Targeted Introgression of a Wheat Stem Rust Resistance Gene by DNA Marker-Assisted Chromosome Engineering

Abstract

Chromosome engineering is a useful strategy for transfer of alien genes from wild relatives into modern crops. However, this strategy has not been extensively used for alien gene introgression in most crops due to low efficiency of conventional cytogenetic techniques. Here, we report an improved scheme of chromosome engineering for efficient elimination of a large amount of goatgrass (Aegilops speltoides) chromatin surrounding Sr39, a gene that provides resistance to multiple stem rust races, including Ug99 (TTKSK) in wheat. The wheat ph1b mutation, which promotes meiotic pairing between homoeologous chromosomes, was employed to induce recombination between wheat chromosome 2B and goatgrass 2S chromatin using a backcross scheme favorable for inducing and detecting the homoeologous recombinants with small goatgrass chromosome segments. Forty recombinants with Sr39 with reduced surrounding goatgrass chromatin were quickly identified from 1048 backcross progenies through disease screening and molecular marker analysis. Four of the recombinants carrying Sr39 with a minimal amount of goatgrass chromatin (2.87–9.15% of the translocated chromosomes) were verified using genomic in situ hybridization. Approximately 97% of the goatgrass chromatin was eliminated in one of the recombinants, in which a tiny goatgrass chromosome segment containing Sr39 was retained in the wheat genome. Localization of the goatgrass chromatin in the recombinants led to rapid development of three molecular markers tightly linked to Sr39. The new wheat lines and markers provide useful resources for the ongoing global effort to combat Ug99. This study has demonstrated great potential of chromosome engineering in genome manipulation for plant improvement.

MODERN genetic improvement has increased crop productivity worldwide, but it also erodes genetic variability of crops (Allard 1996; Hoisington et al. 1999). Narrowed genetic bases make modern crops fragile to global climate change and disease and insect epidemics (Tanksley and McCouch 1997; Chan 2010). Broadening the genetic variability will make crop production more sustainable under various biotic and abiotic stresses. One approach to increase genetic diversity is to incorporate genes from the crop's wild relatives using chromosome engineering (Brar and Khush 2005; Jellen and Leggett 2005; Singh 2005; Jauhar et al. 2009). However, progress of chromosome engineering has been limited due to difficulties in generating, recovering, and identifying meiotic recombinant chromosomes. Recent advances in genomics and high-throughput genotyping technologies have enhanced the competence of chromosome engineering for crop improvement (Ceoloni et al. 2005).

Wheat (Triticum aestivum L., 2n = 6x = 42, AABBDD) is one of the major food crops and its production is currently threatened by a new stem rust (Puccinia graminis Pers.:Pers. f. sp. tritici Eriks. & E. Henn.) race, Ug99 (or TTKSK), identified in Uganda in 1999 (Pretorius et al. 2000; Wanyera et al. 2006; Singh et al. 2008; Jin et al. 2008; Jin et al. 2009). Although Ug99 has broad virulence to currently deployed Sr genes, it is avirulent to most of the Sr genes derived from relatives of wheat (Singh et al. 2006; Jin et al. 2007). However, large amounts of alien chromatin surround most of the alien Sr genes transferred to wheat (Xu et al. 2008). The genes other than the targeted gene on the alien chromatin usually cause linkage drag, a deleterious effect on yield and quality (The et al. 1988; Lukaszewski 2000; Labuschagne et al. 2002). Thus, additional chromosome engineering is needed to minimize alien chromatin before the alien Sr genes are deployed in wheat cultivars.

The alien Sr gene Sr39 is effective against Ug99 and was transferred from goatgrass (Aegilops speltoides Tausch, 2n = 2x = 14, SS) to wheat cultivar Marquis (Kerber and Dyck 1990). The original wheat line, RL5711, carried both Sr39 and a gene for leaf rust resistance (Lr35) on a translocated chromosome between Ae. speltoides chromosome 2S and wheat chromosome 2B designated as T2B/2S#2 (Friebe et al. 1996). Later, Sr39 and Lr35 were transferred from RL5711 into wheat line RL6082 through six backcrosses using wheat cultivar Thatcher as the recurrent parent (Seyfarth et al. 1999). The translocated chromosome T2B/2S#2 consisted of the 2S long arm, a large portion (∼85%) of the 2S short arm, and a small terminal 2B segment (Yu et al. 2010). To make Sr39 usable to fight Ug99, the excess Ae. speltoides chromatin surrounding Sr39 needs to be eliminated.

The common procedure for reducing alien chromatin in wheat is to induce meiotic recombination between the alien chromatin and its homoeologous (i.e., partially homologous) region of wheat chromosome. Regular pairing between homologous chromosomes is ensured by the major gene Ph1 on chromosome 5B in wheat (Riley and Chapman 1958; Gill et al. 1993; Martinez-Perez et al. 2001; Griffiths et al. 2006; Sidhu et al. 2008). Absence of Ph1 due to nullisomy for chromosome 5B or mutation (e.g., ph1b and ph1c) enhances meiotic pairing and recombination between homoeologous chromosomes (Sears 1954, 1966, 1977; Joppa and Williams 1988). The Ph1-deficient genetic stocks such as substitutions of chromosome 5B by 5D and ph1b mutant have been successfully employed for inducing meiotic recombination between wheat chromosomes and their alien homeologues (see review by Qi et al. 2007; Faris et al. 2008; Marais et al. 2010). By using the ph1b mutant, Mago et al. (2009) developed four wheat lines carrying Sr39 on shortened Ae. speltoides chromosome segments. On the basis of their characterization of the two best lines, we estimate that the amount of Ae. speltoides chromatin has only been reduced by 20–25%.

