Pyramiding stem rust resistance genes to race TTKSK (Ug99) in wheat

Abstract

There has not been a major wheat stem rust epidemic in Canada since 1954. The detection of race TTKSK, with virulence to Sr31 and most other stem rust resistance genes in wheat, presents a threat to the wheat crop worldwide. Single disease resistance genes are usually effective for only several years before the pathogen changes genetically to overcome the resistance. To produce more durable resistance to race TTKSK, as well as to leaf rust and Fusarium head blight (FHB), we produced pyramided lines with several resistance genes. A population of 68 doubled haploids was produced from the hybrid ‘AC Cadillac’ (SrCad, Lr34)/‘Carberry’ (Lr34, Fhb1) × ‘RL5405' (Sr33)/‘Carberry’ (Lr34, Fhb1). A total of 16 combinations of variable numbers of the four genes were recovered by marker assisted selection. The most notable combination with four of the genes was Sr33, SrCad, Lr34 and Fhb1. Two such lines were recovered. All pyramided lines with Lr34 and SrCad or Sr33 were resistant to race TTKSK, while two control wheat varieties 'Shaw’ and ‘Little Club’ were susceptible. The two pyramided lines (SrCad-Sr33-Lr34-Fhb1) were relatively more resistant than the lines containing only one gene for stem rust resistance. The pyramided lines with Lr34 showed good resistance to leaf rust races 12–3 MBDS and TDBG. The pyramids with Fhb1 did not show a lowered expression of Fusarium head blight.

Résumé

Il n’y a pas eu de grave épidémie de rouille de la tige du blé au Canada depuis 1954. La détection de la race TTKSK, virulente à l’égard du gène de résistance Sr31 et de la majorité des autres gènes de résistance à la rouille de la tige chez le blé, constitue une menace pour les cultures de blé, et ce, partout dans le monde. La résistance à la maladie attribuable à un seul gène n’est habituellement efficace que pour quelques années avant que l’agent pathogène change génétiquement pour finalement vaincre la résistance. Pour développer une résistance plus durable à la race TTKSK, ainsi qu’à la rouille des feuilles et à la brûlure de l’épi causée par Fusarium (FHB), nous avons produit des lignées cumulant plusieurs gènes de résistance. Une population de 68 dihaploïdes a été créée à partir de l’hybride ‘AC Cadillac’ (SrCad, Lr34)’/‘Carberry’ (Lr34, Fhb1) × ‘RL5405' (Sr33)/‘Carberry’ (Lr34, Fhb1). En tout, 16 combinaisons affichant un nombre variable des quatre gènes ont été récupérées par sélection effectuée à l’aide de marqueurs moléculaires. La sélection la plus remarquable possédant quatre des gènes était Sr33, SrCad, Lr34 et Fhb1. Deux telles lignées ont été récupérées. Toutes les lignées possédant la combinaison Lr34 et SrCad ou Sr33 étaient résistantes à la race TTKSK, tandis que deux variétés témoins, ‘Shaw’ et ‘Little Club’, lui étaient réceptives. Les deux lignées cumulant les gènes de résistance (SrCad-Sr33-Lr34-Fhb1) étaient relativement plus résistantes à la rouille de la tige que les lignées en contenant qu’un seul. Les lignées cumulant les gènes, comportant Lr34, ont affiché une bonne résistance aux races 12-3 MBDS et TDBG de la rouille des feuilles. Les lignées comportant le gène Fhb1 n’ont pas affiché de diminution de l’expression de la brûlure de l’épi causée par Fusarium.

Keywords:

• doubled haploid
• Fusarium head blight (FHB)
• gene pyramiding
• leaf rust
• MAS
• race TTKSK
• resistance
• Ug99

Mots clés:

• brûlure de l’épi causée par Fusarium (FHB)
• dihaploïde
• empilage de gènes
• MAS
• race TTKSK
• résistance
• Ug99

