Stripe rust resistance genes in a set of Ethiopian bread wheat cultivars and breeding lines

Stripe rust caused by Puccinia striiformis f. sp. tritici (Pst) is one the most important diseases of wheat in Ethiopia and worldwide. To identify resistance genes, 90 bread wheat lines and 10 cultivars were tested at the seedling stage against one Pst race from Ethiopia and six races from China as well as evaluated for the stripe rust response in an inoculated field nursery at Yangling, Shaanxi province and in a naturally infected field in Jiangyou, Sichuan, China. Resistance genes were postulated using molecular assays for Yr9, Yr17, Yr18, Yr26, Yr29, Yr36, Yr44 and Yr62. Of the 100 entries tested, 16 had all stage resistance to all races. Molecular markers were positive for Yr9 in five genotypes, Yr17 in 21 genotypes, Yr18 in 27 genotypes, Yr26 in ten genotypes, Yr29 in 22 genotypes, Yr36 in 12 genotypes, Yr44 in 30 genotypes, and Yr62 in 51 genotypes. No line had Yr5, Yr8, Yr10 or Yr15. Complete or all stage resistance was observed in genotypes carrying gene combinations Yr9 + Yr18 + Yr44 + Yr62, Yr29 + Yr62 + Yr26 and Yr9 + Yr17 + Yr26 + Yr44 + Yr62. The results are helpful for developing wheat cultivars with effective and more durable resistance to stripe rust both in China and Ethiopia.


Introduction
Stripe (yellow) rust, caused by Puccinia striiformis f. sp. tritici (Pst) is a significant threat in the majority of wheat growing regions of the world (Wellings 2011) including Ethiopia (Abeyo et al. 2014). Yield losses caused by stripe rust in Ethiopia have ranged from 40 to 100% depending on the degree of susceptibility of cultivars, the time of the initial infection and environmental conditions during epidemic development (Badebo et al. 2001). In 2010, Ethiopia experienced one of the most serious stripe rust epidemics in recent times, with more than 600,000 ha of wheat affected and an estimated $US3.2 million spent on fungicides (Abeyo et al. 2014).
Cultivation of resistant cultivars is the most cost effective and environmentally sensible approach to control stripe rust (Line and Chen 1995). However, durable resistance depends on the use of genetically diverse sources of resistance (Wellings 2011). Periodic outbreaks of stripe rust occur in Ethiopia due to lack of knowledge regarding the genetic resistance present in commercial cultivars and breeding populations, and inadequate monitoring of the pathogen race population (Badebo 2002). In Ethiopia, breeding for resistance is solely based on field observations at naturally infected trial sites. There is only limited information on the genetic composition of current cultivars and even less on materials undergoing selection (Dawit et al. 2012).
Following the discovery of the gene-for-gene interaction between plant hosts and their pathogens (Flor 1971), host resistance genes and their corresponding pathogen avirulence genes could be postulated. Dawit et al. (2012) tested 22 Ethiopian bread wheat cultivars and 24 differential lines with 20 Pst races collected from Ethiopia, France and Germany. They postulated different combinations of Yr2, Yr3a, Yr4a, Yr6, Yr7, Yr8, Yr9, Yr27, Yr32 and YrSU in tested materials. Hovmøller (2007) reported Yr1, Yr2, Yr3, Yr4, Yr6, Yr9, Yr15, Yr17, Yr25 and Yr32 in 98 Danish wheat cultivars; Xia et al. (2007) detected Yr2, Yr3a, Yr4a, Yr6, Yr7, Yr9, Yr26, Yr27, YrSel and YrSd in 72 Chinese wheat cultivars and advanced lines; and Sharma et al. (1995) reported Yr2, Yr6, Yr8, Yr9, Yr10, Yr15, YrA and YrSu in tests of 52 wild emmer derivatives and advanced bread wheat lines from Nepal. The drawback of this method of gene postulation is that while it is the most effective for identifying race-specific single resistance genes or simple gene combinations at the seedling stage, it cannot be easily used to identify adult-plant, often non-specific resistance genes that tend to be more common in the case of stripe rust (Wang and Chen 2017). Knowledge of pedigrees is also very valuable and can be used to support gene postulation. For example, given that Ethiopian wheat cultivars are largely based on the CIMMYT germplasm, the presence of 'Milan' in a pedigree gives a clue as to the possible presence of Yr17 in its derivative; the presence of a synthetic (or Ae. squarrosa) in a pedigree is a clue to the possible presence of Yr24 (Yr26) (R.A. McIntosh personal communication). Yr26 is located on a Triticum aestivum/Haynaldia villosa translocation chromosome lB-1V (Ma et al. 2001).
For the latter, more emphasis has to be placed on pedigree and molecular markers. Zeng et al. (2014) identified several Yr resistance genes in 330 leading cultivars and 164 advanced breeding lines from China using pedigree information and molecular markers together with gene postulation. In the present study, we opted to use seedling tests with an array of Pst races and adult plant responses in the field combined with pedigree and marker tests to postulate Yr resistance genes in the Ethiopian bread wheat cultivars and breeding lines.

