Resistance to wheat rusts identified in wheat/Amblyopyrum muticum chromosome introgressions

Abstract Wheat (Triticum aestivum L.) rusts are a worldwide production problem. Plant breeders have used genetic resistance to combat these fungi. However, single‐gene resistance is rapidly overcome as a result of frequent occurrence of new virulent fungal strains. Thus, a supply of new resistance sources is continually needed, and new resistance sources are limited within hexaploid wheat genetic stocks. Wild relatives are able to be a resource for new resistance genes but are hindered because of chromosome incapability with domesticated wheats. Twenty‐eight double‐haploid hexaploid wheat/Amblyopyrum muticum (Boiss.) Eig introgression lines, with introgressions covering the majority of the T genome, were evaluated for resistance to Puccinia triticina Erikss., P. graminis Pers.:Pers. f.sp. tritici Erikss. & E. Henning, and P. striiformis Westend. f.sp. tritici Erikss.. At the seedling level, four lines were resistant to races of P. triticina, six lines were resistant to P. graminis, and 15 lines were resistant to P. striiformis. At the adult stage, 16 lines were resistant to P. triticina. Line 355 had resistance to all three rusts and line 161 had resistance to all tested races of P. triticina. Some of these lines will require further work to reduce the size of the introgressed segment; however, lines 92 and 355 have very small fragments and can be used directly as new resistance donors.


INTRODUCTION
Wheat rusts are the most formidable group of pathogens influencing production. Three species, P. triticina, P. graminis f.sp. tritici, and P. striiformis f.sp. tritici, are the causal fungi for Crop Science broad-spectrum resistance genes, such as the pleiotropic gene Lr34/Yr18/Sr57, providing various levels of resistance to all three rusts (Lagudah et al., 2006). Single-gene resistance is often effective during the first few years after a variety release. However, large acreages of genetically similar cultivars lead to selection of rust genome mutations that can overcome resistance, resulting in the emergence of new virulent races each year. One of the most notable examples is the P. graminis f. sp. tritici race Ug99, which appeared in Uganda in 1998 and is virulent on most commercial varieties (Pretorius, Singh, Wagoire, & Payne, 2000). Puccinia striiformis has historically been found in cooler climates such as the U.S. Pacific Northwest; however, new races adapted to higher temperatures began appearing in the U.S. Great Plains in 2000 (Milus, Kristensen, & Hovmøller, 2009). Puccinia triticina is found in all world regions of wheat production and >70 different races are found each year in North America alone (Kolmer, 2019;Kolmer, Long, & Hughes, 2005).
Rust resistance breeding is difficult because of the limited available sources of genetic variation within the gene pool of wheat. As a result, geneticists have turned to the wild relatives. The effort required to transfer interspecifc variation into wheat depends on the closeness of the wheat-wild relative relationship. For example, transfer from closely related species Aegilops tauschii Coss., the D genome progenitor, and Triticum monococcum L., one of the A genome progenitors of wheat, can be achieved through crossing elite germplasm with synthetic wheats followed by backcrossing. Transfer of wild accession genes occurs through normal chromosome pairing and recombination (Cox, 1998;Warburton et al., 2006;Zohary, Harlan, & Vardi, 1969). Linkage of target variation with undesirable traits can be resolved through further recombination with selection. The transfer of genetic variation from wild relatives carrying related but not identical genomes can be achieved using strategies like mutagenesis (Sears, 1977(Sears, , 1993 or the removal of the Ph1 locus, which restricts recombination to homologous chromosomes (Al-Kaff et al., 2008).
Amblyopyrum muticum (Aegilops mutica Boiss; 2n = 2x = 14; genome TT) is a wild relative of wheat originating from the Middle East and Armenia. Am. muticum possesses a Ph1 suppressor gene that facilitates recombination between Am. muticum chromosomes and the homoeologous chromosomes of wheat (Dover & Riley, 1972). Relatively little research has previously been undertaken on Am. muticum, but merit as a genetic resource has been shown. Lines homozygous for a 5D/5T introgression provide winter hardiness (Iefimenko, Fedak, Antonyuk, & Ternovska, 2015), while a complete 7T chromosome substitution line provides resistance to powdery mildew (Blumeria graminis f. sp. tritici; Eser, 1998). With the development of new marker systems and higher throughput karyotyping, the majority of the Am. muticum genome has now been introgressed into hexaploid wheat (King et al., 2017. In this report, we begin the characterization of the first 28 lines containing overlapping segments of the T genome by screening for resistance to cereal rusts.

