisms, single nucleotide polymorphisms, and specific-locus amplified fragments (SLAFs) [11,12,19,[21][22][23][24][25]. However, there are still relatively few of these markers, and they do not cover the whole Thinopyrum genome. Consequently, there is an urgent need to develop more molecular markers, especially those with a wide genomic distribution and that are amenable to high-throughput genotyping [12]. Genotyping-by-sequencing (GBS) is a high-throughput, highly accurate, inexpensive, and relatively simple method for developing several markers in Triticale species [26]. However, developing specific molecular markers for tetraploid Th. elongatum remains difficult. Therefore, generating chromosome-specific molecular markers is critical for detecting and tracing alien chromosomes in wheat-tetraploid Th. elongatum hybrid derivatives.
Tetraploid Th. elongatum is an important genetic resource for improving wheat, and some wheat-tetraploid Th. elongatum derivative lines have been developed by crossing hexaploid Trititrigia with Sichuan wheat cultivars [17]. Li et al. [18] produced a fluorescence in situ hybridization (FISH) karyotype of the E-genome chromosomes of tetraploid Th. elongatum based on various repetitive sequence probes, which may enhance the utility of tetraploid Th. elongatum for the introgression of alien genes into wheat. The main objectives of the current study were to: (1) characterize the chromosomal constitution of a wheat-tetraploid Th. elongatum substitution line and evaluate its effects on stripe rust resistance and agronomic traits and (2) develop and validate specific and easy-to-use molecular markers based on GBS, which may be useful for the efficient and reliable identification of tetraploid Th. elongatum chromatin in several species, including common wheat.

Results
Chromosomal constitution of K17-841-1 Genomic in situ hybridization (GISH), FISH, and SSR marker analyses were performed to determine the chromosomal constitution of wheat-tetraploid Th. elongatum line K17-841-1. When tetraploid Th. elongatum total genomic DNA and the J-11 genomic DNA were used as the probe and the blocking DNA, respectively, we observed that line K17-841-1 carried 40 wheat chromosomes and two E chromosomes (Fig. 1a). The GISH signals were sequentially removed and the slides were used in a FISH analysis involving pSc119.2 and pTa535 probes. A pair of E chromosomes produced strong terminal pSc119.2 signals on both arms as well as strong pTa535 signals on the subterminal regions of the short arm and near the centromeric region of the long arm (Fig. 1b). These results are consistent with the previously reported FISH pattern for the 1E chromosome of tetraploid Th. elongatum [18]. Thus, the FISH karyotype of the E-genome chromosomes of tetraploid Th. elongatum and common wheat based on probes pSc119.2 and pTa535 suggested that K17-841-1 is a 1E (1D) chromosomal substitution line. To determine the cytological stability of K17-841-1, we used GISH and FISH to analyze 20 randomly selected seeds of K17-841-1 self-progeny. We revealed that all seeds contained 14 A-(1A-7A), 14 B-(1B-7B), and 12 D-(2D-7D) genomes, as well as a pair of 1E chromosomes (Fig. 1c, d).

Morphology of K17-841-1
We analyzed the agronomic traits of K17-841-1 and the donor parents in two growing seasons. Line K17-841-1 displayed stable morphological traits, which were similar to those of the wheat parents SM482 and SM921 (Table 1; Fig. 3). The average plant height and spike length of line K17-841-1 were significantly lower than those of the Triticum durum-tetraploid Th. elongatum partial amphidiploid 8801, but were greater than those . Arrows indicate 1E-genome chromosomes. Scale bar: 10 μm of SM482 and similar to those of SM921 (Fig. 3a, b). The number of spikelets per spike was lower than that of SM482 or SM921, but was similar to that of 8801. The grain number per spike was greater than that of 8801, but was lower than that of SM921 and similar to that of SM482 (Fig. 3c). There were no significant differences between K17-841-1 and either SM482 or SM921 regarding the tiller number, thousand-kernel weight, and seed setting rate (Fig. 3d).
Development and validation of specific molecular markers for the 1E chromosome of tetraploid Th. elongatum The GBS approach was applied to identify tetraploid Th. elongatum 1E chromosome-specific sequences. Details regarding the sequencing, including raw reads, clean reads, effective rate, error rate, Q20, Q30, and GC content, are summarized in Table 2. The results revealed a high sequencing quality (Q20 ≥ 94% and Q30 ≥ 86%) and a normal GC content distribution. A total of 45,115,502, 7,389,702, 8,710,288, and 8,831,260 effective GBS sequences were obtained for PI531718, PI531750, 8801, and K17-841-1, respectively ( Table 3). The sequencing depth was more than 10.32×. A sequence comparison revealed 73,744 K17-841-1 clean reads that were less than 23% homologous with CS sequences. Additional sequence alignments uncovered 2952 K17-841-1 reads that were more than 23% homologous with 8801 and  Data in the columns indicate means ± standard errors Lowercase letters following the means indicate significant differences at P < 0.05 as determined by the least significant differences PI531750 fragments obtained by GBS. Finally, 1194 K17-841-1 unique reads (597 specific fragments) that were less than 10% homologous with sequences from PI531718 and the other six substitution lines were obtained and were considered the tetraploid Th. elongatum 1E chromosome-specific fragments.