In addition to the wheat lines reported by Mago et al. (2009), R. E. Knox developed a set of breeding lines carrying Lr35/Sr39 from the original translocation line (Knox et al. 2000; Yu et al. 2010). One line had a slightly shortened Ae. speltoides chromosome segment, but the translocation chromosome still retained ∼80–85% of the Ae. speltoides chromatin (Yu et al. 2010). All other lines contained similar amounts of Ae. speltoides chromatin as the original translocation lines. The Ae. speltoides chromatin present in the lines reported by Mago et al. (2009) and Yu et al. (2010) will still be unacceptable to many wheat breeders. Therefore, the objectives of this study were to develop an efficient chromosome engineering procedure to minimize Ae. speltoides chromatin surrounding Sr39 in the wheat–Ae. speltoides translocation line.

MATERIALS AND METHODS

Plant materials:

Wheat line RL6082 containing Sr39 on the translocated chromosome T2B/2S#2, which was transferred from original wheat–Ae. speltoides translocation line RL5711 (Kerber and Dyck 1990) to Thatcher by P. L. Dyck in Winnipeg, Canada (Seyfarth et al. 1999), was used for chromosome engineering. Wheat ‘Chinese Spring’ (CS) and CS ph1b mutant were employed to induce meiotic recombination between the 2B and 2S homoeologous regions. Thatcher, Ae. speltoides accession RL5344 (donor of Sr39), CS N2A-T2B (nullisomic for 2A and tetrasomic for 2B), CS N2B-T2A (nullisomic for 2B and tetrasomic for 2A), and CS N2D-T2A (nullisomic for 2D and tetrasomic for 2A) were used as controls for stem rust evaluation, fluorescent genomic in situ hybridization (FGISH), and molecular marker analysis.

Chromosome manipulation:

Wheat plants homozygous for ph1b and monosomic for both wheat chromosome 2B and translocated chromosome T2B/2S#2 were developed to induce meiotic recombination between 2B and T2B/2S#2 (Figure 1). These plants were created by crossing CS ph1b mutant (ph1bph1b) to RL6082 and backcrossing the F₁ plants to the ph1b mutant. The BC₁F₁ plants were tested for reactions to stem rust. Resistant BC₁F₁ plants were then analyzed with the molecular markers PSR128, PSR574, and AWJL3 to select individuals homozygous for ph1b (Roberts et al. 1999). The resistant ph1bph1b plants, which were monosomic for both 2B and T2B/2S#2, were backcrossed to CS to efficiently recover the gametes containing a recombinant chromosome of 2B and T2B/2S#2. The BC₂F₁ plants were tested with stem rust and genotyped with a molecular marker. The BC₂F₂ progeny of the plants with a shortened Ae. speltoides chromosome segment were screened with stem rust and FGISH to confirm the lines carried Sr39 on shortened Ae. speltoides chromosome segments. The sizes of the Ae. speltoides chromosome segments in the selected wheat lines were calculated as the average percentage of length of Ae. speltoides chromosome segment/total length of the translocation chromosome, which was measured in 18–20 cells with good-quality mitotic metaphases. To verify elimination of the ph1b allele in the selected wheat lines, 16 BC₂F₃ seeds from each of the selected BC₂F₂ plants were analyzed with markers linked to ph1b. The FGISH was performed as described by Yu et al. (2010). The methodologies for stem rust testing and molecular marker development, analysis, and validation are described below.

Figure 1.—

Hybridization procedure to reduce the size of the Aegilops speltoides chromosome segment containing Sr39 in the wheat–Ae. speltoides chromosome 2B/2S translocation line RL6082.

Stem rust resistance evaluation:

The stem rust race TMLK, which can differentiate Sr39 from the Sr genes in Thatcher and CS (Yu et al. 2010), was used for the inoculation of the BC₁F₁, BC₂F₁, and BC₂F₂ populations. The selected wheat lines with shortened Ae. speltoides chromosome segment were tested with six additional races, THTS, TPMK, RTQQ, RHTS, QFMQ, and QFCQ, along with TMLK and Ug99. The seedlings were grown in the greenhouse at 20–23° with a 16/8 hr (day/night) photoperiod. Seven-day-old seedlings were inoculated as described by Williams et al. (1992). Inoculated seedlings were transferred to either a greenhouse or growth chamber maintained at 20–23°. Infection types were scored on the 13th or 14th day after inoculation using the scale described by Stakman (1962), where 0 = immune, ; = necrotic flecks, 1 = small necrotic pustules, 2 = small to medium-sized chlorotic pustules with green island, 3 = medium-sized chlorotic pustules, and 4 = large pustules without chlorosis. Plants with infection type ≥3 were considered susceptible, and plants with an infection type <3 were considered resistant.