Introgression lines
The development, and characterization of the double-haploid wheat/Am. muticum lines are described in King et al. (2017King et al. ( , 2019 and in Supplemental Table S1. In summary, two Am. muticum accessions, JIC2130004 and JIC2130012, were obtained from the Germplasm Resource Unit (John Innes Centre, Norwich, UK) and initially crossed to either 'Chinese Spring' (CI6223) or 'Pavon 76' (PI519847), The F 1 products were crossed to either 'Paragon' (Cereal Variety Handbook, 1997) or Pavon 76, then backcrossed to Paragon to the BC 3 . At the BC 3 generation, double haploids were produced to stabilize the segments using maize pollination, embryo rescue, and chromosome doubling .

Seedling infection testing
Five seeds from each of the 28 doubled-haploid introgression lines were planted in 20-by 20-by 3-cm aluminum cake pans Crop Science containing BM-1 soil media (Berger Peat Moss, LLC). Five seeds of the hexaploid parents Paragon and Pavon 76 along with a susceptible check 'Thatcher' were also included and placed throughout the pan. At the two-to three-leaf seedling stage, 25 mg of P. triticina rust spores from each of the races BBBD, TNRJ (PRTUS35), and 97AZ103 were suspended in 2 ml of Soltrol 170 parafin oil (Phillips Petroleum) and sprayed onto the seedlings with an atomizer (Tallgrass Solutions) at 40 psi. Seedlings were transferred to a Percival humidity chamber at 100% relative humidity with wall settings of 5 • C and water basin of 40 • C for 16 h. Plants were transferred back into the greenhouse. At 14 d post inoculation (dpi), infection types were rated on a scale of 0-4 (McIntosh, Wellings, & Park, 1995): 0, being no infection; a semicolon (;,fleck) being small focused infection points caused by hypersensitive reaction with no pustule formation; 1-4 scale of pustule size with 1 being very small and focused with distinct hypersensitive reaction to 4 being large pustules and no plant defense reaction.
A second set of plants was inoculated with a composite of two P. striiformis tritici isolates collected in Kansas in 2012 and 2014 (Robert Bowden, personal communication, 2018). Plants were inoculated as above but humidity chamber conditions were wall settings of 0 • C and water basin of 29 • C for 24 h. At 14 dpi, plants were scored using a scale of 1 (resistant) to 9 (susceptible) (McNeal, Konzak, Smith, Tate, & Russell, 1971).
A third set of seedlings were inoculated with a composite of two races of P. graminis tritici, RKQQC and QFCFC (obtained from Robert Bowden in 2018), as above except humidity chamber conditions were with wall settings of 6 • C and water basin of 42 • C for 16 h. At 14 dpi, infection types were scored on a scale of 0, ;, 1-4 (see description for leaf rust; McIntosh et al., 1995). Seedlings were also tested at the USDA-ARS Cereal Disease Laboratory (St. Paul, MN) BL-3 containment facility with P. graminis tritici Ug99 Race TTKSK (isolate 04KEN156/04) as described in Hundie et al. (2019).

Adult infection testing
Two plants from each line including Paragon, Pavon 76, and the susceptible check Thatcher were grown in a 4-L pot containing BM-1 soil media in the greenhouse as described above. At spike emergence, plants were inoculated with a field composite of P. triticina collected in 2017 from Thatcher growing in Alfalfa County, OK, and transferred to a humidity chamber as described previously. At 21 dpi, flag leaves were scored for infection using percentage of leaf coverage and infection type (McIntosh et al., 1995).

Lr34 and Lr46 screening
Twenty-one lines were tested for the presence of Lr46 and Lr34 using Kompetitive allele-specific polymerase chain reaction (KASP) markers. Fifty to 100 ng of DNA was dried in a 384-well plate and resuspended using 3 μl of KASP assay master mix containing primer, 2× KASP buffer and enzyme (Keygene), and MgCl 2 . For Lr46, the primer used was Lr46-Yr29_JF2-2-KASP (Brown-Guedira, G. and Fellers, J.P., unpublished), and for Lr34, two separate primers Lr34Exon11-KASP (Lagudah et al., 2009) andLr34JagExon22-KASP (Yan, L., personal communication, 2016) were used to check for the Lr34 gene and also confirm there was not a false-positive gene. For Lr34exon11 and Lr46, the thermal cycler conditions used were as follows: 94 • C for 15 min, 97 • C 15 sec, 68 • C for 1 min at −0.8 • C per cycle for 10 cycles, 97 • C for 20 sec, 60 • C for 1 min for 30 cycles, 59 • C for 30 sec at −1.0 • C per cycle for 34 cycles, and a final cool down of 10 • C for 5 min. For Lr34JagExon22, the thermal cycler conditions were as follows: 94 • C for 15 min, 94 • C for 20 sec, 60 • C for 1 min for 30 cycles, and a cool down for 5 min at 10 • C. Reactions were run on a GeneAmp Systems 9700 (Applied Biosystems) and results were processed using Klustercaller (LGC Biosearch Technologies).