Plant materials
One hundred bread wheat genotypes (10 cultivars and 90 breeding lines) from the Ethiopian National Wheat Research program were tested for resistance to stripe rust in multi-race seedling tests in a greenhouse in the State Key Laboratory of Crop Stress Biology for Arid Areas and College of Plant Protection, Northwest A&F University, Yangling, Shaanxi, China. The materials were also planted at Yangling, China during 2016-2017 and 2017-2018 growing seasons for adult plant field tests, following inoculation with a mixture of races. Lastly, the material was grown at Jiangyou in Sichuan province, China, in the same crop seasons, and exposed to a natural infection. Sichuan is one of the most stripe rust prone areas in China and resistance is an obligatory trait for cultivar release in that province. The wheat cultivars (lines) 'Mingxian 169' for seedling and 'Xiaoyan 22' for adult plant tests were used as susceptible controls, respectively. Wheat lines or cultivars such as'AvSYr18NIL','92R137'(Yr26), 'AvSYr29NIL', 'RSL65(Yr36)' and 'PI 192252(Yr62)' were included in comparative response tests in the in the stripe rust assessment Supplementary Table 2. A set of 18 wheat genotypes that carry different single known genes for resistance (Wan and Chen 2014) was included in the study as references (Supplementary Table 1).

Seedling tests
Seedling tests were conducted in a greenhouse at Yangling, China. For each wheat entry, 5-7 seedlings were grown in a 10 cm diameter pot in a disease free area. Seedlings were inoculated with fresh urediniospores mixed with talc powder in a 1:20 (v/v) ratio when the first leaves were fully expanded, and incubated in a dew chamber for 24 h at 10°C in darkness. The seedlings were then transferred to a growth chamber with 16 h day at 16°C and 8 h night at 13°C. Six Chinese and one Ethiopian stripe rust races (culture EK3-12, Wan et al. 2017) were used in the study. The virulence: avirulence profiles of the actual isolates of those races are given in Table 1.
Mingxian 169 is susceptible to all known Chinese Pst races, and was used as the susceptible check. Infection types were recorded using the 0-9 scale, i.e. starting 15 days post-inoculation. Infection types 0-3 were considered resistant, 4-5 moderately resistant, 6-7 moderately susceptible and 8-9 susceptible, as described by Line and Qayoum (1992).

Field tests
Field tests were conducted during the 2016-2017 and 2017-2018 cropping seasons at the experimental site of Northwest A&F University, Yangling. Seeds of the 100 wheat entries were planted in two row plots, with the susceptible check, Mingxian 169, included every 20 entries. Mingxian 169 was also planted on both sides of the nursery as an inoculated stripe rust spreader. Inoculations were done during late February and early March. The inoculum used in both seasons was a mixture of urediniospores comprised of CYR33 (30%), CYR32 (30%), CYR31 (10%), CYR29 (10%), CYR23 (10%) and V26 (10%). The nurseries were managed according to agronomic practices common in the region, except that the field was irrigated twice during the epidemic development. A similar experimental design was used in the Sichuan province where the material was subjected to natural infection. In the latter trial, Xiayan22 was used as the susceptible control. Infection types and disease severities on the uppermost three leaves were recorded twice according to Line and Qayoum (1992) during mid and late May when the plants had reached the booting and milk stages respectively, and disease severities (DS) on the susceptible check exceeded 80%.

DNA extraction
The entries were grown in a greenhouse up to the two leaf stage for isolation of genomic DNA. A total of 100 mg of leaf tissue was collected from each genotype and immediately frozen in liquid nitrogen. DNA of each entry was extracted following the protocol described by Song and Henry (1994).
DNA concentration was measured using a ND-1000 spectrophotometer (Thermal Scientific, Wilmington, DE), and quality was checked by the agarose gel electrophoresis.