Molecular marker analysis:

Yu et al. (2010) showed that microsatellite marker Xgwm319 generated a 173-bp fragment from chromosome 2B and a 162-bp fragment from Ae. speltoides 2S chromatin. This marker was mapped to the centromeric region of chromosome 2B and 2S (Somers et al. 2004; Yu et al. 2010) and was used to select new 2B/2S recombinants. DNA samples of the BC₂F₁ population were extracted as described in the following steps. Young leaf tissue was collected in Qiagen 1.2 ml 96-well plates with addition of one 3-mm tungsten carbide bead (Qiagen, Valencia, CA) to each well. After loaded plates were frozen in liquid nitrogen, the plates were shaken on a MM300 shaker (Retsch) at 30 Hz for 60–90 sec two to three times until the leaf tissue was ground into a fine powder. Four hundred microliters of preheated DNA extraction buffer (500 mm NaCl, 100 mm Tris-HCl, pH 8.0, 50 mm EDTA, 0.84% SDS, and 0.38 g/100 ml sodium bisulfate freshly added) was added to each well. Plates were briefly (5–10 sec) shaken, and the samples were incubated at 65° for 30–40 min. Chloroform (300–400 μl) was added to each well. After vigorously mixing, the samples were centrifuged at 4000–4500 rpm for 15 min and ∼300 μl of supernatant from each well was transferred into new 96-well plates. About 0.8–0.9 volume of isopropanol was added to each well and plates were vigorously hand shaken. The plate was centrifuged at 4000–4500 rpm for 15 min and the supernatant was discarded. Six hundred microliters of 70% ethanol was added to each well and the plate was centrifuged at 4000–4500 rpm for 15 min. The supernatant was discarded and the pellet was air dried. The DNA was dissolved in 300 μl TE buffer (10 mm Tris, pH 8.0, 1 mm EDTA).

Marker genotyping was performed as described by Tsilo et al. (2009). Polymerase chain reaction (PCR) was performed at an annealing temperature of 50°. Amplified PCR products were labeled with four different fluorescent dyes (6-FAM, VIC, NED, and PET) and separated by a 16-capillary electrophoresis system ABI 3130xl genetic analyzer (Applied Biosystems, Foster City, CA), and genotype calls were analyzed using GeneMapper software v3.7 (Applied Biosystems). Following testing with marker Xgwm319, seven additional PCR-based markers (Sr39#22r, Sr39#50s, BE500705, Xbarc18, Xbarc183, Xbarc200, and Xwmc025) that detect the Ae. speltoides 2S chromatin carrying Sr39 (Mago et al. 2009) were tested on the selected lines. In addition, STS (sequence-tagged site) markers developed as described below were used to test all BC₂F₁ plants in which Sr39 disassociated from the Ae. speltoides allele of Xgwm319.

Development and validation of new STS markers linked to Sr39:

On the basis of the physical location of Ae. speltoides chromatin detected with FGISH, the wheat EST (expressed sequence tag) sequences from the deletion bin 2BS4-0.75-0.84 (http://wheat.pw.usda.gov/data-bin/graingenes/report.data?class=breakpointinterval;name=2BS4-0.75-0.84;show=locus) were selected to design primers using Primer3Plus (Rozen and Skaletsky 2000; http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) under general settings. These primers were used to screen for polymorphisms among Thatcher, CS, RL6082, a bulk of four resistant BC₂F₂ plants, a bulk of four susceptible BC₂F₂ plants, CS N2A-T2D, CS N2B-T2A, and CS N2D-T2A. Genomic DNA was extracted using the 2× CTAB method (Li and Quiros 2001). PCR were carried out as follows: 95° for 5 min, 95° for 40 sec, 55° for 40 sec, 72° for 40 sec, repeated for 36 cycles, with a final extension at 72° for 10 min. The PCR products were separated on an 8% nondenaturing polyacrylamide gel and stained with 2× GelRed. The gel was scanned using a Typhoon 9410 scanner (GE Healthcare Biosciences, Pittsburgh, NJ). The EST sequences from primers detecting polymorphisms were then used to BLAST against the wheat EST database (http://wheat.pw.usda.gov/GG2/blast.shtml), and the top hit contig sequence (NSFT03P2_Contig11068) was selected to BLAST against the rice (Oryza sativa L.) genomic sequence (http://www.shigen.nig.ac.jp/wheat/komugi/blast/blast.jsp). The top hit clone sequence (tp1b0012I12) from the rice Indica group was selected to design additional primers. The polymorphic STS markers were tested on a BC₂F₂ population inoculated with TMLK. The new STS markers linked to Sr39 on reduced Ae. speltoides chromatin were then validated in a set of 40 durum (T. turgidum L. subsp. durum) and common wheat cultivars/lines from China and the United States.