Introduction

The most important biotic constraints to wheat production in temperate regions of the world are rust diseases caused by Puccinia triticina (leaf rust), P. striiformisf. sp. tritici (stripe rust), and P. graminis f. sp. tritici Pers.: f. sp. tritici Eriks. & E. Henn (stem rust, Pgt), as well as Fusarium head blight (FHB) primarily caused by F. graminearum. FHB is the most important wheat disease in Canada and causes major grain yield and quality losses for producers. More importantly, however, F. graminearum produces mycotoxins, such as deoxynivalenol (DON), which pose a risk to the safety of human food and livestock feed (Marasas et al., 1984). FHB resistance is controlled by many QTL (Buerstmayr et al., 2009). Fhb1 is the first QTL to be defined as a distinct locus (Cuthbert et al., 2006) and the markers developed for it were used in the present study. The last wheat stem rust epidemic in North America occurred from 1953 to 1955 (Peturson, 1958). Since then, resistant cultivars have controlled wheat stem rust in North America. However, recently, the evolution of a new Pgt pathotype TTKSK (also known as Ug99) in Africa has threatened the global wheat industry. TTKSK was initially detected in Uganda in 1999 and later spread to neighbouring countries of Kenya and Ethiopia and now has reached Iran (Singh et al., 2006; FAO, 2008) and many other countries in Africa (Singh et al., 2006). TTKSK is virulent on many of the stem rust resistance genes currently used in different wheat growing regions (Jin & Singh, 2006). Among 56 designated and a few undesignated stem rust resistance genes in wheat, only eight designated genes confer resistance to TTKSK (Pretorius et al., 2000; Jin et al., 2007; Hiebert et al., 2011). The genes SrCad and Sr33 were among the genes shown to be effective against race TTKSK. A survey confirmed that SrCad is the basis for all of the seedling resistance to this race in Canadian wheat cultivars (Hiebert et al., 2011). The similarity of infection type, marker order and genetic distance between SSRs linked to resistance in the LMPG/‘Norin 40' and ‘RL6071'/‘Peace’ populations suggest that SrCad and Sr42 may represent the same allele or different alleles of the same locus (Hiebert et al., 2011; Ghazvini et al., 2012). However, for the purpose of this study and to avoid overstating the results of Ghazvini et al. (2012), we will use the SrCad designation throughout.

Most leaf rust resistance genes are race-specific and therefore succumb to new variants of the pathogen soon after their deployment, while slow leaf-rusting resistance has proven to be more durable. The slow-rusting gene cluster Lr34/Yr18, located on chromosome arm 7DS, has provided durable resistance to leaf rust and stripe rust since the early 20th century (Dyck, 1977, 1987; Singh, 1992). Lr34/Yr18 also confers resistance to powdery mildew (Blumeria graminis) (Spielmeyer et al., 2005), stem rust (Dyck, 1987) and Barley yellow dwarf virus (Singh, 1993).

Host resistance is the most effective and environmentally sound method to control plant diseases (Mesterhazy et al., 2003; McIntosh et al., 2010). Pyramiding several rust resistance genes is desirable to extend their effectiveness, but is difficult to detect due to epistatic effects and similar phenotypic responses. However, development of molecular markers closely linked with the desirable resistance genes has enabled the use of marker-assisted selection to combine several genes. The objective of this study was to pyramid SrCad and Sr33 genes that provide seeding stem rust resistance to race TTKSK. Since the parents used in crosses also contained the leaf rust resistance gene Lr34 and FHB resistance gene Fhb1, these loci were also monitored.

Materials and methods

Plant material and doubled haploid production

The F₁ hybrid used for producing the DH population was: ‘AC Cadillac’ (SrCad, Lr34)/‘Carberry’ (Lr34, Fhb1) × ‘RL5405' (Sr33)/‘Carberry’ (Lr34, Fhb1). The pedigrees of ‘AC Cadillac’ and ‘Carberry’ have been described (Depauw et al., 1998, 2011). The pedigree of experimented line ‘RL5045' is ‘Tetra Canthatch’/Ae. tauschii var. strangulata accession RL5288.

Doubled haploid plant production

The wheat × maize pollination method was used to produce the DH progeny according to standard methods (Fedak et al., 1997). The seven F₁ hybrid wheat plants were grown in single 17 cm Fiber-Grow pots in growth chambers maintained at day/night temperatures of 20/15°C, a 16 h photoperiod, and light intensity of 350 μmol⁻²s⁻¹. The corn variety ‘Quickie’, the source of pollen, was grown in a greenhouse maintained at a 16 h photoperiod and light intensity of 300 μmol⁻²s⁻¹. The temperature regime was maintained at 20/15^oC day/night.

