Production and identification of wheat-Agropyron cristatum 6P translocation lines

The narrow genetic background of wheat is the primary factor that has restricted the improvement of crop yield in recent years. The kernel number per spike is the most important factor of the many potential characteristics that determine wheat yield. Agropyron cristatum (L.) Gaertn., a wild relative of wheat, has the characteristics of superior numbers of florets and kernels per spike, which are controlled by chromosome 6P. In this study, the wheat-A. cristatum disomic addition and substitution lines were used as bridge materials to produce wheat-A. cristatum 6P translocation lines induced by gametocidal chromosomes and irradiation. The results of genomic in situ hybridization showed that the frequency of translocation induced by gametocidal chromosomes was 5.08%, which was higher than the frequency of irradiated hybrids (2.78%) and irradiated pollen (2.12%). The fluorescence in situ hybridization results of the translocation lines showed that A. cristatum chromosome 6P could be translocated to wheat ABD genome, and the recombination frequency was A genome > B genome > D genome. The alien A. cristatum chromosome 6P was translocated to wheat homoeologous groups 1, 2, 3, 5 and 6. We obtained abundant translocation lines that possessed whole-arm, terminal, segmental and intercalary translocations. Three 6PS-specific and four 6PL-specific markers will be useful to rapidly identify and trace the translocated fragments. The different wheat-A. cristatum 6P translocation lines obtained in this study can provide basic materials for analyzing the alien genes carried by chromosome 6P. The translocation line WAT33-1-3 and introgression lines WAI37-2 and WAI41-1, which had significant characteristics of multikernel (high numbers of kernels per spike), could be utilized as novel germplasms for high-yield wheat breeding.


Introduction
Wheat is a major cereal crop produced worldwide. The spike number per acre, kernel number per spike and 1,000kernel weight are the main parameters that inXuence its yield. Long-term breeding practices have shown that the kernel number per spike is the factor that can most easily be improved to increase crop yield. A fundamental method of increasing wheat yield would be to widen its narrow genetic background, which has restricted earlier attempts to improve yield. The wild relatives of wheat possess many excellent genes and play an important role in wheat genetic improvement (Wu et al. 2006;Faris et al. 2008).
Agropyron Gaertn. (P genome), distributed in the low temperature Altiplano and Sandlot in Eurasia, is resistant to abiotic stress and major diseases, and has many high-yield characteristics, e.g., more tillers, more spikelets and more Xorets (Dewey 1984;Dong et al. 1992;Badaeva et al. 2007). Therefore, the Agropyron P genome can be used as donor to provide genes for the genetic improvement of wheat crops. Wide crosses between Agropyron and common wheat began in the 1940s (White 1940;Smith 1942) and was Wrst successfully reported in 1989 (Chen et al. 1989). Till now, the hybridization of common wheat with Agropyron species such as A. cristatum, A. disertorum, A. michnoi and A. fragile was obtained at the end of the 1980s and the beginning of the 1990s (Chen et al. 1989;Limin and Fowler 1990;Ahmad and Comeau 1991;Dong 1991, 1993;Li et al. 1995;Jauhar 1992). The production of wheat-Agropyron cristatum (L.) Beauv. disomic addition lines has provided the possibility of introducing A.cristatum genes into wheat (Li et al. 1997(Li et al. , 1998. The previous results of our laboratory have reported that a gene(s) controlling high numbers of Xorets and kernels per spike is located on chromosome 6P of the wheat-A.cristatum addition and substitution lines (Wu et al. 2006). Therefore, the production of wheat-A.cristatum 6P translocation and introgression lines, which had transferred the chromosome 6P fragment carrying the multikernel gene into a wheat background, may improve kernel number and thus increase crop yield.
Ionizing radiation is an eVective method of inducing chromosomal structural variations and of transferring genes (i.e., for disease resistance and high yield) from wild relatives to common wheat (Sears 1956;Sears and Gustafson 1993). In addition, gametocidal chromosomes originating from Aegilops can also cause structural mutation of both the wheat chromosomes (Endo 1996(Endo , 2007Friebe et al. 2000Friebe et al. , 2003Nasuda et al. 2005) and the wild relative chromosomes in a wheat background (Li et al. 2003;Sun et al. 2004;Chen et al. 2005).
The objectives of this work were to produce wheat-A. cristatum 6P translocation lines with the characteristics of multikernel and to stabilize their inheritance in a wheat background. These newly created translocation lines provide novel germplasms and valuable materials for both basic and applied research for high-yield wheat breeding.

