Production of synthetic wheat lines to exploit the genetic diversity of emmer wheat and D genome containing Aegilops species in wheat breeding

Due to the accumulation of various useful traits over evolutionary time, emmer wheat (Triticum turgidum subsp. dicoccum and dicoccoides, 2n = 4x = 28; AABB), durum wheat (T. turgidum subsp. durum, 2n = 4x = 28; AABB), T. timopheevii (2n = 4x = 28; AAGG) and D genome containing Aegilops species offer excellent sources of novel variation for the improvement of bread wheat (T. aestivum L., AABBDD). Here, we made 192 different cross combinations between diverse genotypes of wheat and Aegilops species including emmer wheat × Ae. tauschii (2n = DD or DDDD), durum wheat × Ae. tauschii, T. timopheevii × Ae. tauschii, Ae. crassa × durum wheat, Ae. cylindrica × durum wheat and Ae. ventricosa × durum wheat in the field over three successive years. We successfully recovered 56 different synthetic hexaploid and octaploid F2 lines with AABBDD, AABBDDDD, AAGGDD, D1D1XcrXcrAABB, DcDcCcCcAABB and DvDvNvNvAABB genomes via in vitro rescue of F1 embryos and spontaneous production of F2 seeds on the Fl plants. Cytogenetic analysis of F2 lines showed that the produced synthetic wheat lines were generally promising stable amphiploids. Contribution of D genome bearing Aegilops and the less-investigated emmer wheat genotypes as parents in the crosses resulted in synthetic amphiploids which are a valuable resource for bread wheat breeding.

Crossing. Crosses between tetraploid wheat genotypes as the female parents and Ae. tauschii genotypes as the male parents and crossing between tetraploid Aegilops species (Ae. crassa, Ae. cylindrica and Ae. ventricosa) as female parents and tetraploid wheat genotypes as male parents were made by hand between the months of May and June in 2017, 2018 and 2019 at the research farm of the University of Kurdistan. Approximate temperature and humidity during the crossing period ranged from 18 to 37 °C during the day and 5-17 °C at night, with low humidity and precipitation. Only the two outermost florets of spikelets were pollinated. No hormone treatment was applied. The spikes of wild emmer (T. dicoccoides) and tetraploid Aegilops species were bagged after crossing in order to prevent spikelet dispersal and to enable seed collection.
In late summer, the embryos from the dried mature shriveled F 1 seeds belonging to each cross combination between Ae. tauschii and tetraploid wheat genotypes were rescued. For this, the shriveled seeds were firstly sterilized in 5% sodium hypochlorite solution for 15 min with shaking, rinsed in sterilized distilled water for 2 × 10 min and kept in sterilized distilled water overnight at 4 °C. The seed coat was carefully removed, the embryo was placed on ½ MS media (pH 5.8, including vitamins) 24 in a sterile jar and maintained under photoperiod of 16 h of light and 8 h of darkness at 22-24 °C. Grown seedlings of about 10 cm long were washed to remove the media and transferred to soil in small pods. The established seedlings were finally transplanted to the field. Chemical treatment for chromosome doubling was not applied, and the production of F 2 seeds from F 1 plants in the upcoming spring was relied on the ability of the hybrids to form unreduced male and female gametes.
Hybrid seeds between tetraploid Aegilops species and wheat genotypes had endosperm and did not require embryo rescue. In the spring of the following year, non-hybrid plants were weeded out at the flowering stage and only the true hybrid plants-which were morphologically distinguishable-were retained in the field. The F 1 spikes were bagged to enforce self-pollination. As a rule to indicate the cross direction or genome designation of each hybrid or amphiploid in the present study, the female parent or maternal genome is listed first followed by the male parent or paternal genome.
Pollen viability analysis. Pollen viability was calculated as the percentage of pollen stained with Alexander's solution 25 . Immature anthers were randomly selected from four spikelets of different tillers in each hybrid combination and 10 anthers were analyzed to measure the percentage of viable pollen grains in each F 1 cross combination. Strongly stained swollen pollen grains were assumed to be viable.
The number of seeds used for cytogenetic works was case-dependent: chromosome analysis of the F 1 seeds/ embryos was done using the root tips of single seeds/seedlings separately and the corresponding seed/seedling was replanted to recovery and grow. For the analysis of the parental lines, three different seeds were commonly used from each line. FISH analysis and chromosome counting of Ae. cylindrica-T. durum, Ae. ventricosa-T. durum and Ae. crassa-T. durum amphiploids were applied using the single seed roots. C-banding. The C-banding technique described by Gill et al. 32 was used with mirror modifications in slide preparation. Slides were prepared as for FISH and stored in 96% ethanol at − 20 °C for at least 24 h. Slides were then dried at room temperature and used for C-banding.