RESULTS

We used an improved procedure (Figure 1) to develop ph1b-induced homoeologous recombinants. From the cross and backcross of RL6082 with CS ph1b mutant, 93 BC₁F₁ plants (pedigree: CS ph1bph1b*2/RL6082) were generated and tested with TMLK (supporting information, Table S1). The segregation of 53 resistant plants to 40 susceptible plants fit a 1:1 ratio (χ² = 1.82), indicating that stem rust resistance was conditioned by a single gene. The 53 resistant plants were analyzed with markers PSR128, PSR574, and AWJL3 to detect the presence of ph1b. Sixteen resistant plants were identified as homozygotes for ph1b (Figure S1 and Table S1) and were backcrossed to CS to develop 16 families composing a large BC₂F₁ population. Among 1048 BC₂F₁ plants from 12 families, 554 were resistant and 494 were susceptible to TMLK (Table 1 and Table S2). While this segregation was close to a 1:1 ratio, analysis of the populations on a family basis indicated that segregation in only three families fit a 1:1 ratio (Table 1). A heterogeneity χ² test of the families (χ² = 204.6) indicated that the data could not be pooled. These results indicated that significant segregation distortion was present among families.

TABLE 1.

Segregation for resistance to stem rust race TMLK among BC2F1 plants in the 12 families derived from crossing Chinese Spring (CS) to 12 BC1F1 plants having the pedigree CS ph1bph1b*2/RL6082

Family no.	Resistant	Susceptible	χ2 (1:1)	Prob. (1:1)
81-3	49	35	2.33	0.127
81-5	27	63	14.40	<0.001
81-6	51	38	1.90	0.168
81-11	48	40	0.73	0.394
81-20	21	61	19.51	<0.001
81-35	60	29	10.80	0.001
81-38	72	16	35.64	<0.001
81-39	17	71	33.14	<0.001
81-40	66	24	19.6	<0.001
81-42	57	29	9.12	0.003
81-56	67	18	28.24	<0.001
81-63	19	70	29.22	<0.001
Total	554	494	3.44	0.064

Prob., probability.

The 1048 BC₂F₁ plants tested with TMLK were screened for dissociation between marker Xgwm319 and Sr39. Xgwm319 amplified 193-bp and 191-bp fragments from chromosome 2B of CS and Thatcher, respectively, and a 182-bp fragment from RL5344 and RL6082 (fragment sizes include a 20-bp M13 primer tail) (Figure 2). The results showed that only 40 of 532 resistant BC₂F₁ plants carried the Xgwm319 allele from CS, while the remaining 492 plants had the RL6082 allele (Table 2 and Table S3). Of 476 susceptible plants, 265 carried the RL6082 allele and the remaining 211 plants had the CS allele. The overall dissociation frequency of Sr39 from Xgwm319 was 30.3% (305/1008) (Table 2); however, the dissociation frequency was 7.5% (40/532) and 44.3% (211/476) among resistant and susceptible plants, respectively. The difference in dissociation frequency among resistant and susceptible plants supports the conclusion that significant segregation distortion occurred in the populations.

Figure 2.—

Electropherograms showing the polymerase chain reaction (PCR) products of microsatellite marker Xgwm319 in Thatcher (Tc), Chinese Spring (CS), three CS nullisomic-tetrasomic (N-T) lines, RL6082, and Aegilops speltoides RL5344 (Sr39 source). The three CS N-T lines, including N2A-T2D (nullisomic 2A and tetrasomic for 2D), N2B-T2A, and N2D-T2A, were used as checks for the PCR amplification on chromosome 2B. Fragment sizes include a 20-bp M13 primer tail. The peak represents the PCR products, whereas the horizontal and vertical scales represent fragment sizes in base pairs and fluorescent signal intensity, respectively.

TABLE 2.

Stem rust test and marker analysis for BC2F1 plants having the pedigree Chinese Spring (CS)//CS ph1bph1b*2/ RL6082

Reaction to TMLK		No. of plants with marker (Xgwm319) allele from
Type	No. of plants	RL6082	CS	Missing data
R	554	492	40	22
S	494	265	211	18
Total	1,048	755	253	40

R, resistant; S, susceptible.

The original translocation line RL6082 and the BC₂F₂ plants derived from 11 of the resistant BC₂F₁ plants exhibiting dissociation of Sr39 from Xgwm319 were screened with stem rust and FGISH to confirm the BC₂F₁ plants with Sr39 on shortened Ae. speltoides chromosome segments. One BC₂F₁ plant from family 81-3 was identified to carry a 2B/2S translocation chromosome, but with only a slightly reduced Ae. speltoides chromatin in the short arm (Figure S2A). Six BC₂F₁ plants carried a telocentric chromosome, with the entire 2S long arm (2SL) being absent (Figure S2, B–G). Four BC₂F₁ plants were identified that carried a 2B/2S translocation chromosome with a very small amount of Ae. speltoides chromatin. From each of these four plants, new wheat lines (RWG1, RWG2, RWG3, and RWG4) with homozygous short translocations were selected from the BC₂F₂ progeny by rust testing (Table S4) followed by FGISH analysis (Figures 3 and 4A). The Ae. speltoides chromatin carrying Sr39 in RWG1, RWG2, RWG3, RWG4, and RL6082, in average comprised 2.87%, 4.72%, 3.60%, 9.15%, and 95.03% of the interchanged chromosome, respectively (Table S5), indicating that ∼90–97% of the Ae. speltoides chromatin surrounding Sr39 had been eliminated. Except for the Ae. speltoides chromatin carrying Sr39, other small hybridization signals at the telomeres of some chromosomes were also frequently detected in RL6082 and the four new wheat lines (Figures 3 and 4A). We were unable to determine the identity of these signals, which could be caused by the Ae. speltoides chromatin or the highly repetitive sequences shared by the wheat and Ae. speltoides genomes.

