Introduction

Stem rust, caused by the fungal pathogen Puccinia graminis Pers.:Pers. f. sp. tritici Eriks & E. Henn (Pgt), poses a threat for wheat production. While resistance to stem rust had been stable in wheat cultivars for many years, Pgt race TTKSK (Ug99) with virulence on many genes was detected in Uganda in 19981. The Ug99 race group (TTKSK and 13 variants) and some non-Ug99 races such as TRTTF, TTTTF, and TKTTF continue to spread across different countries and carry virulence to widely deployed stem rust resistance (Sr) genes1,2,3,4,5,6,7. To mitigate these threats at global and local levels, it is important to genetically characterize known resistance resources while also searching for new resistance. Globally, scientists have been exploring the gene pools of wheat and its wild ancestors to identify novel and effective resistance resources against stem rust.

Wheat line Tr129 carries resistance to both leaf rust and stem rust8,9. This line was reportedly developed by hybridizing the hexaploid common wheat cultivar Marquis (2n = 6x = 42, AABBDD) with the tetraploid species Aegilops triuncialis (2n = 4x = 28, CCUU) accession CN 344058. An initial inheritance study conducted by Fetch and Zegeye (2009) found a single dominant gene in Tr129 conditioning resistance to race TPMKC. Further testing indicated the presence of two Sr genes in Tr129 that conferred resistance to Pgt race MCCFC10. Preliminary mapping suggested that these two genes were located on chromosome arms 2BL and 6AS, but the locations were not validated11. The objectives of this study were to determine the origin of stem rust resistance in the wheat line Tr129, determine the chromosomal locations of Sr genes present in Tr129, and identify the Sr genes as known or novel.

Results

GISH analysis

Genomic in situ hybridization (GISH) with probes produced from DNA of Ae. umbellulata, Ae. caudata or Ae. triuncialis did not detect Ae. triuncialis chromatin in Tr129. The control lines showed positive GISH signals as expected, indicating that an Sr gene in Tr129 was not derived from Aegilops triuncialis (2n = 4x = 28, CCUU) as reported by Aung & Kerber8 or that a translocation segment, if present, was too small to be detected by GISH (Supplementary Fig. S1). Moreover, SSR genotyping of the chromosome region flanking a resistance gene in chromosome 2B (Supplementary table S1) showed that the haplotype was more representative of Neepawa (postulated genes Sr5, Sr7a, Sr9b, Sr12, Sr1612) than Marquis (Sr7b). This alone indicated that the published pedigree of TR129 (Marquis*6/3/Marquis/CN 34,405//2*Marquis)8 was incorrect.

Phenotyping

Wheat line Tr129 showed resistance to Pgt races A (TTKSK), B (RRTTF), C (QTHJF), D (TMRTF), E (TPMKC), and F (MCCFC) (Table 1) with infection types (IT) 2, 2−, 2−, 11+, 22+, and 1− respectively. RL6071 was susceptible to all Pgt races used in the current studies with IT ranging from 3+ to 4. The RL6071/Tr129 F2 population segregated 57 resistant (R): 25 susceptible (S) plants to race A, fitting an expected 3 resistant: 1 susceptible ratio for a single dominant gene (χ2 = 1.31, P = 0.25). Segregation data for the entire (or near entire) doubled haploid (DH) population with three races (A, E, and F) allowed postulation (Table 1) of genes Sr8a, Sr9b and temporarily name SrTR129 either singly or in combination based on avirulence/virulence attributes of each race. Although the result with race E showed a significant overabundance of susceptible lines, genetic mapping (below) confirmed that a single gene was involved. Gene SrTr129 showed close similarity to SrCad, Sr42, and SrTmp13,14,15.

Table 1 Phenotypic ratios, chi squared analysis and Sr gene postulation for response to six Puccinia graminis races tested on the RL6071/Tr129 DH population.
Full size table

Subsets of the DH population were selected to confirm the postulations (Table 1, Supplementary table S2). Gene SrTr129 conferred resistance to races A, B, and C; Sr8a conferred resistance to F, D, and B; Sr9b conferred resistance to races E and F; and Sr7b conferred resistance to race C (Table 2).

