Introduction

Wheat (Triticum aestivum L., 2n = 6x = 42, genomes AABBDD) is one of the most important crops worldwide. A major challenge in modern agriculture is to breed elite wheat varieties with enhanced agronomic traits to meet the growing demands for food. However, common wheat exhibits a relatively narrow range of genetic variation, which has become a bottleneck for yield improvement (Dubcovsky and Dvorak 2007; Haudry et al. 2007; Reif et al. 2005). To broaden the genetic basis of common wheat, its wild relatives harboring superior agronomic traits have often been used in wheat breeding (Wang 2011). The tribe Triticeae contains around 400 species and 25 genera, conferring ample genetic diversity for wheat improvement (Mujeeb-Kazi et al. 2013). To date, more than 100 alien genes/QTLs conferring superior traits have been transferred into cultivated wheat (Bao et al. 2009; Cao et al. 2011; Faris et al. 2008; Friebe et al. 1990, 1996a; Fu et al. 2012; He et al. 2009; Hua et al. 2009; Jiang et al. 1994; Kang et al. 2011; Luo et al. 2009; Monneveux et al. 2003; Sharma et al. 1995; Singh et al. 1998; Wang et al. 2011). Producing wheat-alien species translocation lines and elucidating their genetic constitutions are key steps for effective transfer of desirable genes into common wheat (Chen et al. 1995, 2005; Friebe et al. 1992, 1996a; Gill et al. 2011; Klindworth et al. 2012; Larkin et al. 1995; Larson et al. 2012; Niu et al. 2011, 2014; Qi et al. 2011; Singh et al. 1998; Wang and Zhang 1996; Yu et al. 2009). Some translocation lines, especially 1BL∙1RS, 6VS∙6AL, and 7DL∙7Ag have played important roles in wheat improvement. A 1BL∙1RS translocation line was widely used in bread wheat-breeding programs throughout the world due to the presence of powdery mildew resistance gene Pm8 and rust resistance genes Sr31, Lr26, and Yr9 in 1RS (Hsam and Zeller 1997; Lukaszewski 2000; Mago et al. 2002; Singh et al. 1998). A 6VS∙6AL translocation line (Chen et al. 1995, 2005) has been widely used throughout the world (Duan et al. 1998; Liu et al. 1999; Qi et al. 1995). The 7DL∙7Ag translocation lines produced from Lophopyrum elongatum were reported to be valuable reservoirs of desirable genes conferring resistance to leaf rust, salinity, and waterlogging (Deal et al. 1999; Ma et al. 2000; McDonald et al. 2001; Niu et al. 2014).

Agropyron cristatum (L.) Gaertn (2n = 4x = 28; genomes PPPP) is a perennial wheatgrass that possesses many desirable traits such as enhanced fertile tiller number, high grain number per spike, and resistance to numerous diseases (Dewey 1984; Dong et al. 1992). It has long been considered a useful genetic resource for wheat improvement. The F1 hybrids were successfully obtained between common wheat cv. Fukuhokomugi (Fukuho) and A. cristatum accession Z559, followed by backcrossing with Fukuho for several generations (Li et al. 1995, 1998b; Li and Dong 1991, 1993). A series of disomic addition lines was obtained, and the 6P disomic addition line 4844-12 was one of them. This addition line had significantly higher grain number per spike and floret number per spikelet, as well as enhanced resistance to powdery mildew, compared to its wheat parent (Han et al. 2014; Li et al. 1997, 1998a, b; Luan et al. 2010; Wu et al. 2006). These results indicated the existence of desirable genes on the A. cristatum 6P chromosome. To produce germplasm useful in wheat-breeding programs, wheat-A. cristatum 6P translocation lines were then produced by both gametocidal chromosomes and ionizing radiation (Luan et al. 2010). WAT31-13 (M2 generation), produced from irradiated hybrid seeds (wheat-A. cristatum addition line/Gaocheng 8901), is a 5A-6P reciprocal translocation line with 44 chromosomes. Although it displays superior agronomic traits, it is genetically unstable (Luan et al. 2010).