Utility of the tetraploid Th. elongatum-specific markers in the F 2 population
To verify the utility of the newly developed molecular markers, 80 F 2 individuals of K17-841-1 and wheat cultivar Shumai969 (SM969) were analyzed by PCR. We detected specific amplicons for only 20 individuals (Fig. 9, Additional file 3: Table S3). The results of a GISH indicated these 20 plants with specific amplicons carried 1E (1D) chromosomal substitutions, while the 60 plants without specific amplicons had no GISH signal in their chromosome preparation. More importantly, an evaluation of stripe rust resistance at the seedling stage revealed that 8801, K17-841-1, and the 20 positive F 2 individuals carrying 1E chromosomal markers were highly resistant to Pst race CYR-34. In contrast, the 60 F 2 plants lacking amplicons as well as SY95-71 and the parental lines SM482, SM921, and SM969 were highly susceptible ( Fig. 10, Additional file 3: Table S3). These observations implied that the specific markers developed in this study may be useful for detecting the stripe rust resistance gene on the 1E chromosome of tetraploid Th. elongatum during the breeding of new disease-resistant wheat varieties.

Discussion
Distant hybridizations may be useful for transferring agronomically valuable genes from wild relatives to common wheat varieties. Creating intermediate lines carrying alien chromosomes of wheat relatives provides the basis for using these germplasm resources to improve domesticated wheat [27]. Wheat-alien chromosomal substitution lines are important bridge materials for transferring alien genes to common wheat. In recent decades, there have been many attempts by wheat breeders to generate wheat-Th. elongatum substitution lines [8]. However, most of the introgressions from Th. elongatum into wheat involved the diploid Th. elongatum and decaploid Th. ponticum. For example, CS-  diploid Th. elongatum addition and substitution lines have been produced [28]. Additionally, Zheng et al. [29] generated the wheat-Th. ponticum 4Ag (4D) disomic substitution line Blue 58 from a hybridization between Th. ponticum and common wheat. Fu et al. [30] developed a wheat-diploid Th. elongatum 7E (7D) substitution line resistant to Fusarium head blight. Wang et al. [31] produced a wheat-Th. ponticum St (6A) disomic substitution line exhibiting powdery mildew resistance. The 7J St (7B) substitution line CH1113-B13, which is resistant to stripe rust, was identified from among the progeny of a cross between common wheat and Th. ponticum [32]. The tetraploid Th. elongatum harbors many beneficial genes that provide protection from diseases as well as cold, drought, and salinity stresses. Thus, it is a valuable genetic resource for improving the tolerance of common wheat to biotic and abiotic stresses [9,10,17].
To date, there are few reports regarding wheat-tetraploid Th. elongatum substitution lines. Li et al. [17] developed 50 wheat-tetraploid Th. elongatum introgression lines by crossing the T. durum-tetraploid Th. elongatum partial amphidiploid line 8801 with wheat In the current study, we further characterized the wheat-tetraploid Th. elongatum 1E (1D) disomic substitution line K17-841-1 derived from the 8801/SM482/SM921 F 5 progeny. The FISH and SSR analyses confirmed that common wheat chromosome 1D was substituted with the homologous chromosome 1E from tetraploid Th. elongatum, providing further evidence that the E genome of Th. elongatum is closely related to the D genome of common wheat [33,34].
Interestingly, an evaluation of the stripe rust resistance indicated that the wheat parents SM482 and SM921 are susceptible and that the resistance of K17-841-1 was derived from tetraploid Th. elongatum. Some stripe rust resistance genes have been identified in Thinopyrum species and have been transferred to common wheat, including Yr50 from Th. intermedium, YrE from diploid Th. elongatum, and Yr69, YrTP1, YrTP2, and YrCH7056 from Th. ponticum [35,36]. To the best of our knowledge, this study is the first to demonstrate the successful transfer of a new and high-level stripe rust resistance  gene from tetraploid Th. elongatum. Furthermore, the grain yield of line K17-841-1 is similar to that of the wheat parents SM482 and SM921. Accordingly, K17-841-1 represents an appropriate bridge breeding material for the introgression of alien genes to enhance wheat disease resistance. Molecular markers are widely used to detect and track alien chromosomes and/or chromosomal segments during crossing and selection [10]. Conventional methods for developing markers, including SSRs, RAPDs, and amplified fragment length polymorphisms, have low success rates because of the high genomic sequence homology between Th. elongatum and common wheat. For example, using 40 SSR primers, You et al. [37] amplified 108 diploid Th. elongatum-specific fragments, but only one genome-specific molecular marker was developed. In another previous study, 94 diploid Th. elongatum-specific fragments were generated with 26 pairs of RAPD primers, but only three 1E or 3E chromosome-specific markers were developed [38]. Additionally, Ge et al. [39]  developed one diploid Th. elongatum-specific marker from 65 specific fragments based on suppression subtractive hybridization. Furthermore, inter-SSR, interretrotransposon amplified polymorphism, and retrotransposon microsatellite amplified polymorphism techniques