Phenotypic evaluation. Amphiploid lines were grown in autumn each in a row at the research farm of the University of Kurdistan main campus. Rows were 5 m long and 0.5 m apart with a sowing rate of 20 seeds per row. The field was watered by both rainfall and irrigation but no fertilizer was applied. Nine representative plants from each line were used for phenotypic evaluation at maturity. The plant height and spike length (both excluding awns), awn length, total spikelets per spike, nodes number, flag leaf width and length, flowering time (from the first day of the spring) and peduncle length were measured from the main tillers.
For the analysis of the iron (Fe) and zinc (Zn) contents, one gram grain samples were digested in a mixture of concentrated HNO 3 (two parts) and HCl (one part) according to Zarcinas et al. 33 until a white residue was obtained. The required volume was made up after completion of the digestion process, and digests were analyzed using an atomic absorption spectrophotometer (GBC 902 AA, Australia). Three biological replications from each amphiploid line were analyzed. Fe and Zn concentrations were presented in microgram per gram dry weight of seed (µg/g DW).
Statistical analysis. The crossability of each parental genotype was calculated as the percentage of the F 1 embryos or seeds obtained over the total florets pollinated for that cross. Bar graphs of the crossability rates were prepared in the base package of R version 3.6.1 (The R Project for Statistical Computing, Vienna, Austria). Pollen viability data between the cross combinations were analyzed based on completely randomized design in R where the assessed anthers in each cross combination were considered as replications. Because the pollen viability rates in cross combinations were correlated with variance, logarithm of the data were used for the analysis of variance (ANOVA). Pearson's correlation coefficient was used to demonstrate whether F 1 seed set rate is correlated with viable gamete rate. Principal component analysis of morphological data were performed in R based on data of morphological traits.

Results
192 different cross combinations between diverse genotypes of emmer wheat × Ae. tauschii (2n = DD or DDDD), durum wheat × Ae. tauschii, T. timopheevii × Ae. tauschii, Ae. crassa × durum wheat, Ae. cylindrica × durum wheat and Ae. ventricosa × durum wheat were made in the field over three successive years. We successfully recovered 56 different synthetic hexaploid and octaploid F 2 lines with AABBDD, AABBDDDD, AAGGDD, D 1 D 1 X cr X cr AABB, D c D c C c C c AABB and D v D v N v N v AABB genomes via in vitro rescue of F 1 embryos and spontaneous production of F 2 seeds on the F l plants. The crossing schemes and corresponding seed morphology of parental species and resulting amphiploids is shown in Fig. 1. Phenotypic diversity in the parent lines. Emmer wheat and Ae. tauschii genotypes showed highly variable spike morphologies (Fig. 2). A high level of diversity overall was found among the parental emmer wheats for most morphological traits measured. Diversity was especially high for flag leaf width (5.6-19.6 mm), spikelets per spike (10.3-29.6; can also be seen in Fig. 2), spike length (6.1-13.6), flowering time (44-93 days from the first of the spring), seed Fe (24.2-66.3 µg/g DW) and Zn (16-62.3 µg/g DW) contents under no fertilizer conditions. Most emmer wheat landraces collected from the Kurdistan province grouped together based on phenotypic traits ( Supplementary Fig. S2). Of these, some genotypes had high Fe and Zn contents in the seed: 'Bainjub' , 'Tirgaran' , 'TazeabadAesef ' , ' Arandan' and 'Hawraman' (Fig. 2 www.nature.com/scientificreports/ between T. turgidum and Ae. tauschii plus four cross combinations between T. timopheevii and Ae. tauschii). Crossability of T. turgidum and Aegilops genotypes based on the