DNA isolation, PCR amplification and molecular markers

DNA isolation for PCR analysis was carried out by standard methods (Dellaporta et al., 1983) with minor modifications. Previously developed PCR markers for all resistance genes used for analysis are listed in Table 1. For Sr33 and Fhb1, the amplification programme consisted of 45 cycles of 94°C for 1 min, 55–61°C (depending on annealing temperature for each primer pair) for 1 min, 72°C for 2 min, and a final extension at 72°C for 7 min. The programme for L34 was carried out as 94°C for 30 s, 55°C for 30 s, 72°C for 45 s for 50 cycles, 72°C for 4 min. The PCR condition for SrCad consisted of 94°C for 1 min, 42°C for 1 min, 72°C for 2 min for 35 cycles, 72°C for 10 min. PCR products of Sr33, SrCad, and Lr34 were separated by 2% agarose gel electrophoresis in TAE buffer and the PCR products of Fhb1 were separated by Elchrom’s precast spreadex gel with Elchrom Origins electrophoresis unit.

Table 1. Resistance genes, their chromosome location, linked molecular markers, and PCR primers used for marker assisted selection in this study.

Gene (trait)	Chromosome location	Linked marker	Primer pair	Reference
Sr33	1D	Xcfd15	5ʹ CTCCCGTATTGAGCAGGAAG 3’	Sambasivam et al. (2008)
			5ʹ GGCAGGTGTGGTGATGATCT 3’
SrCad	6D	FSD_RSA	5ʹ GTTTTATCTTTTTATTTC 3’	Hiebert et al. (2011)
			5ʹ CTCCTCCCCCCA 3’
Lr34	7D	L34SPF/L34DINT13R2	5ʹ GGGAGCATTATTTTTTTCCATCATG 3’	Laguda et al. (2009)
			5ʹ ACTTTCCTGAAAATAATACAAGCA 3’
		L34DINT9F/L34MINUSR	5ʹ TTGATGAAACCAGTTTTTTTTCTA 3’
			5ʹ GCCATTTAACATAATCATGATGGA 3’
Fhb1	3B	UMN10	5ʹ CGTGGTTCCACGTCTTCTTA 3’	Liu et al. (2008)
			5ʹ TGAAGTTCATGCCACGCATA 3’

Stem rust assays

Selected DH progeny, three parents and two stem rust susceptible wheat varieties were evaluated with stem rust race TTKSK. Ten plants from each DH progeny were assessed for seedling resistance at 14 d post-inoculation using a 0–4 scale (Stakman et al., 1962). Infection types from 0–2+ were considered resistant responses, whereas those from 3–4 were considered susceptible responses.

Leaf rust assays

The DH lines and checks were grown as two sets of lines, with three seeds per line, in flats in a greenhouse. They were inoculated with urediniospore suspensions in an oil carrier with isolates 12–3 MBDS and 06–1-1 TDBG (Long & Kolmer, 1989; McCallum et al., 2017). The inoculated flats were misted for 12 h and symptoms recorded at 14 days after inoculation. Ratings of 0, 1 and 2 indicated a resistant reaction whereas readings of 3, 3+ indicated susceptible reactions.

Fusarium head blight assay

The FHB reaction of the doubled haploid lines was determined by point inoculation using standard methods. Central florets of spikes at 50% anthesis were inoculated with 10 µL of a 50 000 spores mL⁻¹ suspension of a mixture of three virulent isolates of F. graminearum obtained from Canadian Collection of Fungal Cultures (DAOMC), Agriculture and Agri-Food Canada. Plants with inoculated spikes were placed in a mist chamber at 100% RH for 48 h, then returned to normal growth chamber conditions of 16 h light, and day/night temperature regime of 20/15°C. At 21 days after inoculation, the symptoms were scored and expressed as per cent infected florets.

Results

DH population production

A total of 5328 florets were pollinated on F₁ plants, 168 embryos were rescued (3.2%) and 79 haploid plantlets produced. Sixty-eight doubled haploid plants were recovered following the colchicine treatment (1.3% of pollinated florets, 40.5% of rescued embryos, 86.1% recovery after colchicine treatment).

Production of plants with different gene combinations

The observed frequencies of each genotype were in general agreement with expected frequencies (Table 2). The one exception was lines with none of the four loci, where 2.85 lines were expected but six were observed. The 68 DH lines were classified into 16 combinations of the 4 loci, of a possible 32 combinations. Two lines with a pyramid of Lr34-SrCad-Sr33-Fhb1 and five lines with a pyramid of Lr34-SrCad-Sr33 were developed.

Table 2. Frequencies of gene combinations in 68 pyramided DH lines.