were applied to develop 13 specific sequence-characterized amplified region markers based on 34 specific diploid Th. elongatum sequences [40]. In a recent study, 135 randomly selected fragments of diploid Th. elongatum generated by SLAF-sequencing (SLAF-seq) were used to develop 89 specific molecular markers, including 61 7E chromosome-specific markers [11]. Chen et al. [25] developed 20 diploid Th. elongatum 1E chromosome-specific and two E genome-specific molecular markers based on SLAF-seq data, with an efficiency of up to 60%. Approximately 89% of the mined SNPs of diploid Th. elongatum may be authentic with respect to their polymorphisms and chromosomal locations as determined by a high-resolution melting curve assay [12]. Moreover, SLAF-seq data were used to develop 67 Th. ponticum-specific molecular markers from a wheat-Th. ponticum translocation line, with a success rate of up to 39% [19]. Furthermore, GBS, which is a robust, simple, and high-throughput next-generation sequencing approach that can decrease the complexity of large genomes, has been applied to develop many molecular markers in Triticale species [26,41]. This approach can generate considerable sequence information and is appropriate for all whole-genome density distributions. The flexibility and low cost of GBS makes it ideal for genomics-assisted breeding [42]. To date, there have been no reports describing the development of tetraploid Th. elongatum-specific molecular markers. In this study, the GBS method was applied to develop 165 specific molecular markers, including 132 tetraploid Th. elongatum chromosome 1E-specific PCR markers, with a success rate of up to 70%. The greater success rate for our study relative to those of earlier investigations indicates that GBS technology may be used to efficiently develop alien-specific molecular markers in a wheat background.. The markers developed in this study may be useful for the high-throughput and accurate detection of tetraploid Th. elongatum DNA in diverse materials. Most importantly, the verification of their utility for analyzing a F 2 breeding population implies that these specific markers are appropriate for tracing the stripe rust resistance gene carried on chromosome 1E of tetraploid Th. elongatum for the breeding of enhanced materials.
Even after removing K17-841-1 sequences that were more than 90% homologous with sequences from six other substitution lines (2E-7E), 12 markers still amplified specific sequences in the other substitution lines, confirming that the Th. elongatum E chromosomes sequences are highly homologous, likely because of chromosomal rearrangements. Earlier investigations involving biochemical and molecular markers as well as SLAF-seq revealed similar sequence identity levels for the diploid Th. elongatum [11,43]. After analyzing the hybrids from crosses between tetraploid Th. elongatum and diploid Th. elongatum and Th. bessarabicum, Dvořák [44] concluded that tetraploid Th. elongatum is a segmental allotetraploid with two closely related E genomes. On the basis of examinations of chromosomal pairs, karyotypes, and C-banding patterns,Liu and Wang [45] also determined that tetraploid Th. elongatum is a segmental allotetraploid carrying two E genomes with slight differences. In a previous study involving a FISH with repetitive clones, two tetraploid Th. elongatum accessions were analyzed to uncover the significant chromosomal polymorphisms between the E genome of the putative diploid progenitor Th. elongatum and the other genome [18]. The verification of 153 markers specific to tetraploid Th. elongatum confirmed that 21, 106, 50, and 73 markers can detect E genome chromosomes in diploid Th. elongatum, Th.  Table S2). Our results provide further evidence that tetraploid Th. elongatum is an allotetraploid harboring two different E genomes, and Thinopyrum E genomes were differentiated during polyploidization events. Two specific markers, TTE1E-212 and TTE1E-222, were unique to diploid Th. elongatum, tetraploid Th. elongatum, and Th. ponticum, suggesting they can detect the E chromosomal DNA in all materials carrying the E genome. Interestingly, 58, 32,5,8,28,8, and 13 markers based on specific tetraploid Th. elongatum sequences successfully amplified products from Th. bessarabicum (E b ), Pse. libanotica (St), D. villosum (V), H. bogdanii (H), Ag. cristatum (P), S. cereale (R), and P. huashanica (Ns), respectively. Our specificity analysis indicated that the marker amplification frequencies are considerably higher in E, E b , and St genomic species than in other wheat-related species, which is consistent with the findings of previous studies, in which the E genomes of Th. elongatum and Th. ponticum were more closely related to the E b genome of Th. bessarabicum and the St genome of Pseudoroegneria species than to the other genomes [33,46]. Additionally, 48 and 74 markers uncovered polymorphisms between Thinopyrum E-genomecontaining species and Th. bessarabicum and Pse. libanotica, respectively, indicating that they will likely be applicable for investigating the genetic differences between E, E b , and St genomes and for revealing the genetic diversity among species (populations) carrying E, E b , and/or St genomes. The resulting information will further characterize these complex species and may be relevant for the genetic improvement of these important forage crops.