PCR amplification and electrophoresis
Polymerase chain reactions (PCR) were done using a S1000 Thermal Cycler (BIO-RAD, CITY). The PCR reaction (total volume = 15 ll) contained 1.0 unit of Taq DNA polymerase, 1.5 ll of 109 buffer (50 mmol KCl, 10 mmol Tris-HCI, pH 8.3), 2.0 mmol MgCl 2 , 200 lmol of dNTP, 0.6 lmol of each primer and 50-100 ng of template DNA. Primers of markers for identifying specific Yr genes were synthesized by Sangon Biotech Co, Ltd (Shanghai, China) The PCR conditions were: denaturation at 94°C for 4 min, followed by 35 cycles (each consisting of 1 min at 94°C, 30 s at 40-60°C (depending upon the primers), 1 min at 72°C), final extension for 10 min at 72°C and a 4°C holding step. In the case of the STS-7/8, STS-9/10 and URIC/LN2 primer pairs, after implication, the PCR products were digested with DpnII (New England Biolabs, USA) as explained by Chen et al. (2003), with some modifications. The reaction solution for enzymatic digestion contained 10 ll of PCR product, 1 ll of restriction enzyme DpnII (New England Biolabs, Beverly, MA) and 2 ll of 109 buffer for DpnII (New England Biolabs). Samples were kept at 37°C for 4 h and the digested products were separated on agarose gels.
Yr26 was detected using SNP markers WRS303 and WRS467. SNP genotype calling and clustering were done using Illumina Genome Studio Polyploid Clustering v1.0 and Affymetrix Genotyping Console TM (GTC) software, respectively. Details of the publicly available molecular markers used to detect other resistance genes are listed in Table 2.

Disease data
Seedling infection type data for the 100 tested entries inoculated with seven Pst races are listed in Supplementary Table 2. Fifty-six lines were highly resistant (IT 0-3) and nine lines were moderately resistant (IT 4-5) to the Ethiopian race; 40 lines were highly resistant and 8 lines moderately resistant to all Chinese races; 42 lines were highly resistant and 10 lines moderately resistant to both the Ethiopian and Chinese races. Fourteen lines and two cultivars (DANDA'A, and HULLUKA) were highly resistant (IT 0-3) to all races. Only three lines were considered susceptible in field tests in China, with infection severity ranging from 50 to 70%, including cv. MILLENIUM. This was not unexpected given the comprehensive international screening program conducted by CIMMYT and the strong emphasis on the adult plant resistance which is often provided by combinations of resistance genes individually conferring low to intermediate levels of resistance.
The presence of the 'Pastor' Yr46 gene carrier cultivar in the pedigree of nine breeding lines such as ETBW8823, ETBW8827, ETBW8571, ETBW8663, ETBW8491, ETBW8492, ETBW8311, ETBW9470 and ETBW8848, and the presence of a synthetic (that is, of Ae. squarrosa) in the pedigrees of 8 lines possessing Yr24 (Yr26) ( Table 2) may have contributed to improved field resistance. Pedigrees, seedling and adult plant responses, and molecular marker detection for 10 bread wheat cultivars and 23 breading lines are presented in Table 3.
Each individual accession had similar adult plant stripe rust disease responses except for three entries       presence of pedigree to those tested Yr genes, ab molecular marker detection and pedigree, bc gene postulation and molecular marker detection across testing sites and years (Supplementary Table 2).

Detection of Yr resistance genes using molecular markers
The outcomes of molecular marker tests to predict the presence of Yr9, Yr17, Yr18, Yr26, Yr29, Yr36, Yr44 and Yr62 in the tested entries are provided in Supplementary Table 2. At least two molecular markers were used to identify each gene in the tested genotypes and the Yr gene (s) was considered to be present if both markers predicted its presence.

Yr9
Primer pairs H20 (Liu et al. 2008) and P6M12P (Mago et al. 2005) were selected to identify Yr9. Primer H20 F/R produced a 1598 bp diagnostic fragment (Fig. 1); primer P6M12-P F/R produced diagnostic fragments of 250 and 350 bp. Five bread wheat breeding lines and none cultivar had the Yr9 marker.