Order	Lr34	SrCad	Sr33	Fhb1	Expected Freq.	Expected #	Observed #
1	Lr34	SrCad	Sr33	Fhb1	0.03	2.12	2
2	Lr34	-	Sr33	Fhb1	0.07	4.43	6
3	Lr34	SrCad	-	Fhb1	0.09	5.88	7
4	Lr34	SrCad	Sr33	-	0.04	2.85	5
5	Lr34	SrCad	-	-	0.12	7.91	6
6	Lr34	-	Sr33	-	0.09	5.96	4
7	Lr34	-	-	Fhb1	0.18	12.30	12
8	-	-	Sr33	Fhb1	0.01	0.76	1
9	Lr34	-	-	-	0.24	16.55	16
10	-	SrCad	-	-	0.02	1.36	2
11	-	-	-	Fhb1	0.03	2.12	1
12	-	-	-	-	0.04	2.85	6
13	-	-	Sr33	-	0.02	1.03	0
14	-	SrCad	-	Fhb1	0.01	1.01	0
15	-	SrCad	Sr33	-	0.01	0.49	0
16	-	SrCad	Sr33	Fhb1	0.01	0.37	0
TOTALS					1	68	68

The F₁ plants that were used to generate the DH lines were heterozygous for the loci SrCad, Sr33 and Fhb1, but not all were heterozygous for Lr34. Consequently, a good fit was observed for a 1:1 segregation ratio among the DH lines for the first three loci but not for Lr34. There were more loci recovered than expected for Lr34 and less than expected for the other three, although the values were not significantly different. Segregation distortion may have been a factor affecting the expected frequencies of some of the loci, but it was more likely the small population size that affected the frequencies.

Disease resistance

Ten of the 30 lines putatively containing Lr34 with either SrCad or Sr33 were tested with race TTKSK. All lines with Lr34 and SrCad or Sr33 were resistant to TTKSK, while the standard check cultivars ‘Shaw’ and ‘Little Club’ displayed susceptible ITs (Table 3 and Fig. 1). The three parents were also tested with race TTKSK. As expected ‘AC Cadillac’ and ‘RL5405' displayed resistance while ‘Carberry’ displayed a susceptible reaction (Table 3).

Table 3. Stem rust infection type of various gene stacks.

				Leaf rust
Order	Gene type	Stem rust infection type	FHB infected florets (%)	TDBG	MBDS
1	Lr34-SrCad-Sr33-Fhb1	1-, 1	27.7	1	1+
2	Lr34-Sr33-Fhb1	1-, 1	26.6	1	1-
3	Lr34-SrCad-Fhb1	1-, 1+	35.0	1	1
4	Lr34-SrCad-Sr33	1-, 1	46.0	1+	1+
5	Lr34-SrCad	1-, 1+	48.0	1	1
6	Lr34-Sr33	1-, 1+	45.6	1-	1-
7	Lr34-Fhb1	-	29.5	1-	1-
8	Sr33-Fhb1	2-, 3	32.5	3+	3
9	Lr34	3-, 3+	43.0	1-	1-
10	SrCad	1-, 2	74.5	3	3
‘AC Cadillac’	Lr34-SrCad	1-, 1	100	1-	1-
‘RL5405’	Sr33	2-, 3	86.4	3	3
‘Carberry’	Lr34-Fhb1	3-,	80.0	1-	1-
‘Shaw’	-	3-	85.0	3	3
‘Little Club’	-	3-, 3+	86.0	3	3
‘Sumai3’	-	-	7.3	-	-
‘Roblin’	-	-	96.0	-	-

Fig. 1. (Colour online) Phenotype of susceptible (A) and resistant (B) lines to TTKSK.

Eleven of the derived lines and all parents used in this experiment were inoculated with leaf rust races TDBG and MBDS. Eight of the derived lines contained Lr34 and all gave a resistant reaction (Table 3). The check cultivars ‘AC Cadillac’ and ‘Carberry’ both gave resistant reactions as would be expected since both carried Lr34. The derived lines that gave a susceptible reaction were Sr33-Fhb1, SrCad and Fhb1. They all lacked Lr34. Similarly, the checks giving a susceptible reaction were ‘RL5405', ‘Shaw’ and ‘Little Club’ and all were deficient in Lr34.

The genotypes shown in Table 3 were all inoculated with a mixture of three virulent isolates of Fusarium graminearum. The check cultivars ‘Sumai3' (R) and ‘Roblin’ (S) gave readings of 5.5 and 92.3% infected florets, respectively. All lines in the test had much higher levels of infected florets than ‘Sumai3'. Lines that combined the Fhb1 locus tended to have lower values than lines without, regardless of the presence of other genes.