Conclusions
We characterized a wheat-tetraploid Th. elongatum 1E (1D) disomic substitution line based on GISH, FISH, and SSR analyses. This line is highly resistant to the stripe rust pathogen strains prevalent in China. Moreover, it represents an appropriate bridge breeding material for the introgression of alien genes to improve wheat disease resistance. Additionally, we analyzed GBS data to generate and validate a new 1E chromosome-specific, easy-to-use marker set that may be applicable for identifying and characterizing the tetraploid Th. elongatum chromosomes along with the chromosomes of all other wheat-related species and for investigating the genetic differences and phylogenetic relationships among E, E b , St, and other closelyrelated genomes. The use of this marker set will further increase our understanding of these complex species.

Plant materials
The plant materials used in the current study are listed in Table 5. The hexaploid Trititrigia line 8801 (2n = 6x = 42, AABBEE), which is tolerant to cold, drought, and salt stresses and resistant to Fusarium head blight, rust, and powdery mildew, was originally produced and identified at the Eastern Cereal and Oilseed Research Center, Ottawa, Canada [17]. The native wheat cultivars SM482, SM921, and SM969 exhibit superior agronomic traits, but are susceptible to the stripe rust pathogens prevalent in southwestern China. To produce wheattetraploid Th. elongatum derivative line, we first crossed 8801 with SM482, after which the resulting F 1 plants were further crossed with SM921 to obtain the BC 1 F 1 population. Seeds from the BC 1 F 1 plants were bulked and advanced to the BC 1 F 5 generation by single seed descent, ultimately resulting in the isolation of line K17-841-1 (Fig. 11). The F 2 population comprising 80 individuals was derived from a cross between K17-841-1 and SM969. We developed six wheat-tetraploid Th.

GISH and FISH analyses
Root tips from germinating seeds were treated with nitrous oxide for 2 h and 90% acetic acid for 10 min, after which they were digested with pectinase and cellulase (Yakult Pharmaceutical, Tokyo, Japan) [47]. Slides were then prepared for the GISH analysis as described by Han et al. [48]. The cetyltrimethyl ammonium bromide method was used to extract genomic DNA from the freshly collected leaves of tetraploid Th. elongatum PI531750 and wheat cultivar J-11 [49]. The PI531750 DNA was labeled with fluorescein-12-dUTP according to the nick translation method (Thermo Fisher Scientific, Eugene, OR, USA) and was used as the hybridization probe. The GISH analysis was completed according to the method described by Han et al. [50], with a 1:150 probe DNA:blocking DNA ratio. Specifically, 10 μL hybridization solution containing 2 × saline sodium citrate (SSC), 10% dextran sulfate, and 10 ng/μL labeled probe DNA together with blocking DNA was added to each slide. Samples were denatured by heating at 85°C for 5 min, incubated at 37°C for 8 h, and washed with 2× SSC. Finally, the chromosomes were counterstained with 4,6-diamino-2-phenylindole solution (Vector Laboratories, Burlingame, CA, USA). The GISH slides were examined with the BX-63 microscope (Olympus, Tokyo, Japan) and images were captured with a DP-70 CCD camera.
The GISH slides were washed sequentially with 70% (v/v) ethanol for 5 min, 2× SSC at 60°C for 30 min, ddH 2 O for 10 min, and 100% (v/v) ethanol for 5 min. A FISH analysis was subsequently completed to identify the chromosomal constitution of line K17-841-1, with pSc119.2 and pTa535 as probes [47,51]. The FISH was performed as described by Han et al. [48], with minor