Table 2 Phenotypic ratios, chi squared analysis and Sr gene postulation for response to four Puccinia graminis races tested on the RL6071/Peace DH population.
Full size table

Genotyping and mapping

Genotyping the F2 population with the 90K iSelect SNP array showed linkage between the race A resistance gene and markers on chromosome arm 6DS in the region associated with SrCad, Sr42, and SrTmp13,14,15. A 6DS linkage map spanning an 8.47 centimorgan (cM) genetic region was developed for the DH population using 13 KASP, two SSR, and FSD_RSA markers. Gene SrTr129 conferred resistance to race A and was located to position 7.73 cM on the linkage map. It was flanked by markers gpw5182 and kwm71 and co-segregated with 12 KASP markers (Fig. 1), 11 of which were previously used to map SrTmp, SrCad, and Sr4213,14. Fourteen KASP markers were used to compare haplotypes in the region of the chromosome arm 6DS carrying race A resistance in wheat cultivars/lines Triumph 64 (SrTmp), Tr129 (SrTr129), Peace (SrCad), and Norin 40 (Sr42). All four haplotypes were unique, and Tr129 was differentiated from Peace, Norin 40, and Triumph 64 by three, 12, and nine SNP markers, respectively (Table 3). Previous data14 indicated SrTr129 conferred resistance specificity that differed from Sr42 and SrTmp, thus a comparison with SrCad was needed. Phenotyping the RL6071/Peace DH population revealed single gene segregation for Pgt races D, E, and F (Table 2). All three races virulent to SrCad in this population were also virulent to SrTr129 as shown above. One Sr gene in Peace conferred resistance to races E and F, whereas independent Sr genes conferred resistance to races A and D (Supplementary table S3). SrTr129 and SrCad both conferred resistance to races A, B, and C. The races used in this study detected no difference between SrTr129 and SrCad.

Figure 1
figure 1

Comparison of the SrCad genomic region in four mapping populations: LMPG-6S/Triumph 64 (a), RL6071/Tr129 (b), RL6071/Peace (c), and LMPG-6S/Norin 40 (d). Markers shown in dark blue font were mapped in more than one mapping population and resistance genes are shown in dark orange font. Mapping distances are in centimorgans (cM).

Full size image
Table 3 SNP marker haplotypes of wheat lines Triumph 64, Tr129, Peace, and Norin 40 with genes for stem rust resistance on chromosome arm 6DS and susceptible lines LMPG and RL6071.
Full size table

A 19.6 cM partial linkage map for chromosome arm 6AS was constructed with nine KASP markers to map the postulated Sr8a effective against races B, D, and F (Fig. 2). This gene was distal to KASP markers kwm53 and kwm54 previously reported to be closely linked with the Sr8 locus16. The 90K iSelect SNP array was also used to genotype the Subset 2 DH lines that lacked SrTr129 and identified the postulated Sr7b and Sr9b genes on chromosome arms 4AL and 2BL, respectively. The linkage map of chromosome arm 4AL located Sr7b, which conferred resistance only to race C. This gene was mapped to position 63.1 cM on the 4AL linkage map and was flanked by SNP markers IWB47901 and IWB24693, located in the region previously associated with Sr717 (Fig. 3a, Supplementary table S4). Phenotyping of parents and differential wheat lines carrying the Sr7a and Sr7b alleles showed that Sr7a was ineffective against race C (Table 4, Supplementary Fig. S2), whereas Tr129 and near-isogenic line ISr7b-Ra (CI 14,165) were resistant (IT 11 + and 22-, respectively). SNP marker data used to map the postulated Sr9b on chromosome arm 2BL that was effective against races E and F was consistent with the region known to carry Sr9 (Fig. 3b, Supplementary table S5). The Sr9 region has seven resistance alleles (Sr9a, Sr9b, Sr9d, Sr9e, Sr9f., Sr9g, and Sr9h)18,19,20,21,22,23, of which Sr9a and Sr9b are effective against races E and F (Fig. 3b). Since race C is virulent for Sr9b and avirulent for Sr9a, and the Subset 2 lines lacking resistance to race A (Table 1) segregated for a single gene located on 4B, that gene is likely Sr9b.