Discussion

Pyramiding resistance genes – the combination of different resistance genes against the same pathogen in one breeding line or cultivar – likely will increase the longevity of resistance of a genotype – a potentially useful strategy to create more durable and broad-spectrum resistances (Liu et al., 2000; Mago et al., 2011). Successful marker-assisted pyramiding of genes with resistance to either leaf rust, powdery mildew or stem rust has been reported previously (Kloppers & Pretorius, 1997; Liu et al., 2000; Mago et al., 2011). In the present study, to verify resistance, genes for stem rust as well as leaf rust and FHB resistance were assayed. Pyramiding different disease resistance genes would be an ideal strategy to reduce the loss of crop production due to multiple pathogens.

Doubled haploid technology can accelerate the breeding process and when combined with marker-assisted selection, can provide a route for producing lines homozygous for multiple genes. In the present study, marker-assisted selection combined with doubled haploid breeding technology was applied for pyramiding disease resistance genes. Some of the derived lines were tested with race TTKSK of stem rust and all showed low disease values, indicating that recombination between genes and markers likely did not occur. By the end of the gene pyramiding process, two pyramided lines with homozygous SrCad, Sr33, Lr34 and Fhb1 had been selected. The lines with Lr34, SrCad, Sr33 and Fhb1 should provide long-term stem rust resistance and leaf rust resistance plus a degree of resistance to FHB. The focus of this study was the production of pyramids containing more than one seedling stem rust resistance gene. But the addition of Lr34 will provide much needed leaf rust resistance in Canadian wheat breeding programmes.

In addition to leaf rust resistance, Lr34 has been shown to enhance the effectiveness of stem rust resistance genes (Hiebert et al., 2011; Kolmer et al., 2011) and provides some resistance to stripe rust (Singh, 1992) and powdery mildew (Blumeria graminis), where it is called Pm38 (Spielmeyer et al., 2005). In addition, SrCad is tightly linked in coupling to the bunt resistance gene Bt10 (Menzies et al., 2006), thus bunt resistance is also included in the gene pyramids containing SrCad.

Race TTKSK is virulent on most Sr genes (Jin et al., 2007). Among 56 designated and a few undesignated stem rust resistance genes in wheat, only eight designated genes in the primary gene pool (Sr9h, Sr13, Sr14, Sr22, Sr28, Sr35, SrCad and Sr45) confer resistance to race TTKSK (Pretorius et al., 2000; Jin et al., 2007; Hiebert et al., 2011; Rouse et al., 2014). Almost all of the other genes in that list were derived from alien sources. Stem rust race TTKSK was used in the present study to verify resistance. There are about 10 other known races of stem rust. However it is well known that SrCad and Sr33 are effective against all of these other races (Tom Fetch, personal communication).

FHB is also a devastating disease of wheat in all temperate regions of the world. Inheritance of resistance is oligogenic and complex. It has been shown (Buerstmayr et al., 2009) that FHB resistance QTL are present on virtually every wheat chromosome. To add to the complexity, it has been shown (Somers et al., 2003) that Type I, Type II and Type III resistances are controlled by separate QTL. In addition, it is becoming obvious (Buerstmayr et al., 2009) that to obtain a high level of resistance, minor, unmapped QTL should supplement the resistance provided by the major QTL. Fhb1 is currently the most effective gene for FHB resistance (Cuthbert et al., 2006). So its addition to the Sr pyramids will offer some degree of FHB resistance in addition to the leaf and stem rust resistance.

There are two factors that may explain in part the high FHB values shown in Table 3. First is that the post inoculation environmental conditions in our growth facilities are excessive and thus overwhelm any FHB QTL. Second, the genetic control of FHB resistance in wheat is complex. It has now been repeatedly shown (Buerstmayr et al., 2009) that a number of QTL, both major and minor, are required as a package to provide some degree of FHB resistance. ‘Sumai3' is known to carry a number of major and minor QTL for FHB resistance, one being Fhb1. Perhaps other genotypes shown in Table 3 have only Fhb1 and thus have much lower FHB resistance than ‘Sumai3'.

One of the objectives of this study was to produce germplasm with stem rust/leaf rust resistance added to lines with some degree of FHB resistance. This objective has only been partially achieved. The pyramided lines with disease resistance produced in this research will serve as donors of multiple resistance genes and some of the lines with good agronomic traits also can be used as enhanced germplasm.

Acknowledgements

Financial assistance in the form of grants from the Grain Farmers of Ontario, Western Grains Research Foundation and Science and Technology Planning Project of Qinghai Province, China [2011-Z-716] is gratefully acknowledged.

Additional information

Funding

This work was supported by the Grain Farmers of Ontario grant [W2011GE02]; Western Grains Research Foundation grant [AIP-CL06].