Hybrid combination and induction techniques
Hybrid seeds were produced between the wheat-A.cristatum disomic addition line as the female parent and the 'Gaocheng8901' as the male parent. The seeds were exposed to 60 Co gamma rays at a radiation dose of 250 Gray (Gy) and a dose rate of 1.472 Gy/min.
The mature spikes of wheat-A.cristatum disomic substitution line were exposed to 60 Co gamma rays at a radiation dose of 9 Gy and a dose rate of 0.861 Gy/min. Fresh 60 Co-irradiated pollen was collected from dehiscent anthers and used to pollinate the female parent 'Fukuhokomugi', which had been artiWcially emasculated 1-2 days before.
The other combination was between the CS-G2C addition line as the female parent and wheat-A.cristatum addition line as the male parent.

Chromosome preparation
Plant root tips were used to prepare chromosome preparations by the conventional squashing method (Cuadrado et al. 2000). Cytological observations were performed under a BX51 Olympus phase-contrast microscope (Olympus Corp., Tokyo, Japan) and the images were taken with a digital camera. Slides with good chromosome spreading were frozen in liquid nitrogen until needed for GISH/FISH detection.

GISH detection of the translocation lines and FISH identiWcation of the translocated wheat chromosomes
To identify the translocations of progeny, GISH was carried out in root tip cells using A.cristatum P-genomic DNA as probe. 'Fukuhokomugi' genomic DNA alone or an equivalent mixture of 'Fukuhokomugi' and 'CS' (or 'Gaoch-eng8901', according to crossing parents) genomic DNA was used for blocking.
The P-genomic DNA was used as probe to detect P-genome chromosome fragment; the barley clone pHvG39, the Aegilops tauschii Coss. clone pAs1 or the rye clone pSc119.2 was used as probe to identify wheat translocated chromosomes.
The clones pAs1, pSc119.2 and pHvG39 have speciWc hybridization signals on wheat A, B and D chromosomes. The clone pAs1 is used to detect mostly the D-genome chromosomes. The clone pSc119.2 can identify the B-genome chromosomes. Although the clone pHvG39 can hybridize with A, B and D genomes, the hybridization signal is particularly strong in the centromere of the B-genome chromosomes while it hybridizes weakly with the A and D genomes. All the chromosomes of wheat can be distinguished using pAs1 and pHvG39 as probes together. Dual-color FISH and GISH/ FISH were carried out to identify the homoeologous group of the translocated wheat chromosomes. Dual-color FISH was performed using pAs1 and pHvG39 (Pedersen and Langridge 1997) or pAs1 and pSc119.2 (Mukai et al. 1993) plasmid DNAs as probes. Subsequently, GISH detection was carried out on the same slide using the P-genomic DNA as a probe. Dual-color GISH/FISH was performed using the P-genomic DNA and one of pHvG39, pAs1 or pSc119.2 as probes. Finally, the homoeologous groups of translocated wheat chromosomes were analyzed and veriWed according to the type of probe, the intensity and location of hybridization signal, the size of chromosome and the ratio of the arms.
The Biotin (BIO)-Nick Translation Mix, Digoxigenin (DIG)-Nick Translation Mix, avidin-Xuorescein and antidigoxigenin-rhodamine were purchased from Roche (Switzerland). The in situ hybridization methods utilized were previously described by Cuadrado et al. (2000), except that the rinsing steps following hybridization were modiWed with 0.5£ saline sodium citrate (SSC) instead of 0.1£ SSC. In situ hybridization images were observed under a Nikon Eclipse E600 (Japan) Xuorescence microscope and captured with a CCD camera (Diagnostic Instruments, Inc., Sterling Heights, MI, USA).

Molecular marker analysis
Seven 6P-speciWc sequence-tagged site (STS) markers were designed based on suppression subtractive libraries constructed in our laboratory. These STS markers and one sequence-characterized ampliWed region (SCAR) marker (SC5 815 ) were used to characterize the genomic composition of the wheat-A.cristatum translocation lines (primer sequences are shown in Table 1). An A. cristatum accession Z559 and the wheat-A.cristatum addition line were used as positive controls, while 'Fukuhokomugi', 'CS' and 'Gaocheng8901' were used as negative controls. For PCR ampliWcation, 25 l of each reaction mixture containing 100 ng genomic DNA, 1 U Taq polymerase, 0.2 mM of each dNTP, 0.25 mM of primer and 1£ PCR buVer were used. The DNA ampliWcation program comprised incubation at 94°C for 5 min followed by 35 cycles of 1 min at 94°C, 1 min at the annealing temperature speciWc for each primer and 1.5 min at 72°C, followed by a Wnal 10 min extension at 72°C. The PCR ampliWcation products were visualized with electrophoresis as previously described by Wu et al. (2006).