Figure 2
figure 2

Comparison of Sr8a linkage maps from RL6071/Tr129 DH population from the present study (a) and the LMPG-6S/Harvest DH population from Hiebert et al.16 (b). Common markers between two linkage maps are in dark blue font, and purple font was used for the postulated gene. Distances between loci are in centimorgans (cM).

Full size image
Figure 3
figure 3

Linkage maps developed for Subset 2 DH lines. (a) Sr7b mapped at 63.1 cM on chromosome arm 4AL; (b) Sr9b mapped at 79.2 cM on chromosome arm 2BL. Kompetitive allele specific (KASP) PCR markers designated as kwh developed in the present study are shown on the linkage maps. Primer information is provided in supplementary table S7.

Full size image
Table 4 Stem rust reactions of parents RL6071 and Tr129, and five wheat lines carrying Sr7a or Sr7b alleles when tested with Pgt race C.
Full size table

Discussion

Line Tr129 was previously reported to carry stem rust resistance derived from Ae. triuncialis8. However, negative GISH results in this study revealed the absence of detectable Ae. triuncialis chromatin in Tr129 (Supplementary Fig. S1). In addition, no large linkage block normally associated with alien translocations was associated with any of the resistance genes. We concluded that the stem rust resistance in line Tr129 was not derived from Ae. triuncialis. Moreover, a genetic haplotype in the Sr9 region similar to that of Neepawa in Tr129 did not support the reported pedigree with Marquis as the recurrent parent (Supplemental Table S1). Thus, the origin and pedigree of Tr129 is unknown.

Genetic analysis of stem rust resistance in Tr129 revealed the presence of four genes located on chromosome arms 6DS, 6AS, 4AL and 2BL. Theses genes include temporarily named SrTr129, Sr8a, Sr7b, and Sr9b, respectively. SrTr129 conferring resistance to race A (TTKSK or Ug99, Table 1) mapped to a region known to carry resistance genes SrCad, Sr42, and SrTmp13,14,15. A comparison of linkage maps for SrTr129, SrCad, Sr42, and SrTmp showed collinearity and consistency in the Sr gene position on chromosome arm 6DS (Fig. 1). Since SrTr129 conferred resistance to race C (Table 1), and Norin 40 with Sr42 was susceptible to race C15, SrTr129 cannot be Sr42. A line with SrTmp was susceptible to race TRTTF (06YEM34-1)14 whereas Tr129 was resistant (data not shown). SrTr129 showed resistance to related race C (Table 1) indicating that SrTmp and SrTr129 were alleles or located at different closely linked positions. Peace and Tr129 showed the same pattern of response against multiple Pgt races (data not presented) but since Tr129 carried four resistance genes we could not differentiate the response arrays conferred by SrCad and SrTr129. We then tested 73 random DH lines from a RL6071/Peace population used to map SrCad24 with races D, E, and F using the methods listed previously. Many lines resistant to race A (SrCad) were susceptible to races D, E, or F (Table 2, Supplementary table S3). It appears that Peace also has multiple Sr genes that need to be confirmed in future studies. Thus, SrTr129 could be SrCad previously described in AC Cadillac and Peace24. Although we could not differentiate SrTr129 from SrCad using several Pgt races, a comparison of SNP haplotypes for chromosome arm 6DS showed a unique haplotype for Tr129 compared to SrCad, SrTmp, and Sr42. However, cloning of these genes will clear the ambiguity of whether these are alleles or located at closely linked loci.