To further localize these desirable genes on the A. cristatum 6P chromosome and acquire translocation materials with genetic stability, strict self-pollination was carried out over further generations of WAT31-13. Forty-three stable translocation lines in the M8 generation were selected. Surprisingly, most of them contained 6P translocation chromosome segments that were different from those in WAT31-13, although the various elite agronomic traits were retained. In this study, we not only examined the genomic constitutions of the 43 translocation lines by GISH and FISH, but also ascribed superior agronomic traits (especially the high fertile tiller number) to specific A. cristatum 6P chromosomal segments.

Materials and methods

Materials

WAT31-13 is a 5A-6P reciprocal translocation line (2n = 44) produced from irradiated hybrid seeds using the wheat-A. cristatum 6P disomic addition line 4844-12 as the female parent and common wheat Gaocheng 8901 (2n = 6x = 42, genomes AABBDD) as the male parent (Luan et al. 2010). In this study, 43 stable M8 translocation lines were obtained from the progenies of WAT31-13. Wheat-A. cristatum 6P disomic addition line 4844-12, common wheat Gaocheng 8901, and Fukuho were used as contrasting parents in this study. Wheat-A. cristatum 6P disomic addition line 4844-12 and translocation line WAT31-13 were originally produced in our laboratory. Seeds of Gaocheng 8901 and Fukuho were provided by the Institute of Crop Sciences, Chinese Academy of Agricultural Sciences.

GISH and FISH analysis

Genomic in situ hybridization (GISH) and dual-color fluorescence in situ hybridization (FISH) were carried out as previously described (Han et al. 2003). Genomic DNA was isolated using the CTAB method (Allen et al. 2006). A. cristatum genomic DNA (labeled with Dig-Nick-Translation Mix) and Fukuho genomic DNA were used as probe and blocker, respectively. Wheat and A. cristatum chromosomes were pseudo-colored as blue and red, respectively. Dual-color FISH was performed using pAs1 and pHvG39 plasmid DNAs as probes in all translocation subtypes, while probes CRW (centromeric retrotransposon of wheat) and pAcCR1 (centromeric retrotransposon of A. cristatum) were applied to analyze the centromere of the arm translocation chromosome (TrA) in the Subtype II. Probe pAs1 from Aegilops tauschii preferentially hybridizes to most D-genome chromosomes, and the probe pHvG39 could hybridize with A, B, and D genome chromosomes (Pedersen and Langridge 1997). Probe CRW was identified from a Triticum boeoticum library and closely associated with the centromeres (Liu et al. 2008). The A. cristatum centromere-specific probe pAcCR1 was developed using the wheat-A. cristatum 6PS ditelosomic lines in our laboratory. To develop the probe pAcCR1, microdissection and degenerate oligonucleotide-primed PCR (DOP-PCR) were conducted as previously described (Vega et al. 1994). All cytological images were taken under a Nikon Eclipse E600 fluorescence microscope and captured with a CCD camera.

Translocation breakpoint analyses

To determine the translocation breakpoints, a total of 25 wheat SSR markers physically or genetically mapped on wheat chromosome 5A were chosen (Somers et al. 2004; Sourdille et al. 2004). Common wheat ‘Chinese Spring’ (CS) nulli-tetrasomic lines (N5AT5B and N5AT5D), CS ditelosomic lines (Dt5AL and Dt5AS), and CS 5AL deletion lines (5AS10-0.98-1.00, 5AS3-0.75-0.98, 5AS1-0.40-0.75, C-5AS1-0.40, C-5AL12-0.35, 5AL12-0.35-0.57, 5AL10-0.57-0.78, 5AL17-0.78-0.87, and 5AL23-0.87-1.00) were applied to verify the chromosomal localization of wheat SSR markers. Nine A. cristatum-specific markers developed from EST sequences of A. cristatum mRNA transcriptome and seven A. cristatum-specific markers (For3-G02, For5-E08, For22-B10, For15-D06, For8-G11, For14-B02, and For22-E03) obtained from Luan et al. (2010) were chosen to distinguish different chromosomal regions of A. cristatum chromosome 6P. All wheat SSR primers were obtained from the Graingenes website (http://wheat.pw.usda.gov/ggpages/maps.shtml).