GISH detection of hybird progeny
The GISH results of the hybird progeny are shown in Table 2. GISH analysis of 236 M 2 progeny from irradiated hybrid seeds (wheat-A.cristatum addition line/'Gaoch-eng8901') showed that Wve plants (2.12%) had a translocation between the chromosome 6P and wheat chromosomes. Among them, plant WAT36 had a terminal translocation and one pair of the 6P chromosomes (Fig. 1a). Plant WAT30 had a translocation involving a large 6P fragment. The fragment ratios (large/small) were diVerent in two reciprocal translocation plants WAT31 and WAT32, indicating that the breakpoints occurred at diVerent sites on chromosome 6P (Fig. 1b, c). Plant WAT33 was homozygous for a translocation involviong a 6P terminal fragment (Fig. 1d). In addition, plants WAI37 and WAI41, which  Fig. 1e), was detected by GISH among the 36 M 2 progeny of 'Fukuhokomugi' pollinated with the irradiated pollen of the wheat-A.cristatum substitution line.
Among the 354 F 2 progeny of the CS-G2C and wheat-A. cristatum addition lines, 18 translocation plants were veri-Wed. The frequency of plants carrying translocations generated by the gametocidal chromosome was 5.08%. Various translocation types were found, including nine plants with whole-arm translocations (Fig. 1f, g), three plants with fragmental translocations (Fig. 1h), two plants with terminal translocations, one plant with fragmental and intercalary translocations in the same chromosome (Fig. 1i), one plant with a whole-arm and a terminal translocation in diVerent chromosomes (Fig. 1j), and one plant with a dicentric translocation (Fig. 1l). Moreover, we found a chimera, which had a dicentric or terminal translocation in diVerent root tip cells (Fig. 1k). Of the 18 translocation plants, 4 died or did not set seeds.

IdentiWcation of wheat chromosomes involved in the translocations
The wheat chromosomes translocated with chromosome 6P were identiWed by FISH, as previously described by Pedersen and Langridge (1997) and Mukai et al. (1993), combined with GISH. The results are shown in Table 3.

Molecular marker identiWcation of the translocation lines
Fifteen translocation plants and two introgression plants were characterized using A.cristatum chromosome 6P-speciWc markers (partial results are shown in Fig. 3). The molecular marker analysis of diVerent whole-arm translocation lines (including short-arm and long-arm translocations) conWrmed that For3-G02 260 , For5-E08 121 and For22-B10 185 were 6PS-speciWc markers, because they were positive in the short-arm translocation lines (WAT18-1 and WAT24-1). Simultaneously, For8-G11 181 , For14-B02 114 , For15-D06 610 and For22-E03 150 were 6PL-speciWc markers because they were positive in the long-arm translocation lines (WAT17-1, WAT19-1, WAT23-1 and WAT26-3). The sites of the SCAR marker (SC5 815 ) were distributed on both the short and long arms of A.cristatum chromosome 6P. Further molecular analysis of the remaining translocation lines was carried out to determine whether the translocated fragments of chromosome 6P belonged to the short or the long arm (Table 4).
The breakpoints of chromosome 6P were estimated according to relative length of the translocated 6P fragment compared with the complete chromosome 6P. The position of the centromere is considered as 0, while the terminal end of the short and long arms is considered as 1.
Part of the long-arm speciWc markers were positive in WAT21-1, while all the short-arm speciWc markers were negative. So, the translocated fragment of chromosome 6P of WAT21-1 was (0.7-1)L. The translocated fragment in WAT21-2 was deWned as S + (0-0.7)L because bands for all of the short-arm and part of the long-arm speciWc markers were observed. Using the same method, we identiWed the translocated fragment of WAT25-2 as L + (0-0.4)S and the translocated fragment of WAT27-2 as (0.2-1)S. Positive identiWcation of the translocated fragment (i.e., belonging to the short or long arm) of WAT25-1, WAT29-7 and WAT33-1-3 could not be made, because the translocated homozygous terminal translocation, e WAT29: homozygous intercalary translocation, f WAT23: whole-arm translocation, g WAT26: whole-arm translocation and a partial translocated wheat chromosome was deleted, h WAT34: a large fragmental translocation, i WAT27: fragmental and intercalary translocation occurred in the same chromosome, j WAT35: whole-arm and terminal translocation occurred in diVerent chromosomes, k WAT20: a chimera, which had dicentric translocation and terminal translocation detected in diVerent root tip cells of one plant, l WAT19: dicentric translocation. The P-genomic DNA signal to detect P chromosomes appears as a yellow-greenish color. Arrows point to the diVerent translocated chromosomes fragments were too small to be characterized by any of the markers. Similarly, the large fragment of WAT34-2 was not identiWed because it was close to a complete A.cristatum chromosome 6P and was positive for all the markers. WAT31-13-2 and WAT31-13-3 were diVerent types of translocations, which originated from WAT31-13 (5A-6P fragmental reciprocal translocation are described in Table 3). WAT31-13-3 with reciprocal translocations involving chromosome 6P was positive for seven markers, which indicated that the large and small fragments constituted a complete chromosome 6P. WAT31-13-2 with a small fragment was positive for three short-arm speciWc markers and was identiWed as (0.3-1)S.
Based on the relationship between the size of the translocated fragment and molecular markers, it was inferred that two 6PS-speciWc markers For3-G02 260 and For5-E08 121 were located in the (0.4-1)S region of the short arm of chromosome 6P and that another 6PS-speciWc marker For22-B10 185 was located in the (0.3-0.4)S region. There was one 6PL special marker For15-D06 610 located in the (0-0.7)L region, and three other 6PL special markers (For8-G11 181 , For14-B02 114 and For22-E03 150 ) were located in the (0.7-1)L region of chromosome 6P. The multikernel plants WAI37-2 and WAI41-1 had no GISH signals, but were positive for SC5 815 detection. These results implied that the translocated fragment of chromosome 6P in WAT33-1-3, WAI37-2 and WAI41-1 may control the characteristics of multikernel. EVorts to obtain a uniform genetic background by single-parent backcrossing is currently in progress and will lay the foundations for further analysis of the multikernel gene(s).