The RL6071/Tr129 DH population also segregated for a single gene that conferred resistance to race D and mapped to the chromosome arm 6AS region known to carry Sr8 (Fig. 2). Collinearity between chromosome 6AS linkage maps in RL6071/Tr129 and LMPG-6S/Harvest DH populations16 support the hypothesis that resistance was conferred by an allele of Sr8 (Fig. 2). Three resistance alleles (Sr8a, Sr8b, and Sr8155B1) have been reported at the Sr8 locus25,26. Sr8b was excluded as a candidate since race D is virulent (Fetch, unpublished data). Both Sr8a and Sr8155B1 conferred resistance to TRTTF26; the IT 2- recorded in this study matched the expected response for Sr8a and not the IT 0; documented for Sr8155B1, hence the latter was excluded as a candidate.

DH lines in Subset 1 inoculated with Pgt race C segregated 3:1 indicating that Tr129 carries two resistance genes effective against this race (Table 1). SrTr129 was one of these genes (Supplementary table S2), and data from Subset 2 (lines that lacked SrTr129) allowed mapping of another gene for race C resistance to chromosome arm 4AL (Fig. 3a). This gene was flanked by SNP markers IWB47901 and IWB24693 at 137.3 and 166.7 cM, respectively17(Supplementary table S4). The consensus map positions of Sr7a-linked STARP markers Xrwgsnp10 and Xrwgsnp1127 and SNP marker IWA106728 are close to the Sr7 locus (Supplementary table S6). The locations of Sr7 and SrND643 were reported on chromosome arm 4AL29. Gene SrND643 was excluded as a candidate as it was known to confer resistance to race A29. Since Sr7a is ineffective against race C it is likely that allele present in Tr129 is Sr7b (Table 4).

To test for the presence of Sr9b in Tr129, the phenotypic data for Subset 2 DH lines (lacking SrTr129) tested with Pgt races E and F was used. Although the chi squared analysis did not fit a single gene ratio for response to race E, QTL analysis indicated that only the 2BL genomic region was involved. Six Sr genes (Sr9, Sr16, Sr20, Sr28, Sr47, and Sr883-2B have been reported on 2BL30,31. As race E is virulent for Sr16, Sr20, and Sr28 these genes were eliminated as candidates. Genes Sr47 and Sr883-2B can also be eliminated as they confer resistance to race A. Thus, Sr9b was the only candidate allele that matched the specificity of the gene on chromosome arm 2BL in Tr129.

In conclusion, we could not identify the origin of Sr resistance in Tr129, but it was not derived from Ae. triuncialis. Tr129 also differed from Peace as that cultivar does not carry Sr8a (Hiebert, unpublished data). By using a mapping population, subsets of the population to remove confounding effects of SrTr129, and several Pgt races with differing virulence, we were able to identify and map stem rust resistance genes present in Tr129. Tr129 has at least four Sr genes: SrTr129 located on chromosome arm 6DS, Sr8a on 2BL, Sr7b on 4AL, and Sr9b on 2BL. Gene SrTr129 conferred resistance to races A, B, and C, Sr8a conferred resistance to races B, D, and F, Sr7b conferred resistance to race C, and Sr9b conferred resistance to races E and F. Two Sr genes in Tr129 conferring resistance to race F were reported by Ghazvini et al.10 (Supplementary table S2). Genes Sr7b and Sr9b are very common in Canadian wheat cultivars and provide moderate protection against individual races in the North American Pgt population. Gene Sr8a is a common source of resistance to race B in Canadian wheat cultivars16. Although SrTr129 could not be differentiated from SrCad, Tr129 had a unique SNP haplotype in the SrCad genomic region and therefore could represent a new source of resistance effective against the widely virulent Ug99 lineage of Pgt races. Insight into the effectiveness of SrCad to virulent exotic races and ineffectiveness against some North American races will guide breeders on how to utilize SrCad in breeding for stem rust resistance.