Evaluation of agronomic traits

Evaluation of traits was conducted in a field trial in Beijing with three replications in each of 2 years (2012 and 2013). For each replication, 20 grains of each line were evenly planted in 2.0 m rows, spaced 0.3 m apart. All the translocation lines were evaluated for a number of agronomic traits, including fertile tiller number, grain number per spike, spikelet number per spike, kernel number per spikelet, and thousand-kernel weight. Traits were measured on ten plants randomly selected from each line of each translocation subtype. Statistical analyses were conducted using the Statistical Analysis System version 9.2 (SAS Institute Inc., Cary, NC, USA), and the t-test was used to test the difference of the agronomic traits between each of the translocation subtypes and two common wheat parents (Gaocheng 8901 and Fukuho).

Results

Distinct translocation types were identified in the progenies of WAT31-13

We previously reported that the 5A-6P reciprocal translocation-addition line WAT31-13 (2n = 44) (M2 generation) harbored a number of superior agronomic traits but displayed genetic instability (Luan et al. 2010). By strict self-pollination by spike bagging for six generations, 43 translocation lines were acquired at the M8 generation. GISH was carried out at each generation to determine the genetic constitutions of individual plants. Plants that lost the A. cristatum chromosome segments were excluded from further investigation. Because plants at the M8 generation were all identical in genetic constitution relative to the M6 and M7 generations, we concluded that these 43 M8 lines were genetically stable. GISH indicated that there were three types of translocation chromosomes in these various lines (Table 1), with each line harboring a single or combinations of three translocation types (Table 2). As shown in Table 1, the small translocation chromosome (TrS) contained a small segment of A. cristatum chromosome (shown in red) of approximately 60 % of the distal portion of one chromosome arm, whereas the remainder of the chromosome including the centromere was from wheat (shown in blue). In contrast, the large translocation chromosome (TrL) contained a small segment of wheat chromosome (shown in blue) at the end of one arm with the rest of the chromosome from A. cristatum (shown in red). The arm translocation chromosome (TrA) seemed to include only the translocated arm found in the TrL type.

Table 1 Three types of translocation chromosomes identified from 43 wheat-A. cristatum translocation lines
Full size table
Table 2 Seven subtypes of translocations categorized from 43 wheat-A. cristatum translocation lines
Full size table

According to the translocation chromosomes they harbored, these 43 translocation lines were categorized into seven translocation subtypes (Table 2; Fig. 1). In Subtype I, there were 42 wheat chromosomes as well as one pair of TrSs and one pair of TrLs (Fig. 1a; Table 2). Subtype II contained 42 wheat chromosomes, a pair of TrSs and a pair of TrAs (Fig. 1b; Table 2). Subtype III contained 42 wheat chromosomes and a pair of TrLs (Fig. 1c; Table 2). Subtype IV contained 42 wheat chromosomes and a pair of TrSs (Fig. 1d; Table 2). In Subtype V, there were 40 wheat chromosomes and a pair of TrSs (Fig. 1e; Table 2). Both Subtype VI and VII possessed 42 chromosomes in total, consisting of all wheat chromosomes except one chromosome 5A but add one TrS and TrL, respectively (Fig. 1f, g; Table 2). In Subtype II, we hypothesized that TrA might have derived from TrL by misdivision at the centromere. If this was the case, the centromere of TrA should come from A. cristatum rather than wheat. Indeed, the wheat centromere-specific probe CRW successfully stained all chromosome centromeres except for those of the TrAs, but the A. cristatum centromere-specific probe pAcCR1 stained only the TrA centromeres (Fig. 2). These results supported our assumption that TrA was a telocentric derivative of TrL.