F 1 seeds per total pollinated florets is shown in Fig. 3. Crossing success between tetraploid wheat and Ae. tauschii was relatively moderate, but production www.nature.com/scientificreports/ of viable hybrids mostly required embryo rescue. Mean crossabilities in T. turgidum × Ae. tauschii was 0.062, ranging from 0 to 0.38. Of the Ae. tauschii genotypes, accessions ' AE 1211' and 'G 299' showed the highest mean crossability of 0.15 and 0.08 both with emmer wheat genotypes, respectively (Fig. 3A). However, a similar set of emmer wheat couldn't be crossed with each Ae. tauschii genotype due to differences in flowering time of both Ae. tauschii and emmer wheat accessions, therefore the genotypic effect might also be involved. The range of flowering time was especially high for emmer wheat (sown in autumn) ranging from 44 to 93 days from the first day of the spring ( Supplementary Fig. S1, Supplementary Spreadsheet S2). From the 101 cross combinations between T. turgidum (subsp. dicoccum and dicoccoides and durum) and Ae. tauschii, 346 F 1 seeds were obtained belonging to 44 different cross combinations. Of the 346 F 1 seeds produced from T. turgidum wheat × Ae. tauschii crosses, only fourteen had endosperm and could autonomously germinate. These endosperm-containing F 1 seeds belong to T. dicoccum 'Bainjub' × Ae. tauschii ' AE 1211' (2 seeds out of 10), (1 out of 2) and T. durum '1477' × Ae. tauschii 'G 299' (1 out of 6) crosses. The remaining F 1 seeds were shriveled or lacked endosperm and hence required embryo rescue ( Supplementary  Fig. S3). From which, 52% of the F 1 seeds (47 from 90) were successfully rescued, and resulted plants grown to maturity. F 1 seeds from 25 cross combinations successfully reached to maturity and produced amphiploid F 2 seeds (Supplementary Spreadsheet S1). T. timopheevii showed a relatively high crossability (0.085 on average) with Ae. tauschii and produced 167 thin healthy F 1 seeds from four cross combinations that could autonomously germinate. By the end, all crosses produced a total of 29 different synthetic hexaploid and octaploid F 2 lines with www.nature.com/scientificreports/ AABBDD, AABBDDDD or AAGGDD genomes. The number of obtained F 1 and F 2 seeds from these hybrids is shown in Supplementary Spreadsheet S1.
Crosses between tetraploid Aegilops and tetraploid wheat. Different genotypes of D genome containing tetraploid Aegilops species (Ae. crossa, Ae. cylindrica and Ae. ventricosa) and tetraploid wheat lines were used in 87 different cross combinations. The number of F 1 seeds produced per total wheat florets pollinated by each Aegilops is shown in Supplementary Spreadsheet S1 and the crossability rates for each cross combinations is indicated in Fig. 3. Crossing success between tetraploid Aegilops and tetraploid wheat was relatively high and the F 1 seeds could autonomously germinate. A total of 262 octaploid F 2 amphiploid seeds with D 1 D 1 X cr X cr AABB, D c D c C c C c AABB and D v D v N v N v AABB genomes were recovered from 27 different cross combinations. Mean crossability rates in crosses between Ae. crassa, Ae. cylindrica and Ae. ventricosa with tetraploid wheat lines were 0.52, 0.49 and 0.43, respectively. The overall mean crossability in all the crosses was 0.51. Similar to wheat-Ae. tauschii crosses, a similar set of emmer wheat couldn't be crossed with each Aegilops genotype due to differences in flowering time of both Aegilops and wheat accessions. 1322 seeds were produced from crosses between tetraploid wheat and tetraploid Aegilops (e.g. Ae. crossa, Ae. cylindrica and Ae. ventricosa). All of these F 1 seeds had endosperm, but from a sample of 50 F 1 seeds, 28 (0.56) germinated in Petri dishes. 74% of these germinated seeds resulted plants grown to maturity. The number of obtained F 2 seeds from these hybrids is shown in Supplementary Spreadsheet S1.