Material and methods

Plant material

Tr129 is a hexaploid stem rust resistant wheat line bred by Dr. T. Aung (AAFC, retired). Wheat line RL6071 (Prelude/8*Marquis*2/3/Prelude//Prelude/8*Marquis) is a hexaploid wheat line developed by Dr. P. L. Dyck (AAFC) and was selected as a susceptible parent for genetic studies; it does not carry Sr7b, which is present in Marquis. We used F2 and doubled haploid (DH) populations developed32 from the cross RL6071/Tr129 to study the inheritance of stem rust resistance in Tr129 and for genetic mapping.

Genomic in situ hybridization (GISH)

Genomic DNA of tetraploid Aegilops triuncialis L. (genome UUCC, 2n = 28) TA1752 and diploid accessions Aegilops umbellulata Zhuk. (genome UU, 2n = 14; TA1851) and Ae. caudata L. (genome CC, 2n = 14; TA1908) representing different subgenomes were used to prepare GISH probes. Chromosomal preparations of chromosome addition lines TA7562 (DA1U, 2n = 44) and 99-247-4 (DA3CtL, 2n = 44) were used as controls in GISH experiments. Accessions TA1752, TA1851, TA1908 and TA7562 are maintained by the Wheat Genetics Resource Center at Kansas State University. GISH was performed according to Zhang et al.33 with modifications described in Liu et al.34. Chromosome preparations were mounted and counterstained with propidium iodide (PI) in Vectashield (Vector Laboratories, Burlingham, CA, cat # H-1300). Images were captured with a Zeiss Axioplan 2 microscope using a cooled charge-coupled device camera CoolSNAP HQ2 (Photometrics, Tucson, AZ) and AxioVision 4.8 software (Zeiss). Images were processed using Adobe Photoshop software (Adobe Systems Incorporated, San Jose, CA, USA).

Stem rust assays

All stem rust assays were performed on seedlings with fully emerged first leaves. Plants pre- and post-inoculation were grown in a greenhouse at 20 ± 2 °C with a 16 h photoperiod. Inoculation, incubation, and rating of stem rust infection types were performed as described by Hiebert et al.24. Briefly, Pgt urediniospores suspended in light mineral oil (4 mg per 0.7 ml oil) were sprayed onto seedlings. After allowing the oil to evaporate, seedlings were incubated overnight in darkness at 100% relative humidity and allowed to dry slowly under light before removal to greenhouse benches. Seedlings were rated for infection type (IT) 14 days post-inoculation following the 0–4 scale described by Stakman et al.35 and modified by Roelfs and Martens36. Six races with wide differences in virulence based on the Pgt letter-code nomenclature of Roelfs and Martens36 were used in the study to differentiate individual Sr genes. DH lines with ITs 0–2+ were classified resistant and ITs 3–4 were classified as susceptible. The parents (RL6071 and Tr129), F2 population (n = 85), and DH population (n =  ~ 276) were screened with Pgt races A (TTKSK accession SA31), E (TPMKC isolate W1373), and F (MCCFC isolate W1541). For QTL analysis, ITs were converted in to a linearized 0–9 scale37. Subset 1 of the DH population was randomly selected and phenotyped with Pgt races B (RRTTF isolate 10PAK05-1), C (QTHJF isolate W1347), and D (TMRTF isolate W1311). To better resolve the four Sr genes that were detected by multi-race testing with six Pgt races listed above, an additional subset (Subset 2) of DH lines lacking resistance to race A was phenotyped with race C. For the DH population or subsets of the population, three to five seedlings were rated per line for each Pgt race.

A multi-pathogen test was conducted with Pgt races D, E, and F to compare the Tr129-derived Sr gene on chromosome arm 6DS with SrCad. This analysis was performed on 73 randomly selected DH lines from a previously developed population from cross RL6071/Peace (BW90*3/BW553//BW90’S’/Katepwa)24. This population had already been phenotyped with race A. To differentiate between alleles Sr7a and Sr7b, four genotypes carrying Sr7a were phenotyped with the Sr7a-virulent Pgt race C along with Tr129, RL6071, and wheat line ISr7b-Ra (CI 14165)36 (Table 4).