Fig. 1
figure 1

GISH discrimination of seven subtypes of wheat-A. cristatum translocation lines. a Subtype I contained 22 pairs of chromosomes including 20 pairs of wheat chromosomes, one pair of TrSs and one pair of TrLs. b Subtype II consisted of 20 pairs of wheat chromosomes, one pair of TrSs and one pair of TrAs. c, d Both Subtype III and Subtype IV consisted of 44 chromosomes including 42 wheat chromosomes, but including one pair of TrLs and one pair of TrSs, respectively. e Subtype V possessed 42 chromosomes consisting of 40 wheat chromosomes and one pair of TrSs. f, g Both Subtype VI and Subtype VII possessed 42 chromosomes in total, but harbored one TrS and one TrL, respectively. Arrows indicate wheat-A. cristatum translocation chromosomes. Chromosomes or chromosomal segments painted in blu e and red belong to wheat and A. cristatum, respectively (color figure online)

Full size image
Fig. 2
figure 2

Identification of the centromere of TrA in Subtype II using the CRW and pAcCR1 probes. a Green signals appeared on all 21 pairs of chromosomes except one pair of TrAs when the probe CRW (centromeric retrotransposon of wheat) was used; b Green signals are present on one pair of TrAs with the probe pAcCR1 (centromeric retrotransposon of A. cristatum). Chromosomes in blue and red are wheat and A. cristatum chromosomes, respectively. Arrows indicate the TrA chromosomes in Subtype II (color figure online)

Full size image

Translocation occurred between the wheat chromosome 5A and A. cristatum chromosome 6P

To further determine the identity of the wheat chromosomes involved in the translocation, dual-color FISH was performed on each of the seven translocation subtypes using pAs1 and pHvG39 probes, which were used to distinguish all 21 pairs of wheat chromosomes. As shown in Fig. 3, A. cristatum 6P chromosome segments were translocated onto wheat chromosome 5A in all seven subtypes. In the TrSs that were present in Subtype I, II, IV, V, and VI, A. cristatum 6P chromosome segments were translocated onto wheat chromosome arm 5AL, replacing 60 % of the distal portion of 5AL. In the TrLs that were present in Subtype I, III, and VII, A. cristatum 6P chromosome segments replaced wheat chromosome 5AS and a 40 % portion of 5AL proximal to the centromere.

Fig. 3
figure 3

FISH identification of seven subtypes of wheat-A. cristatum translocation lines using pHvG39 and pAs1 as probes. Images on the right of each panel show FISH results, and those on the left showed corresponding GISH patterns. Chromosomes in blue and red are wheat and A. cristatum chromosomes in the GISH patterns, respectively; whereas pHvG39 and pAs1 signals were pseudo-colored as green and red in the FISH patterns, respectively. a–g, FISH patterns of Subtype I (a), Subtype II (b), Subtype III (c), Subtype IV (d), Subtype V (e), Subtype VI (f), and Subtype VII (g). Note that A. cristatum 6P chromosome segments were translocated to wheat chromosome 5A in all seven subtypes. Arrows indicate wheat-A. cristatum translocation chromosomes (color figure online)

Full size image

Identification of the breakpoint in the Subtype V translocation line

To pinpoint the location of the breakpoint in the TrS and TrL translocation chromosomes, Subtype V was the only subtype that could be used, since it lacked an intact wheat chromosome 5A. Twenty-five wheat SSR markers physically mapped on wheat chromosome 5A were chosen to analyze the breakpoint in Subtype V. CS nulli-tetrasomic lines (N5AT5B and N5AT5D), CS ditelosomic lines (Dt5AL and Dt5AS), and CS 5AL deletion lines were used to confirm the physical locations of these markers. As shown in Fig. 4a, seven and 18 markers were located to the wheat chromosome arms 5AS and 5AL, respectively. Among them, five markers (Xgwm186, XbarcM158, XbarcM135, Xbarc155, and Xbarc186) were located in bin C-5AL12-0.35, four markers (Xbarc40, XksuM56, Xbarc1, and XksuM5) in bin 5AL12-0.35-0.57, three markers (Xcfa2163, Xgwm617, and Xgwm666) in bin 5AL10-0.57-0.78, three markers (Xwmc110, Xcfa2155, and Xgpw1086) in bin 5AL17-0.78-0.87, and three markers (Xgpw2120, Xgpw2136, and Xgpw2172) in bin 5AL23-0.87-1.00. PCR results showed that only SSR markers in the 5AS and bin C-5AL12-0.35 were amplified in Subtype V, but SSR markers from other bins were not. Xbarc40 and XksuM56 were present in Subtype V, but Xbarc1 and XksuM5 were absent (Fig. 5a), indicating that the translocation breakpoint occurred on wheat chromosome arm 5AL and the breakpoint was located in bin 5AL12-0.35-0.57. All 25 wheat SSR markers were present in the other six subtypes, suggesting that at least one intact 5A chromosome or combinations of one intact 5A chromosome was present in these subtypes.