tauschii, one T. turgidum × Ae. cylindrica and one Ae. ventricosa × T. turgidum hybrid plants (Fig. 4). The analyzed hybrids varied significantly in pollen viability (F = 18.76***; df error = 108) and plump seed set. Four T. turgidum × Ae. tauschii hybrids from different cross combinations produced viable pollen grains at 0.86%, 24.81%, 12.10% and 9.14% frequencies on average. Similarly, the mean frequencies of viable pollen grains in the analyzed  Chromosomal constitutions of the synthetic wheat lines. FISH allowed to identify all the chromosomes in the Triticum and Aegilops parental species, hybrids and F 2 lines, with reference to the chromosome length, arm ratio, and pTa535-1 and GAA banding pattern parameters (Fig. 5, 6, 7). pTa535-1 oligonucleotide probe mainly hybridized with the A-and D-genome chromosomes in combination with the GAA probe. C-banding and FISH using GAA-oligonucleotide probes on accession 49667 of T. dicoccum showed a general agreement in banding patterns (Fig. 5A, Supplementary Fig. S4). Although C-banding generally revealed more bands, banding patterns of both methods were generally similar, confirming that GAA microsatellite loci colocalize with C-bands in the genus Triticum. Cytogenetic analysis also unexpectedly revealed that accession ' AE 1211' of Ae. tauschii was autotetraploid with 2n = 4x = 28 chromosomes (Fig. 5B); interestingly, this accession was also one of the most fertile parents in the cross with emmer wheat. Based on GAA and pTa535-1 banding patterns, a balanced reciprocal translocation involving chromosomes 5A t and 6G was identified in T. timopheevii accession '131212' resulting in 5A t S.6GL and 6GS.5A t L translocated chromosomes (Fig. 5C). This translocation was also observed in an amphiploid produced from a cross between this line and Ae. tauschii ' AE 1602' (Fig. 5F). Some chromosomal rearrangements or translocations were identified that are induced by polyploidization in the evaluated synthetic amphiploids, including one small heterozygous deletion at the distal end of the 1BL chromosome arm in T. durum '78' × Ae. tauschii '191' line (Fig. 6D) and a single 2C c chromosome showing deletion/translocation in an amphiploid genotype from a cross between Ae. cylindrica '236'-T. durum '17' (Fig. 7D). This genotype was a monosome with 2n = 8x = 55 chromosomes. The monosomic status of this genotype was further confirmed by chromosome counting in nine different mitotic metaphase cells of this plant (Supplementary Fig. S7). We checked the chromosome number of two Ae. ventricosa-T. turgidum and four Ae. crassa-T. turgidum amphiploid seedlings. Six different mitotic metaphase cells from a single Ae. ventricosa ' AE 1511'-T. turgidum '13' amphiploid plant were checked and all showed 54 chromosomes (Supplementary Fig. S8), while all the checked cell of another single amphiploid seedling (i.e. Ae. ventricosa ' AE 357'-T. turgidum '11') showed 53 chromosomes (Supplementary Fig. S8). On the other hand, all the four Ae. crassa-T. turgidum amphiploids were complete disomic plants with 2n = 8x = 56 chromosomes Phenotypic diversity of the amphiploids. The produced amphiploids showed a high variation for most morphological traits measured (Table 1). We also observed a high level of diversity in spike morphology (Supplementary Fig. S5). Diversity was especially high for flag leaf width (5.6-19.6 mm), spikelets per spike (10.3-29.6;

Discussion
Global wheat production must substantially increase from its current level, to ensure food for a growing world population. This require continuous breeding activity and involvement of new and less exploited genetic resources. Here, in order to incorporate the genetic diversity of emmer wheat and exotic D genomes for the future wheat breeding, a collection of synthetic wheat lines and amphiploids were produced from crosses between tetraploid Triticum species and subspecies (AABB or AAGG genome) and the D genome containing species Ae. tauschii (diploids and one tetraploid with DD and DDDD genomes), Ae. crassa (D 1 D 1 X cr X cr genome), Ae. cylindrica (D c D c C c C c genome) and Ae. ventricosa (D v D v N v N v genome) and amphiploids with AABBDD, AABBDDDD, AAGGDD, D 1 D 1 X cr X cr AABB, D c D c C c C c AABB and D v D v N v N v AABB genomes were generated. Ae. crassa has tetraploid and hexaploid cytotypes but tetraploid cytotypes were used in the present study ( Fig. 7; Supplementary Fig. S9). We already have established FISH-based karyotypes of 'Bookan' and 'Sanandaj' accessions of this species 26 . Other D genome containing Aegilops species include Ae. juvenalis (Thell.) Eig (D 1 D 1 X cr X cr U j U j ) and Ae. vavilovii (Zhuk.) Chennav. (D 1 D 1 X cr X cr S v S v ) which were not used in the present study. However amphiploids from crossing durum wheat with these two species have been recently developed 34 . The produced amphiploids in the present study showed a high variation in morphological traits (Table 1). Based on phenotypic traits, amphiploids with the same genome composition grouped together in principal component analysis, although a higher variation was observed for the Ae. crassa-wheat amphiploids (Fig. 8).