Discussion
Ionizing radiation breaks chromosomes at random and may cause small alien segmental translocation and intercalary translocations (Jiang et al. 1993). In this study, the translocation types we created included segments reciprocal and terminal as well as intercalary translocations. The breakpoints of chromosome 6P were situated near the centromere and the terminal.
Aegilops cylindrica gametocidal chromosome 2C can induce alien chromosomal structural variations in many wild relatives with a wheat background (Endo 1996;Chen et al. 2005;Shi et al. 2005). However, chromosome translocations between wheat and the Agropyron Gaertn. genome P have not been widely reported. Dewey (1984) considered that it might not be possible to transfer genes from Agropyron to Triticum even if the interfeneric hybrids can be obtained. But the addition of chromosome pairs from P genome to wheat may be a practical plant breeding option. In this study, the alien translocation lines were successfully obtained utilizing the wheat-A. cristatum 6P addition line. Abundant types of chromosome variations and a high frequency (5.08%) of translocation were induced by gametocidal chromosome 2C. Half of the translocation lines were whole-arm translocations, which indicated that both the A. cristatum and wheat chromosomes were more easily broken at the site of the centromere. In addition to the proximal centromere, the breakpoints were also distributed near the terminal. Finally, there were also double translocations (diVerent translocation types in two chromosomes), complex translocations (diVerent translocation types in one chromosome) and a chimera (diVerent translocation types in diVerent root tip cells of one plant) detected in the progeny of CS-G2C/wheat-A. cristatum addition line. The chimeric phenomenon may be the outcome of the somaclonal variation that occurred during mitosis.
The identiWcation of partial translocation lines showed that the A. cristatum chromosome 6P could be translocated to wheat ABD genome, and the recombination frequency was A genome > B genome > D genome. The results diVered from the previous reports that the B-genome chromosomes were involved in translocations most frenquently, followed by the A-and D-genome chromosomes (Endo et al. 1994;Li et al. 2003;Badaeva et al. 2007).
Numerous disease resistance genes have been successfully translocated into wheat from wild relatives using methods of chromosome engineering (Friebe et al. 1996;Dubcovsky et al. 1998;Chen et al. 2005;Faris et al. 2008). However, reports on the translocation of genes controlling yield traits are rare.
In this study, the radiation-induced translocation progeny showed excellent agronomic features. SpeciWcally, the pollen parent T. aestivum cv. 'Gaocheng8901' used in irradiated hybrid seeds was high-quality strong gluten wheat, and as such its translocation progeny had more productive potential. These translocation lines might combine high yield and quality characteristics and therefore have a high application value. The small segmental translocation lines (WAT33-1-3) and the introgression lines (WAI37-2 and WAI41-1), which had signiWcant characteristics of multikernel and inherited more stably than the addition lines, could be used directly by wheat breeders.
In this study, we also created a high frequency of wheat-A.cristatum translocation lines induced by a gametocidal chromosome. Though 22.2% of the translocation lines did not survive or produce seeds normally because of imbalanced chromosome compositions, the high frequency of translocation also made the gametocidal chromosome an eVective method to induce mutation and construct a series of translocated chromosome fragments. This series of fragments allowed us to accumulate abundant genetic material for cloning the A.cristatum gene(s). In the future, the gametocidal chromosome-induced translocation lines could be improved by backcrossing.