Fig. 4
figure 4

Physical map of wheat chromosome 5A and A. cristatum chromosome 6P. The map on the left is wheat chromosome 5A consisting of nine deletion bins, and wheat SSR markers are shown on the corresponding regions. The map on the right is A. cristatum chromosome 6P, including 6PS, C-6PL-0.32, 6PL-0.32-0.69, and 6PL-0.69-1.00 chromosomal segments, and A. cristatum-specific markers are shown on the corresponding regions. Fraction breakpoints on chromosomes 5A and 6P are indicated as dashed lines; The location of gene Q is indicated by the arrow according to the previous reports (Galiba et al. 1995; Sutka et al. 1999)

Full size image
Fig. 5
figure 5

PCR amplification patterns of wheat SSR markers and A. cristatum-specific markers. a Wheat SSR markers Ksum5 and Ksum56 were shown in bin 5AL12-0.35-0.57 by CS nulli-tetrasomic lines, CS ditelosomic lines, and CS 5AL deletion lines. b A. cristatum-specific markers Agc3413, Agc12567, and Agc8937 located in different bins of A. cristatum 6P chromosome were present or absent in some subtypes. Arrows indicate specific bands from wheat (a) and A. cristatum (b), respectively. Lanes in a: 1 CS, 2 N5AT5B, 3 N5AT5D, 4 DT5AS, 5 DT5AL, 6 5AL23-0.87-1.00, 7 5AL17-0.78-0.87, 8 5AL10-0.57-0.78, 9 5AL12-0.35-0.57, 10 C-5AL12-0.35, 11 Subtype V. Lanes in b: 1 Z559, 2 4844-12, 3 Gaocheng 8901, 4 Fukuho, 5 Subtype I, 6 Subtype II, 7 Subtype III, 8 Subtype IV, 9 Subtype V, I0 Subtype VI, 11 Subtype VII

Full size image

To determine which arm of A. cristatum chromosome 6P was present in these translocation lines, seven EST makers specific for A. cristatum chromosome arm 6PL or 6PS were used. Three markers (For3-G02, For5-E08 and For22-B10) on chromosome 6PS were absent in subtype V, whereas four markers (For15-D06, For8-G11, For14-B02, and For22-E03) on the A. cristatum chromosome 6PL were present in Subtype V (Figs. 4b, 5b). Besides, another nine A. cristatum-specific EST markers were examined. Three markers (Agc3413, Agc7155, and Agc9322) on A. cristatum chromosome arm 6PS, two markers (Agc6900 and Agc34162) in bin C-6PL-0.32, and Agc12567 in bin 6PL-0.32-0.69 were absent in Subtype V, whereas Agc8937 in bin 6PL-0.32-0.69 and two markers (Agc4543 and Agc24535) in the bin 6PL-0.69-1.00 were present in Subtype V (Fig. 5b). The results indicated that the translocation breakpoint occurred on A. cristatum chromosome arm 6PL and the breakpoint was located in bin 6PL-0.32-0.69. A. cristatum chromosome segments including partial 6PL-0.32-0.69 and whole 6PL-0.69-1.00 were translocated onto wheat chromosome arm 5AL, so the constitution of the translocation chromosome in Subtype V could be described as T5AS∙5AL-6PL (Table 2). Similarly, TrS and TrL could also be described as T5AS∙5AL-6PL and T5AL-6PL∙6PS, respectively, while TrA could be designated as T5AL-6PL (Table 2).