Yr17
Resistance genes Yr17/Lr37/Sr38 (a block of yellow, leaf and stem rust resistance genes) are located within a segment of Ae, ventricosa chromosome 6N that translocated to wheat chromosome 2AS (Tanguy et al. 2005). Primers URIC and LN2 amplify fragments of (?) 285 bp (from the N genome) indicating the presence of the resistance genes while (-) 275 bp (A genome) indicate their absence. Digestion of the PCR products with restriction enzyme Dpn11 facilitates differentiation between these two bands. Primer SC-385 amplifies a 385 bp fragment. Twenty-one entries tested positive for Yr17, including cvs HOGGANA and MERARO in Supplementary  Table 2.

Yr18
A block of genes Yr18/Lr34/Pm38 confers APR to yellow rust, leaf rust and powdery mildew. Krattinger et al. (2009) developed two primer pairs for detecting the presence (? Yr18) and absence (-Yr18) of this block. The first multiplex reaction was developed based on primer combinations L34DINT9F/ L34LUSR and csLV34 F/R, which results in amplification of either two bands (517 bp and 150 bp) which indicate the presence of Yr18, or one band (229 bp) which suggests the absence of Yr18. Of the 100 tested entries, 27 carried Yr18.

Yr29
The STS marker Bac17R and the SSR marker Wmc44 (both suggested by Rosewarne et al. 2006) were used to detect this resistance gene. Primer set Bac17R amplifies a 1700 bp diagnostic fragment and primer set Wmc44 amplifies * a 260 bp diagnostic fragment.
Twenty-two entries had the Yr29 associated marker polymorphisms. The seedling and adult plant resistance responses of the latter lines are shown in Table 2.

Yr36
Two gene-specific primer sets to detect the locus were developed by Fu et al. (2009). Primer set Yr36Ela amplified a 911 bp fragment and primer set Yr36 START amplified 871 or 537 bp fragments. Twelve entries were found to have Yr36.

Yr44
Two markers for Yr44 were developed by Sui et al. (2009). Primer Pwb5/pWNR amplifies a 380 bp diagnostic product whereas wgp100 amplifies a 820 bp fragment. Thirty entries tested positive for Yr44. Among the lines with the gene, ETBW8816, ETBW8800, ETBW8917 and ETBW8260 were resistant to all races, whereas 26 showed resistance to one or more races, indicating that these entries might also carry other effective resistance genes.

Yr62
Two SSR markers were mapped close to Yr62 by Lu et al. (2014). These markers (Gwm192 and Gwm2511) amplified 222 bp and 133 bp diagnostic bands in the present material, respectively. Fifty-one entries were identified as likely carrying Yr62.