We produced novel synthetic wheat lines using durum, T. dicoccum and T. dicoccoides wheat genotypes toward increasing the genetic diversity of all the bread wheat subgenomes. Highly variable emmer wheat and Ae. tauschii genotypes (as reflected by the spike morphologies in Fig. 2 and Supplementary file Figs. S1, S2) were used in the www.nature.com/scientificreports/ www.nature.com/scientificreports/ crosses. In total, 29 different synthetic hexaploid and octaploid F 2 lines were recovered with AABBDD, AABBD-DDD or AAGGDD genomes (Supplementary Spreadsheet S1). A high variation in the crossability rate among Ae. tauschii genotypes, with the highest interspecific crossability observed for G 299 and AE 1211 (the tetraploid accession) (Fig. 3). However, most F 1 seeds recovered lacked an endosperm and required embryo rescue to germinate. In contrast, Ogbonnaya et al. 35 found an Ae. tauschii genotype that produced endosperm-containing F 1 seeds when crossed with wheat. In wheat × Aegilops crosses, wheat genotypes differ in their crossability with Ae. tauschii 36 . In fact, effect of major crossability genes in common wheat such as Kr1 (5BL), Kr2 (5AL), Kr3 (5D) and SKr (5BS) is well known for the obtaining F 1 hybrids 37,38 and wheat genotypes with dominant alleles generally show less crossability. The presence of crossability genes in tetraploid forms of wheat has not been studied in detail. However, there are reports suggesting chromosomes 7A and 4B are involved, so the crossability in tetraploid and hexaploid wheat might be controlled by different genetic systems 39,40 . Many synthetic hexaploids has been produced in CIMMYT, from crosses between durum wheat (T. turgidum) cultivars and Ae. tauschii accessions 23,41 . However, synthetic wheat production has mainly been confined to the crossing of durum wheat with Ae. tauschii. Although, emmer-based hexaploid lines has been developed directly from emmer wheat × hexaploid wheat crosses and backcrossing to hexaploid wheat 42,43 and useful traits, such as protein content and test weight 42 and water-use-efficiency of grain production 43 was introduced from emmer to hexaploid wheat.
In addition to the synthetic wheat lines, 27 different octaploid F 2 amphiploid lines were recovered from crosses between Tetraploid Aegilops species (i.e. Ae. crassa, Ae. cylindrica and Ae. ventricosa) as female parents and T. turgidum genotypes as male parents (Supplementary Spreadsheet S1). A high rate of endosperm containing plump seeds from tetraploid Aegilops species (♀) × tetraploid wheat (♂) crosses were produced in the present study (Fig. 3). The overall mean crossability in all the crosses between tetraploid Aegilops species (i.e. Ae. crassa, Ae. cylindrica and Ae. ventricosa) and T. turgidum was 0.51 which is apparently higher than the mean crossability recorded for Triticum-Ae. tauschii crosses (i.e. 0.062). Tetraploid and hexaploid Aegilops species generally show higher crossability than diploid Aegilops species when crossed with common or tetraploid wheat and generally tend to set endosperm more resulting in plump F 1 seeds 44,45 . Spontaneous F 2 seed production has been reported in Ae. crassa × T. persicum hybrids 21 . Delibes and Garcia-Olmedo 46 reported hybridization between wheat and Ae. ventricosa. Yuan, et al. 47 reported the production of Ae. cylindrica × T. aestivum hybrids that showed 0 to less that 1% seed set in back-crosses. Occurrence of interspecific hybrids between Ae. cylindrica and T. aestivum has also been reported in the field 48 . Fakhri et al. also concluded that lack of F 2 seed in reciprocal crosses between T. aestivum and Ae. cylindrica hybrids might be due to lack of meiotic restitution and low rate of viable gametes 19 . Xu and Dong 21 also reported complete sterility of Ae. cylindrica × T. persicum hybrids. While it was very hard to cross tetraploid Aegilops species as the male parent to T. turgidum ( Supplementary Fig. S6), the reverse cross direction was more successful in the present study.  www.nature.com/scientificreports/ We analyzed the rates of unreduced male gametes in 12 different hybrid plants and found a significant correlation between unreduced gamete and seed set. Hence, we believe that the F 2 seeds are mainly the product of unreduced gametes rather than somatic chromosome doubling. While Ae. cylindrica has been reported to spontaneously cross with T. aestivum and frequently crossed with T. aestivum artificially, the resulting hybrid plants (2n = ABDD c C c ) are sterile and do not produce viable gametes 19,49 . In our work, F 1 plants from the crosses between T. durum (♀) and Ae. cylindrica (♂) were completely sterile and no F 2 seeds were produced ( Fig. 4G; Supplementary Fig. S6), while healthy F 1 seeds were produced from reverse crosses. The germination rate of F 1 seeds from Ae. cylindrica × T. turgidum was very low and only two different F 1 plants (from two different cross types) reached to maturity and 6 F 2 seeds were harvested (Supplementary Spreadsheet S1; Supplementary  Fig. S5). These results confirm that crossability highly depends on parental species, their ploidy level and cross direction. Interestingly, we produced eight auto-allo-octaploid synthetic wheat line with AABBDDDD genome in F 2 using a tetraploid Ae. tauschii as the male parent (Fig. 6C). But the stability of these line may be affected by their auto-allopolyploidy status during the next generations.