SSR marker analysis
Primer pairs for four SSR markers (wmc147, wmc222, gwm337, and Xcfd63) on wheat chromosomes 1DS and 1DL were used to characterize the 1D chromatin in the wheat-tetraploid Th. elongatum line K17-841-1. Details regarding all primers are listed in Table 6. Additionally, CS, SM482, and SM921 were used as positive controls, whereas 8801 was used as a negative control. The PCR amplification was completed as described by Somers et al. [52], with minor modifications.

Agronomic trait evaluation
The morphological traits of K17-841-1 and its parents were evaluated in a field trial in Wenjiang, Sichuan The stripe rust responses were evaluated with three replicates as described by Li et al. [17]. The stripe rust IT was based on the following scale: 0, 0;, 1, 2, 3, and 4, in which 0 = immunity, 0; = necrotic flecks, and 1-4 = increasing sporulation and decreasing necrosis or chlorosis. Plants with an IT of 2 or lower were considered resistant, whereas plants with an IT of 3 or 4 were considered susceptible.
Genotyping-by-sequencing, sequence alignment, and tetraploid Th. elongatum 1E chromosome-specific fragment acquisition Total genomic DNA was extracted from fresh young leaves with the cetyltrimethyl ammonium bromide method, after which the DNA concentration was adjusted to 150 ng/μL. Genomic DNA samples of diploid Th. elongatum PI531718, PI531750, 8801, K17-841-1, and six wheat-tetraploid Th. elongatum (2E-7E) chromosomal substitution lines were subjected to GBS, which was completed by Novogene Bioinformatics Technology Co., Ltd. (Beijing, China). High-Quality DNA libraries were constructed and sequenced with the Illumina HiSeq™ system. The raw reads in FASTQ format were filtered by removing low-quality reads and reads with adapter and/or poly-N sequences to obtain clean reads (140 bp). Tetraploid Th. elongatum 1E chromosome-specific sequences were obtained as follows. First, the high-quality K17-841-1 sequences were compared with the CS reference genome sequences (https://urgi.versailles.inra.fr/download/iwgsc/IWGSC_ RefSeq_Assemblies/v1.0/), after which the sequences with identities greater than 23% were eliminated [54]. Second, the sequences were compared with those of 8801 and PI531750 acquired by GBS in this study, and the sequences with identities greater than 23% were selected. Finally, the retained K17-841-1 sequences were compared with the sequences of PI531718 and the six chromosomal substitution lines. Sequences with identities exceeding 90% were removed. The remaining sequences were considered to be specific to chromosome 1E of tetraploid Th. elongatum.
Development and validation of the tetraploid Th. elongatum-specific molecular markers On the basis of these specific sequences, PCR primers were designed with the Primer3 Plus online tool (http:// www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus. cgi) and then synthesized by TSINGKE (Chengdu, China). Details regarding the PCR primers are presented in Additional file 1: Table S1. The amplified products were examined by 3% agarose electrophoresis. The markers that amplified specific sequences in PI531750, 8801, and K17-841-1, but not in SM482, SM921, CS, PI531718, and the 2E-7E substitution lines, were identified as tetraploid Th. elongatum 1E chromosome-specific molecular markers. The stability, repeatability, and specificity of these markers were validated in the CS, PI531750, 8801, and K17-841-1 lines as well as in 11 wheat-related species. The PCR amplifications were completed in a reaction volume of 25 μL, which included 1.0 μL template DNA (100 ng/μL), 12.5 μL 2× Taq Master Mix for PAGE (Dye plus), 1.0 μL each primer (10 μM), and 9.5 μL ddH 2 O. The PCR program was as follows: 94°C for 5 min; 35 cycles of 94°C for 30 s, appropriate annealing temperature of 50-60°C for 30 s, and 72°C for 1 min; 72°C for 10 min.