Figure 3.—

Fluorescent genomic in situ hybridization (FGISH) results for RL6082 (A), four new wheat lines RWG1 (B), RWG2 (C), RWG3 (D), and RWG4 (E) with Sr39, and Chinese Spring (F). The FGISH result for RWG1 (B) was obtained from a BC₂F₂ plant that was heterozygous for the translocation chromosome. The Ae. speltoides chromatin (green) is indicated by arrows. The green signals on the telomeres of the chromosomes could be caused by the Ae. speltoides chromatin or the highly repetitive sequences shared by wheat and Ae. speltoides. Bar, 10 μm.

Figure 4.—

Molecular, phenotypic, and cytogenetic characteristics of four new wheat lines RWG1, RWG2, RWG3, and RWG4 and their parents RL6082, Chinese Spring (CS), and Thatcher (Tc). (A) Fluorescent genomic in situ hybridization results and stem rust reactions to TMLK. The Ae. speltoides chromatin (green signal) is indicated by arrows. (B) Spike morphology. (C) Images of the polymerase chain reaction (PCR) amplicons of three codominant sequence-tagged site (STS) markers (Xrwgs27, Xrwgs28, and Xrwgs29) associated with Sr39. The numbers at the top of the gels are lane numbers: 1, Tc; 2, CS; 3, RL6082; 4, RWG4; 5, RWG3; 6, RWG2; 7, RWG1; 8-10, heterozygous BC₂F₂ plants for RWG4, RWG3, and RWG2, respectively; 11, one BC₂F₂ susceptible plant; 12, CS N2A-T2D (nullisomic for chromosome 2A and tetrasomic for 2D); 13, CS N2B-T2A; and 14, CS N2D-T2A. Diagnostic bands for the Ae. speltoides chromatin carrying Sr39 are indicated by arrows in RL6082. The numbers on the left represent the fragment size in base pairs. (D) Images of PCR amplicons for markers Sr39#22r, Sr39#50s, and Xbarc183. The lane numbers and genotype designations were the same as those for the markers Xrwgs27, Xrwgs28, and Xrwgs29 in image C. (E) Schematic representation of the interchanged chromosomes in RL6082, RWG1, RWG2, RWG3, and RWG4, showing sizes of shortened Ae. speltoides chromosome segments and locations of Sr39 and six PCR-based markers. (F) A portion of the gel image from validation of three STS markers in a set of durum and common wheat cultivars and lines. The numbers on the right represent the fragment size in base pairs.

The FGISH analysis of the new wheat lines indicated that the Ae. speltoides chromatin carrying Sr39 was approximately located in chromosome bin 2BS4-0.75-0.84. To develop molecular markers closely associated with Sr39, we designed primers on the basis of the wheat ESTs mapped to 2BS4-0.75-0.84 and its collinear region of the rice genome. A total of 40 primer pairs (Table S6) were tested for polymorphisms on Thatcher, CS, RL6082, bulked resistant BC₂F₂ plants, bulked susceptible BC₂F₂ plants, CS N2A-T2D, CS N2B-T2A, and CS N2D-T2A. Three primer pairs detected polymorphisms and generated codominant STS markers, which were designated as Xrwgs27, Xrwgs28, and Xrwgs29 (Table 3 and Figure 4C). The marker Xrwgs27 amplified a 740-bp fragment in RL6082 and a 710-bp fragment in Thatcher and CS; Xrwgs28 amplified 360-bp, 450-bp, and 520-bp fragments in RL6082 and a 350-bp fragment in Thatcher and CS; and Xrwgs29 amplified a 540-bp fragment in RL6082 and a 550-bp fragment in Thatcher and CS. The four introgression lines therefore carried bands of 740, 360/450/520, and 540 bp for Xrwgs27, Xrwgs28, and Xrwgs29, respectively. These three markers were tested on a population of 65 BC₂F₂ plants, and all cosegregated with stem rust resistance (Table S7). Identification of markers based on ESTs mapped to bin 2BS4-0.75-0.84 confirmed that the short Ae. speltoides chromosome segments carrying Sr39 was interstitially located in bin 2BS4-0.75-0.84.

TABLE 3.

Three sequence-tagged site (STS) markers linked to Sr39 located on short Aegilops speltoides chromosome segment in the four new wheat lines

			Band size (bp)		EST accession or genomic groupc
Marker	Primer sequence	Tm (50 mM Na+) (°C)a	RL6082	CSb
Xrwgs27	5′ GCCTTGGTGGATTTTGTGAT 3′	60	740	710	BG275030
	5′ GCGCTTTCAGTACAGGGTTC 3′	60
Xrwgs28	5′ AGAGCCTGGGACTGTTGCTA 3′	60	360/450/520	350	tplb0012l12
	5′ CAATGGCACTCTTCAAAGCA 3′	60
Xrwgs29	5′ CGGCTATTGCTCAAAGAAGG 3′	60	540	550	tplb0012l12
	5′ TGTTTCTGTCAGAGGCAACG 3′	60

^aMelting temperature.