Discussion
The wide deployment and cultivation of bread wheat cultivars with Yr9 led to serious stripe rust epidemics in many countries, including Ethiopia, when the gene became ineffective Wellings 2011). Hence, it is of paramount importance to identify and develop breeding lines with effective resistance to the predominant and emerging races that threaten wheat production. To detect effective stripe resistance genes in wheat germplasm, combinations of Pst races with different virulence spectra should be employed in breeding material (Walker 1965). In this study, 100 entries were evaluated using a predominant Pst race from Ethiopia and six races representing different race categories from China. The disease phenotype data were combined with DNA marker data to postulate likely Yr genes for which dependable markers are available. The entries were more susceptible to Chinese races than to the Ethiopian race. We identified 14 breeding lines and two cultivars that were resistant to all races. Wan et al. (2017) also identified several genotypes resistant at the seedling stage among the tested material, using two additional Pst races from Ethiopia and three from the United States.
Molecular markers are helpful to identify resistance genes and speed up selection, if such markers are available (Chen 2013). In the present study, we used Fig. 1 Electrophoregrams produced with the use of different primers: The marker diagnostic product was confirmed using, well-established different stocks and Yr gene carrier cultivars a identification of Yr9, AVS/6*Yr9 is the NIL known to carry Yr9. b identification of Yr17, AVS/6*Yr17 is the NIL known to carry the Yr17 gene. c identification of Yr18 AvSYr18NIL is the known carry Yr18. d identification of Yr29, Pavon 76 known to carry Yr29, E, identification of Yr36, RSL65 know to carry Yr36, f identification Yr44, AvS/Zak (1-1-35-line1) is known to carry Yr44, and g identification of Yr62, PI 192252 know to carry Yr62 gene. Note, M: bands in the above Fig. 1, 1598 DNA markers to predict the presence of Yr9, Yr17, Yr18, Yr26, Yr29, Yr36, Yr44 and Yr62, all of which were detected with various frequencies among the tested entries. When molecular markers were applied to identify the Yr5, Yr15, Yr10 and Yr8 resistance genes, indicative DNA fragments were observed only among positive controls, indicating that none of the entries carries these genes. Sixteen entries were resistant both at the seedling and adult-plant stages (ETBW8905, ETBW8800, ETBW8917, ETBW8823, ETBW8827, ETBW8583, ETBW8705, ETBW8486, ETBW8491, ETBW9015, ETBW9470, ETBW8260, ETBW8437 and ETBW8387, DANDA'A and HUL-LUKA) suggesting that these entries might carry either a novel resistance gene(s) or an effective combination of resistance genes. Similar to our study, Zheng et al. (2017) obtained multiple gene combinations among Yr resistance genes in Chinese wheat materials that were effective in the field. Chen (2013) suggested that combinations of all stage resistance genes such as Yr5 or Yr15 and HTAP resistance genes such as Yr18, Yr29, Yr36, Yr39, Yr52, Yr59 and Yr62 should be incorporated in breeding material to develop wheat cultivars with strong and durable resistance.
Yr9 was identified in five entries using both rye chromosome specific and STS markers with lines ETBW8815 and ETBW8917 being resistant to all of the tested Pst races, indicating that these entries probably possess another resistance gene(s) or effective gene combinations. These results are consistent with other recent studies (Wu et al. 2016;Zeng et al. 2014).
Molecular marker detection indicated that 21% of the tested entries carry Yr17. Zeng et al. (2014) reported that 45 of 494 Chinese wheat cultivars had Yr17, while Wu et al. (2016) identified Yr17 in 10.3% of the Chinese wheat cultivars that they studied. Using gene specific markers, L34DINT9F and L34PLUSR, we determined that 27% of Ethiopian bread wheat genotypes carry Yr18. Yang et al. (2008) identified Yr18 in 89.6% of 231 Chinese wheat cultivars and 6.1% of 422 landraces. Yr26 was present in 10% of the tested entries of our study (markers WRS303 and WRS467), where as Zeng et al. (2014) reported that 15 Fig. 2 a identification of Yr26 resistance gene using SNP markers, genotyping data plotted using Kluster caller software. Genotyped samples marked red indicated absence of Yr26 in the tested entries for the allele reported with HEX, those marked blue indicated that the presence of Yr26 in the tested entries for the FAM allele (27.27%) of the 494 cultivars in their study had Yr26 (detected with SSR marker WE173).
Using STS markers we identified 22 entries with Yr29. Twelve entries had Yr36 (marker Yr36 START). Yr36 contain a kinase and a putative START lipidbinding domain and confers broad resistance to stripe rust at relatively high temperature (Fu et al. 2009). Zheng et al. (2017) also reported that few of the 672 wheat accessions tested by them had Yr36. In our study, Yr44 was found in 30% of the entries using the STS marker pWB5/pWN1R. Yr44 was first discovered in the spring wheat cv. Zak; it is located on chromosome 2B and confers race-specific or all stage resistance (Sui et al. 2009). Using SSR markers gwm192 and gwm25, we identified 51 entries with Yr62 among 100 tested cultivars. Yr62 was first identified in the spring wheat germplasm PI192252 and provides adult plant stripe rust resistance (Lu et al. 2014).
Pedigree analysis on 79 wheat lines showed that 51 lines had Yr9 in their pedigree but only three lines were confirmed with molecular marker to carry of Yr9. Badebo et al. (1990) also identified 67% of tested 42 commercial wheat contained Yr9. Pedigree analysis of eight lines suggested the presence of Yr17; molecular markers confirmed only three. Another eight lines gives clue carried Yr26 confirmed by molecular detection was one. Nine breeding lines possibly had Lr67/Yr46. These are slow-rusting resistance genes for both leaf and stripe rusts, originally identified in cv. 'Pastor' by Sybil et al. (2011).
The information obtained on the resistance profiles if various stocks tested here will contribute towards diversification and knowledge-based use of effective resistance genes in the Chinese and Ethiopian wheat breeding programs, and will be useful in attempts to develop cultivars with durable or long lasting resistance.