Genotyping and mapping

DNA was extracted from parent lines (RL6071 and Tr129) and F2 and DH progeny using a modified ammonium acetate method38. The F2 (n = 85) plants and parents were genotyped with the 90 K iSelect SNP array17. Linkage maps were constructed using MapDisto 1.8.2 (http://mapdisto.free.fr)39 by setting logarithm of odds (LOD) and Rmax values 3.0 and 0.3, respectively. Genetic distances were calculated with the Kosambi mapping function40. After initial mapping of the Sr gene that conferred resistance to Pgt race A to chromosome arm 6DS in the F2 population, further analyses were undertaken using the DH population (n = 276). SrCad region-specific kompetitive allele specific PCR (KASP) markers on chromosome arm 6DS for genotyping were selected from Kassa et al.13 and analysis of those markers followed the same publication13. In addition, two simple sequence repeat (SSR) markers (cfd49 and gpw5182) and common bunt resistance gene (Bt-10) PCR marker FSD_RSA41 were also included in the chromosome 6DS linkage map. Genotyping of the DH population with marker FSD_RSA followed procedures described by Hiebert et al.24. SSR genotyping was done following described procedures42,43. PCR products were analysed by using an ABI 3100 genetic analyzer (Applied Biosystems, Streetsville, ON, Canada). Wheat lines Triumph 64, Tr129, Peace, and Norin 40 were haplotyped with 14 KASP markers (kwm112, kwm191, kwm196, kwm197, kwm244, kwm871, kwm873, kwm918, kwm929, kwm987, kwm994, kwm997, kwm999, and kwm1000) described by Kassa et al.13 to characterize the region spanning SrCad in chromosome arm 6DS.

To map additional Sr genes present in the Tr129 line, further genetic analysis was done by using two different subsets of the DH population as explained earlier. Subset 1 contained randomly selected DH lines, and Subset 2 consisted of DH lines lacking resistance for race A. DH lines from Subset 1 were used to map two additional Sr genes present in Tr129. Subset 2 was used to map a race E-specific resistance gene from Tr129. Subset 2 was also phenotyped with the race C to support the genotypic analysis conducted on Subset 1.

For mapping the Sr gene effective against Pgt races B, D, and F, we first considered chromosome arm 6AS based on the initial analysis done by Ghazvini et al.11 with race F. To develop a chromosome 6AS linkage map for the Tr129 DH population, nine KASP markers (kwh53, kwh54, kwh58, kwh62, kwh63, kwh223, kwh225, kwh227, and kwh233) were selected from Hiebert et al.16 (Supplementary table S7). The 6AS linkage map was developed with these nine KASP markers, and phenotypic data from races B, D, and F was used to map an additional Sr gene derived from Tr129.

90 K iSelect genotyping was performed on Subset 2 DH lines to determine the chromosomal locations of two additional Sr genes in Tr129. QTL analysis was carried out using QGENE 4.4.0 (https://www.qgene.org/)44 and single-trait multiple interval mapping (MIM) was used to detect genomic regions associated with specific races45. After initial QTL detection, the phenotypes for races C and E were mapped as Mendelian traits. Mapchart 2.32 (https://www.wur.nl/en/show/mapchart.htm)46 was used to develop the linkage map figures. After locating the Sr gene on chromosome arm 4AL, polymorphic iSelect SNPs were converted to KASP markers (Supplementary table S7). An additional KASP marker (kwh703) was developed from the Sr7 region-associated SNP-based semi-thermal asymmetric reverse PCR (STARP) marker Xrwgsnp11 from Saini et al.27 (Supplementary table S7). An Sr gene effective against races E and F was identified on chromosome arm 2BL, and 21 KASP markers were developed from polymorphic iSelect SNPs identified on chromosome 2B. DNA marker genotyping and linkage mapping for all Sr genes were performed according to the procedure described above. An additional genotypic analysis was done on the parental lines Tr129 and RL6071, as well as Neepawa and Marquis, with eight SSR markers (Supplementary table S1) specific for the Sr9 region23,43,47.

Research involving plants

All field experiments were in compliance with Institutional, National and International guideline policies.