Evaluation of agronomic traits in all seven subtypes

Each of these 43 translocation lines was evaluated for five agronomic traits, including fertile tiller number, grain number per spike, spikelet number per spike, kernel number per spikelet, and thousand-kernel weight. Trait observations over 2 years were very similar showing that year-by-year environmental effects were insignificant. We also observed the lines from the same translocation subtype always displayed similar phenotypes, indicating that the genetic constitution was the major factor controlling these phenotypes. Phenotypes for each subtype were calculated using phenotypic data from all lines within the particular subtypes (Table 3; Fig. 6).

Table 3 The agronomic traits of seven translocation subtypes and their parents in year 2012 and 2013
Full size table

Among all the seven subtypes, Subtype V was the only subtype exhibiting inferior agronomic traits including lower fertile tiller number, grain number, spikelet number per spike, kernel number per spikelet, and thousand-kernel weight. Subtype V displayed speltoid spike morphology, and occurred at a very low frequency (only one line). This was probably because Subtype V was the only translocation subtype deficient in a large portion of wheat chromosome arm 5AL. Our results suggested that the distal end of wheat chromosome arm 5AL was essential for normal plant growth and overall fitness and cannot be fully compensated by the A. cristatum chromosome arm 6PL. We compared the agronomic traits among the other six subtypes and their contrast parents. Grain number per spike in Subtype I, VI, and VII were higher than those of two common wheat parents (Gaocheng 8901 and Fukuho), but lower than that of A. cristatum 6P disomic addition line 4844-12. Kernel number per spikelet in all six subtypes were similar to those of two common wheat parents (Gaocheng 8901 and Fukuho), but significantly lower than that of A. cristatum 6P disomic addition line 4844-12. This was also the case for spikelet number per spike. For thousand-kernel weight, no significant difference was observed among the six subtypes as well as two common wheat parents and A. cristatum 6P disomic addition line 4844-12.

Fig. 6
figure 6

Plant morphology of seven subtypes and their parents. Plant morphologies of Subtype I (a), II (b), III (c), IV (d), V (e), VI (f), VII (g), wheat-A. cristatum addition line disomic 4844-12 (h), Gaocheng 8901 (i) and Fukuho (j) were shown

Full size image

Among all traits analyzed, fertile tiller number showed the highest variance. The translocation lines with TrL (Subtype I, III, and VII produced multiple fertile tiller number, which was much higher than those of their two common wheat parents. However, translocation lines without TrL but with TrS (Subtype II, IV, V and VI) displayed reduced fertile tiller number. According to the FISH patterns, TrL included A. cristatum chromosome arm 6PS, chromosome segment 6PL-0.32, and partial 6PL-0.32-0.69, whereas TrS included A. cristatum chromosome segments 6PL-0.69-1.00 and partial 6PL-0.32-0.69. These results indicated that there were positive and negative regulators of fertile tiller number on A. cristatum chromosome 6PS and 6PL, respectively. But when the positive and negative regulators of the fertile tiller number both existed on the same chromosome, as in the case of the A. cristatum 6P disomic addition line 4844-12, the fertile tiller number was not significantly different from those of two common wheat parents, probably because positive and negative effects regulators neutralized each other.

Discussion

Stabilization of the genetic constitution of translocation lines can be achieved spontaneously

Alien genetic resources are important for improving agronomic traits in wheat. Ionizing radiation treatment of alien addition lines, substitution lines, and translocation lines carrying desirable genes is one method to induce chromosome translocation (Chen et al. 1996; Friebe et al. 1996b; Sears and Gustafson 1993; Zhang et al. 2012). In our study, all the translocation lines were identified from the progenies of WAT31-13, which came from an irradiated hybrid. The translocation lines were classified into seven subtypes (I–VII), containing three translocation chromosome types (TrS, TrL, and TrA). TrS and TrL, but not TrA, were present in WAT31-13, indicating that TrA was a new chromosome type which occurred spontaneously. Cytobiological evidences indicated that TrA was derived from TrL as a consequence of a chromosome breakage at the centromere causing loss of the chromosome arm 6PS but retention of the A. cristatum centromere. Such chromosomal breakage happened at a high frequency, since the number of Subtype II individuals (30 lines) was much higher than that of Subtype I (4 lines). The exact biological mechanism of such high frequency of chromosome breakage is currently unknown, but TrA was probably more genetically stable than TrL. Similar spontaneous chromosomal rearrangements have also been reported in wheat-Haynaldia villosa and wheat-Leymus racemosus translocation lines (Cao et al. 2009; Wang et al. 2010).