One small heterozygous deletion at the distal end of the 1BL chromosome arm in the F 2 generation of T. durum '78' × Ae. tauschii ' AE 191' line (Fig. 6D) and a monosomic amphiploid genotype from a cross between Ae. cylindrica '236'-T. durum '17' (D c D c C c C c AABB genome) carrying a single rearranged 2C c chromosome were identified (Fig. 7). No other chromosomal arrangement or translocation induced by polyploidization, was identified in the evaluated synthetic wheat lines implying their genome stability.
All the four evaluated Ae. crassa-T. turgidum amphiploids (D 1 D 1 X cr X cr AABB genome) showed a complete set of 2n = 8x = 56 chromosomes. Contrary to the Ae. cylindrica-T. turgidum F 2 plants which set plump and shriveled seeds in less than 50% of the florets, Ae. crassa-T. turgidum amphiploids F 2 plants were completely fertile and set plump healthy seeds. Naranjo and Benavente 50 observed high levels of chiasmata in Ae. crassa-wheat hybrids with the Ae. crassa cytoplasm, suggesting that Ae. crassa cytoplasm induces homoeologous pairing. Such a cytoplasmic effect-if available in our Ae. crassa-T. turgidum amphiploids-may lead to chromosomal rearrangement in the next generation, however, no rearrangement was detected in an Ae. crassa-T. turgidum individual amphiploid analysed by FISH (Fig. 7B).
The two evaluated Ae. ventricosa-T. turgidum and one Ae. cylindrica-T. turgidum amphiploid individuals were aneuploids with 2n = 54 ( Supplementary Fig. S8, 2n = 53 ( Supplementary Fig. S8) and 2n = 55 (Supplementary Fig. S7) chromosomes. Such a chromosome elimination commonly results from production and union of partially unreduced gametes in F 1 plants which is caused by meiotic irregularities such as uni-and multivalent formation and lagging chromosomes 44 . Chromosome loss may also happens in upcoming generations. The stability of amphiploids from the Triticum or Aegilops is also affected by parental species and genotype, the effect of specific genes and the rate of parental genome affinity 40,45,[51][52][53] . However, chromosome elimination in offspring of high ploidy level amphiploids may lead to stabile partial amphiploids over subsequent generations 54 . More genomic analysis of the offspring is required to find out the genomic stability and transmission, because going through more rounds of meiosis would provide chance for possible genome rearrangements.
Crossing between the synthetic wheat and amphiploid lines produced in the present study with T. aestivum and repeated generations of self-pollination can generate bread wheat lines with recombined new subgenomes. In this way, new introgression lines with useful phenotypic traits can be recovered. Retention of individuals with only D chromosomes is also possible using marker assisted selection which results in new wheat lines with recombined new D subgenomes 18 .

Conclusion
Here, a lot of crosses were made between tetraploid wheat (i.e. emmer wheat, T. durum and T. timopheevii) and the D-containing Aegilops species and successfully recovered various synthetic hexaploid and octaploid F 2 lines with AABBDD, AABBDDDD, AAGGDD, D 1 D 1 X cr X cr AABB, D c D c C c C c AABB and D v D v N v N v AABB genomes via crossing, in vitro rescue of F 1 embryos and spontaneous production of F 2 seeds on the F l plants. Diverse genotypes of emmer wheat and Aegilops species were used in the crosses and various forms of D subgenomes were brought together in the produced amphiploids. Contribution of D genome bearing Aegilops species and the less-investigated emmer wheat genotypes as parents in the crosses resulted in novel synthetic wheat and amphiploids which can further be used as bridges to expand the genetic variation of wheat beyond its current status via crossing and backcrossing.