^bCS, Chinese Spring.

^cWheat EST (expressed sequence tag) accessions were obtained from Website: http://wheat.pw.usda.gov/cgi-bin/westsql/map_locus.cgi (verified on November 27, 2010). The genomic clone tplb0012l12 sequence (4162 bp) was obtained from Website: http://www.shigen.nig.ac.jp/wheat/komugi/ests/cdnaQueryAction.do?cloneName=tplb0012l12&resourceTypeId=2 (verified on November 27, 2010), which contains a hypothetical protein OsI_27446 of rice (Oryza sativa L.) Indica Group.

Analysis of four introgression lines with the seven markers reported by Mago et al. (2009) revealed that marker Sr39#22r detected the Ae. speltoides chromatin carrying Sr39 for all four lines, but Sr39#50s and Xbarc183 detected the Ae. speltoides chromatin only for RWG4 (Figure 4, D and E). The four remaining markers (BE500705, Xbarc18, Xbarc200, and Xwmc025) produced no polymorphism between RL6082 and CS or Thatcher. Because we only analyzed the BC₂F₂ plants derived from 11 of the 40 BC₂F₁ plants exhibiting dissociation of Sr39 from Xgwm319 with FGISH, the remaining 29 BC₂F₁ plants were tested with markers Sr39#22r, Xrwgs27, Xrwgs28, and Xrwgs29. When compared to RL6082 and the four introgression lines, no polymorphisms were observed in the 29 plants (Table S8), and therefore none of the 29 plants had Ae. speltoides chromosome segments even shorter than the four selected introgression lines.

The four introgression lines (RWG1, RWG2, RWG3, and RWG4) all exhibited a similar level of resistance to seven local stem rust races (THTS, TPMK, RTQQ, RHTS, QFMQ, QFCQ, and TMLK) and Ug99 as the original stock RL6082 (Table 4 and Figure 4A, right). They were morphologically similar to CS with normal seed fertility and had bigger spikes than RL6082 (Figure 4B). Analysis of 16 single BC₂F₃ seeds from the original BC₂F₂ plants for each of four lines with ph1b markers showed that plants homozygous for ph1b were not detected in any line (Figure S3), suggesting that the four introgression lines did not carry the ph1b allele.

TABLE 4.

Infection types produced by four wheat lines and their parental lines to Ug99 (TTKSK) and seven locally maintained races of stem rust

	Infection types to racesa
Genotype	TTKSK	TPMK	TMLK	THTS	RTQQ	RHTS	QFMQ	QFCQ
Thatcher		432	4	43	34	43	32	34
Chinese Spring	4	4	43	4	4	43	43	4
RL6082	2−	12	12	12	12	1	12	12
RWG1	2−	123	12	12	12	12	12	12
RWG2	2−2	21	12	12	12	12	12	12
RWG3	2−	21	12	12	12	12	12	21
RWG4	2+	12	12	12	12	12	12	12

^aInfection types follow Stakman (1962) where 0, ;, 1, and 2 were considered low infection types, and 3 to 4 were considered high infection types. For combinations, order indicates predominant types, hence 432 is predominantly infection type (IT) 4, with decreasing amounts of IT 3 and IT 2. 2−, small IT 2; 2+, large IT 2.

The three STS markers (Xrwgs27, Xrwgs28, and Xrwgs29) were validated with 23 spring wheat, 9 winter wheat, and 8 durum cultivars/lines (Table 5, Figure 4F). Xrwgs27 amplified a 710-bp or a 725-bp fragment in 28 or 12 cultivars/lines, respectively (Table 5). For Xrwgs28 and Xrwgs29, amplicons of different sizes were observed between wheat classes (durum wheat vs. common wheat), but were uniform within class. Amplicons derived from Ae. speltoides chromatin could be differentiated from all the durum and common wheat cultivars/lines tested, and these three markers are thus suitable for breeding resistant cultivars using marker-assisted selection of Sr39.

TABLE 5.