Non-homoeologous translocations between chromosome 5A and 4A in hexaploid wheat are well known (Devos et al. 1995), and translocations involving 4L/5L also exist in several other species within the tribe Triticeae (King et al. 1994). In our study, all translocations occurred between wheat chromosome arm 5AL and A. cristatum chromosome arm 6PL. Among all the translocation subtypes, Subtype V was the only translocation subtype that exhibited inferior agronomic traits, and also occurred at a very low frequency. This was probably caused by a low overall fitness of the resulting translocation line due to incomplete compensation of the missing segments of chromosome arm 5AL. Indeed, chromosome arm 5AL contains important genes such as gene Q controlling free-threshing and square spike morphology, Vrn-A1 determining winter/spring growth-habit, and Fr1 conferring frost-resistance (Galiba et al. 1995; Sarma et al. 1998; Sutka et al. 1999). Gene Q was reported to locate in bin 5AL23-0.87-1.00 of chromosome arm 5AL, and its location is indicated on the physical map of the wheat chromosome arm 5A (Fig. 4a). Consistent with their lack of gene Q, Subtype V plants displayed speltoid spike morphology, as well as lower seed-set and thousand-kernel weight. Similar phenotypes were observed in the progenies of the cultivar Biscay, which lacks chromosomes 5A carrying Q gene (Forster et al. 2013). Therefore, our results indicated that the distal end of wheat chromosome arm 5AL is essential for normal plant growth and overall fitness, and could not be fully compensated by the A. cristatum chromosome arm 6PL.

The probe pAcCR1 is sufficient to distinguish the A. cristatum and wheat centromere

The plant centromeres are mainly composed of various types of repetitive DNA elements, including transposons, retrotransposons, and telomere-like repeats, most of which are species- or genome-specific (Galasso et al. 1995; Iwabuchi et al. 1991). Thus, centromeric DNA is often a perfect target to distinguish chromosomes of different species (Jiang et al. 2003). A number of repetitive DNA elements were found in wheat centromeres (Cheng and Murata 2003; Kishii et al. 2001; Ito et al. 2004; Zhang et al. 2004), upon which the wheat centromere-specific probe CRW was developed (Liu et al. 2008). In this study, we developed the probe pAcCR1 by a combination of microdissection and DOP-PCR methods. The probe pAcCR1 specifically hybridizes with A. cristatum centromeres, and could not hybridize with wheat centromeres. The probe pAcCR1 could not hybridize with centromeres from several other genomes including genome H, E, R, St, and U (unpublished data).

The translocation lines developed in this study could be used to mechanistically explore the A. cristatum chromosome segments controlling fertile tiller number

Tillering is a key component of yield for most cereals such as wheat, rice, and barley (Sakamoto and Matsuoka 2004; Sreenivasulu and Schnurbusch 2012). While the molecular mechanism of tillering has been extensively studied in dicots and a few monocots (Aguilar-Martinez et al. 2007; Kebrom et al. 2013; Li et al. 2003; Minakuchi et al. 2010; Schmitz et al. 2002; Wang and Li 2008), it has not been fully explained in wheat, probably due to the lack of genetic materials suitable for dissecting the genetic architecture of tillering. A number of quantitative trait loci (QTLs) for tiller number have been identified (Huang et al. 2004; Kato et al. 2000; Kim et al. 1993; Li et al. 2002; Narasimhamoorthy et al. 2006; Naruoka et al. 2011; Snape et al. 1985), but the underlying genes have not been cloned. In those studies, a large number of QTLs with minor effects were identified, indicating that tillering is a complex trait coordinately controlled by many loci.