Band size of three marker loci in 40 durum and common wheat cultivars or lines

				Marker band size (bp)b
Cultivar or line	Origina	Growth habit	Type	Xrwgs27	Xrwgs28	Xrwgs29
Jimai 22	China	Winter	Common wheat	725	350	550
Yangmai 16	China	Winter	Common wheat	725	350	550
Shanrong 1	China	Winter	Common wheat	710	350	550
Shanrong 3	China	Winter	Common wheat	710	350	550
Jinan 17	China	Winter	Common wheat	710	350	550
Jinan 177	China	Winter	Common wheat	725	350	550
Zhengmai 9023	China	Winter	Common wheat	725	350	550
Amidon	ND	Spring	Common wheat	725	350	550
Howard	ND	Spring	Common wheat	710	350	550
Alsen	ND	Spring	Common wheat	710	350	550
Grandin	ND	Spring	Common wheat	725	350	550
Glenn	ND	Spring	Common wheat	710	350	550
Faller	ND	Spring	Common wheat	710	350	550
Glupro	ND	Spring	Common wheat	725	350	550
Ernest	ND	Spring	Common wheat	725	350	550
Steele-ND	ND	Spring	Common wheat	710	350	550
Reeder	ND	Spring	Common wheat	710	350	550
Mott	ND	Spring	Common wheat	725	350	550
Kulm	ND	Spring	Common wheat	710	350	550
Parshall	SD	Spring	Common wheat	710	350	550
Granger	SD	Spring	Common wheat	710	350	550
Brick	SD	Spring	Common wheat	710	350	550
Russ	SD	Spring	Common wheat	710	350	550
Briggs	SD	Spring	Common wheat	710	350	550
Traverse	SD	Spring	Common wheat	710	350	550
Sabin	MN	Spring	Common wheat	710	350	550
Oklee	MN	Spring	Common wheat	725	350	550
Ulen	MN	Spring	Common wheat	710	350	550
Ada	MN	Spring	Common wheat	710	350	550
Tom	MN	Spring	Common wheat	725	350	550
Newton	KS	Winter	Common wheat	725	350	550
IL06-14262	IL	Winter	Common wheat	710	350	550
Divide	ND	Spring	Durum wheat	710	355	545 + 550
Ben	ND	Spring	Durum wheat	710	355	545 + 550
Tioga	ND	Spring	Durum wheat	710	355	545 + 550
Grenora	ND	Spring	Durum wheat	710	355	545 + 550
Lebsock	ND	Spring	Durum wheat	710	355	545 + 550
Monroe	ND	Spring	Durum wheat	710	355	545 + 550
Alkabo	ND	Spring	Durum wheat	710	355	545 + 550
Mountrail	ND	Spring	Durum wheat	710	355	545 + 550

^aND, North Dakota; SD, South Dakota; MN, Minnesota; IL, Illinois; KS, Kansas.

^bThe fragment sizes from Aegilops speltoides for Xrwgs27, Xrwgs28, and Xrwgs29 are 740 bp, 360/450/520 bp, and 540 bp, respectively.

DISCUSSION

Success of chromosome engineering for targeted introgression of alien genes is dependent upon elimination of the deleterious effects of the introgressed alien chromatin in the crop plant. Thus far, it has been a challenge to transfer a small amount of alien chromatin containing the gene of interest from one genome to another nonhomologous genome. Only a few genes have been isolated from large alien chromosomal segments, such as Sr26 (Dundas et al. 2007) and SrR (Anugrahwati et al. 2008) for stem rust resistance and Pm21 (Chen et al. 1995) for powdery mildew (Erysiphe graminis DC. f. sp. tritici Em. Marchal) resistance transferred into wheat, and Rfo for fertility restoration transferred from radish (Raphanus sativus L.) into canola (Brassica napus L.) (Feng et al. 2009). The end products of most chromosome engineering studies have been chromosome additions or translocations involving large alien chromosomal segments (see reviews by Jiang et al. 1994; Friebe et al. 1996; Fedak 1999; Xu et al. 2009). The present study demonstrated that targeted introgression of alien genes could be more efficiently and precisely accomplished than in the past by improving the protocol in inducing and identifying homoeologous recombinants.

Construction of an optimal population of homoeologous recombinants:

Development of a population of spontaneous or artificially induced homoeologous recombinants containing an alien gene through a proper hybridization scheme is the first step in eliminating unwanted alien chromatin. Such populations should be favorable for inducing homoeologous recombination and should be easily phenotyped and genotyped. The population size should be large enough (e.g., 1000–2000 individuals) to include desired recombinants carrying the gene of interest. On the basis of these considerations, we used a stem rust-susceptible wheat cultivar CS and its ph1b mutant to develop a large backcross population for Sr39 so that the homoeologous recombinants carrying the gene could be quickly identified using a stem rust test at the seedling stage.

The F₂ populations from crosses involving the CS ph1b mutant have been commonly used to reduce alien chromatin in the wheat genome (Friebe et al. 1996; Qi et al. 2007; Mago et al. 2009). In our study, we used a BC₂F₁ population instead of an F₂ population for induction of new homoeologous recombinants. The BC₂F₁ population was developed by backcrossing the resistant BC₁F₁ plants that were homozygous for ph1b and hemizygous for the translocated alien chromosome segment. By making this backcross, any plant detected to have a reduced alien chromatin was selfed, and homozygous progeny for the reduced alien chromatin were selected. Most importantly, the hemizygous alien chromatin was easily detected by marker screening. Selection in a BC₂F₁ instead of a F₂ or F₃ population may have an advantage beyond ease of selection of homozygotes. In our study, we used a BC₁F₁ plant as male to produce the BC₂F₁ population. As a consequence, pollen competition may be a factor in producing lines with shortened alien chromosome segments. Chen et al. (2005) produced wheat–Leymus racemosus translocation lines using irradiated monosomic-addition plants as male in crosses to cultivars susceptible to Fusarium head blight (caused by Fusarium graminearum Schw.). They concluded that preferential transmission favored selection of gametes carrying a translocated chromosome over gametes carrying a complete alien chromosome. In our study, RL6082 carried a pair of translocated chromosomes with a very large Ae. speltoides chromosome segment, and pollen competition may therefore favor gametes carrying a small Ae. speltoides chromosome segment.

Rapid identification of homoeologous recombinants with reduced alien chromatin using robust DNA marker:

The low output of short translocations from most previous chromosome engineering efforts are largely due to the conventional cytogenetic approaches used for chromosome identification (e.g., chromosome karyotype and pairing analysis, banding, and in situ hybridization) that are tedious, laborious, and not suitable for handling large populations. In this study, we quickly selected individuals with reduced alien chromatin by screening a large segregating population with a single molecular marker, Xgwm319, instead of cytogenetic approaches. Our success in using this marker may be partially attributed to the relative position between Xgwm319 and Sr39. Mago et al. (2009) found that Sr39 was located near the original translocation breakpoint in RL5711; and, as a consequence, the Ae. speltoides chromatin needed little or no reduction distal to Sr39. Xgwm319 is close to the centromere of wheat chromosome 2B (Somers et al. 2004; Sourdille et al. 2004), where recombination rates are reduced and linkage blocks are large (Wu et al. 2003). Thus, any resistant plant that lacks the Ae. speltoides allele at the Xgwm319 locus must have lost a relatively large Ae. speltoides chromosome segment. At the same time, Xgwm319 is sufficiently remote from Sr39 to allow for multiple recombination events between Xgwm319 and Sr39. Using this rationale, only 40 resistant recombinants were identified from the original population of 1048 plants.

Segregation distortion of stem rust resistance in a population of ph1b-induced homoeologous recombinants:

We observed significant segregation distortion of stem rust resistance in the BC₂F₁ population in the study. Segregation distortion is a common feature associated with alien chromosomes in the wheat background (Ceoloni et al. 1996; Marais et al. 2010). It is caused by the presence of segregation distortion (Sd) and gametocidal (Gc) factors. Kerber and Dyck (1990) reported distorted segregation ratios in the breeding of RL5711, which resulted in preferential transmission of the rust resistance genes through male gametes. Our observation that the dissociation frequency was only 7.5% in resistant plants indicated a preferential transmission of the alien chromatin carrying both the Ae. speltoides allele at the Xgwm319 locus and Sr39. The higher dissociation frequency of 44.3% among susceptible plants indicated that there was reduced preferential transmission when only the Ae. speltoides segment carrying Sr39 is present. In the present experiment, segregation ratios may also have been altered due to chromosome breaks, which produced telocentric chromosomes having reduced transmission. Finally, the CS ph1bph1b line used in this experiment had been maintained in a homozygous recessive condition, and the line may have accumulated translocations that, despite not seriously affecting fertility of the CS ph1bph1b line, may have affected meiosis to alter transmission of some chromosome segments. Thus, backcross of the selected lines to agronomically acceptable cultivars is important to eliminate ph1b and remove other unwanted translocations from the wheat lines.

Development of new molecular markers linked to Sr39 in new introgression lines:

Pyramiding of multiple stem rust resistant genes into one cultivar will be necessary for long-term control of stem rust. One of the major factors for successful control of stem rust in North America has been the development of cultivars carrying multiple Sr genes (Zhong et al. 2009). To pyramid multiple resistance genes into one cultivar using stem rust testing, multiple races are needed to differentiate resistance genes. However, because Sr genes are usually effective against multiple races, it is difficult, or even impossible, to select specific races for screening the desired genes. Therefore, molecular markers could serve as an important alternate tool for gene pyramiding. Mago et al. (2009) reported several markers that could detect Ae. speltoides chromatin carrying Sr39. Of those markers, we found that only Sr39#22r could detect the Ae. speltoides chromatin in our three shortest translocation lines, RWG1, RWG2, and RWG3. Sr39#22r is a dominant marker, which is used with Sr39#50s or Xbarc183 to detect heterozygous plants. Because Sr39#50s and Xbarc183 cannot be detected in RWG1, RWG2, and RWG3, additional codominant markers are needed to detect heterozygous plants in these lines. In our study, we utilized information from FGISH, chromosome bin-mapped wheat ESTs, the rice genomic sequence, and the BLAST tool to quickly develop three codominant STS markers tightly linked to Sr39. These markers should be useful for pyramiding of Sr39 with other Sr genes into commercial cultivars for controlling Ug99.

This study demonstrated that alien gene introgression could be efficiently accomplished using an improved scheme of chromosome engineering. In <2 years, we successfully developed four wheat lines with Sr39 residing in a small Ae. speltoides chromosome segment. Over 90% of the Ae. speltoides chromatin was eliminated in these wheat lines. We believe that introgression of stem rust resistance genes from wild relatives into wheat will no longer be a challenging task using the optimal procedure reported in this study. The effective Sr genes present in the secondary and tertiary gene pool will become available for wheat breeding when they are needed for fighting newly emerging races. The successful integration of modern DNA marker technology into chromosome engineering will set a benchmark for future alien gene introgression in wheat and other plant species.