To successfully clone QTLs, the parents used to construct the cloning population have to be carefully chosen, since QTLs with large effects can only be revealed when alleles from the parents have dramatic effects on tillering. Few cases have been reported where large-effect QTLs controlling tillering have been fine-mapped in wheat, the most prominent example was the tiller inhibition gene (tin1) detected in the uniculm wheat mutant Line 492 and located on the wheat chromosome arm 1AS (Spielmeyer and Richards 2004). Other examples included tin2 and tin3 (Kuraparthy et al. 2007; Peng et al. 1998). The ftin (fertile tiller inhibition) gene was located on wheat chromosome arm 1AS in Pubing 3558 in our previous report (Zhang et al. 2013). The translocation lines reported here have dramatic differences in fertile tiller number, so there should exist gene(s) controlling fertile tiller number. In this study, we determined that the A. cristatum chromosome arm 6PS and 6PL have positive and negative effects on fertile tiller number, respectively. However, the underlying genes controlling fertile tiller number cannot be cloned from these materials using a traditional map-based cloning approach. To overcome this hurdle, the A. cristatum 6P chromosome segments can first be reduced to smaller segments and transferred into a known elite cultivated wheat background, and then the strategy employed in the cloning of Pm21 can be used (Cao et al. 2011).

Some translocation lines showed potential applications for breeding high-yielding wheat

Wheat cultivars can be classified into large-spike and multi-spike types. Compared to large-spike types, multi-spike type cultivars are considered to be more stable in agronomic performance, and it is easier to achieve higher yields per unit field area (Deng et al. 2011). This effect becomes more pronounced when wheat plants are confronted with environmental stresses such as drought and salinity (Reynolds et al. 2007; Tian et al. 2006a, b). In this study, all translocation subtypes except Subtype V showed higher grain number per spike, thus they are potentially useful for breeding wheat lines with large spikes. However, what attracted our attention more were the subtypes with high numbers of fertile tiller, such as Subtype I, III, and VII, which also showed potential applications for breeding of multi-spike types. Although these subtypes showed potential applications for breeding high-yielding wheat, it would be a challenge to introduce the desirable genes conferring multiple fertile tiller number from A. cristatum chromosome arm 6PS into common wheat without the introduced segments also conferring deleterious effects. To overcome this problem, chromosome translocation induced by ionizing radiation or Ph1-deficient genetic stocks can be used.

The Ph1 system is advantageous in that it can significantly promote the frequency of homoeologous chromosome pairing and recombination, thereby producing genetically compensating translocations (Qi et al. 2008; Niu et al. 2011). However, homoeologous recombination is affected by the structurally rearranged segments of alien chromosomes, the genetic relationship between wheat and its related species as well as the genetic distance between the target gene and the centromere (Qi et al. 2007; Monte et al. 1993; Nasuda et al. 1998). In our previous research, wheat-A. cristatum 6P disomic addition lines displayed obvious chromosome rearrangements (Han et al. 2014). Therefore, ionizing radiation might be the preferred choice in our case. Ionizing radiation treatment can cause alien chromosome breakage at different positions, and then the alien chromosome will be transferred onto different regions of wheat chromosomes, as exemplified in Chen et al. (2013). Therefore, it is possible that the desirable genes conferring multiple fertile tiller number from A. cristatum chromosome arm 6PS would be transferred onto other chromosomes rather than wheat chromosome 5A during further cycles of ionizing radiation.

In conclusion, by studying the chromosomal constitution, behavior, and agronomic traits of wheat-A. cristatum translocation lines, we pinpointed the chromosomal segments of A. cristatum 6P positively and negatively regulating fertile tiller number in wheat, respectively. Our work not only laid the foundation for further research on wheat tillering, but also provided the starting materials for high-yield wheat breeding.

Author contribution statement

Li LH conceived the project. Lu YQ and Ye XL analyzed the data and wrote the manuscript. Ye XL, Han HM, and Chen GY performed cytological experiments. Liu WH created the translocation lines. Zhang JP and Gao AN contributed to the development of molecular markers. Yang XM and Li XQ performed the phenotyping.