RESULTS
Three different races of P. triticina were used to screen seedlings of the introgression lines. Race 1 BBBD, the most avirulent race available with virulence only to Lr14a, Lr14b, Lr20, and Lr50, was used to identify any resistance contained in the lines. Only lines 28, 161, and 355 were resistant with infection type (IT) ; (fleck) meaning a very small, hypersensitive-like point on the leaf with no pustule formation ( Figure 1A; Table 1). Isolate 97AZ103A, originally collected from durum wheat in Arizona, is virulent on Lr2c, Lr3, LrB, Lr10, Lr14b, and Lr20 and represents one of the most virulent races on U.S. durum (tetraploid) wheat. Lines 161 and 355 were resistant to 97AZ103A ( Figure 1B; Table 1).
Resistance phenotypes to P. striiformis tritici have larger lesions and use a different rating scale based on lesion length and number of pustules. The composite used to inoculate the seedlings represents the current Great Plains field population and is most notable for overcoming Yr17 and the unknown 'TAM111' Yr gene within quantitative trait loci QYr.tamu-2B . Paragon appeared to have low level of resistance to this composite ( Figure 3) and was scored as 7 out of 9 (Table 1). Fifteen of the introgression lines were scored with high to medium levels of resistance ranging from 1 to 5. Lines 16, 27, 29, 92, and 121 had the highest level of resistance (Table 1; Figure 3).

DISCUSSION
The use of wheat relatives in crop improvement has been technologically limited by an inability to quickly identify and characterize introgressions. The assumed low levels of recombination between the chromosomes of wheat and those of the wild relatives have also prevented adoption of this source of variation. However, because of the presence of a Ph1 inhibitor in Am. muticum and the application of new marker technology, over 200 introgressions have been generated between wheat and Am. muticum (Dover & Riley, 1972;King et al., 2017, and2019;Grewal et al., 2019). Twenty-four of the 28 lines evaluated here carry single introgressions with the remaining four Crop Science T A B L E 1 Disease reactions of bread wheat/Amblyopyrum muticum introgression double-haploid lines after infection with Puccinia triticina (leaf), P. graminis tritici (stem), and P. striiformis tritici (stripe). Resistance was evaluated at the seedling and adult stages. Lines were also tested for the presence of resistance gene Lr34  Paragon  3  3  3  1-2C  4  7 ins/ins a 0, ;, 1, 2, 3, where 0 is resistant (R) and 3 is susceptible (S). b 0R-90S, adult flag leaf based on percentage leaf coverage; LTN, leaf tip necrosis; C, chlorosis; N, necrosis; MR, medium resistance; Z, more pustules at the base, fewer at the tip (McIntosh et al., 1995). c For seedlings, 0, ;, 1-4, where 0 is resistant and 4 is susceptible (McIntosh et al., 1995). d 1-9, where 1 is resistant and 9 is susceptible (McNeal et al., 1971). e NT, not tested; ins, insertion; del, deletion.
carrying either two or three. The introgressions in two of the 28 lines are whole chromosomes or very large chromosome segments from the T genome but nine are small to telomeric in size (see karyotypes in King et al., 2019). The smaller introgressions are likely to carry fewer deleterious genes and thus can be more rapidly integrated into a breeding program. Larger introgressions, however, frequently need to be reduced in size in order to minimize linkage drag. This can be done via the strategy of overlapping introgressions (Sears, 1956) or via recombination with the B and D genomes (Glémin et al., 2019). In this report, we begin assessing whether the T genome has useful resistance to the three cereal rusts. Useful resistance was found to all three cereal rusts within the introgression lines with some of the resistance because of new resistance genes. Line 355 had seedling resistance to the less virulent P. triticina races and adult resistance to the field composite but was susceptible to TNRJ. Line 355 was also resistant to Great Plains isolates of P. graminis f. sp. tritici   (McIntosh et al., 1995) but not Ug99. Line 355 has a small 1T fragment and because of its resistances may be useful for durum and bread wheat improvement. Line 161 had very good P. triticina resistance at both seedling and adult stages but contains a large segment of 1T. One Lr gene may be shared with Line 355, but line 161 may also have a second gene that provides resistance to TNRJ. The Am. muticum lines also provided new resistance to P. graminis and P. striiformis. Lines 29 and 92 were highly resistant to both species, with 29 also resistant to P. graminis Ug99 race TTKSK. No stem rust resistance gene derived from Am. muticum has been previously described, therefore this resistance is new. Line 29 contains a whole 7T chromosome and thus will need to be reduced in size before it can be used in a breeding program. Unfortunately, the lines in this study with large or small recombined segments derived from 7T did not exhibit resistance to Ug99, suggesting that they are missing the fragment of the whole chromosome with the resistance gene. Line 92 contains a very small introgression from 5T .

Crop Science
Several of the introgression lines tested were produced from the same original BC 3 plants (Supplemental Table  S1) and were therefore expected to contain the same segments. Indeed, the molecular and cytogenetic characterization appeared to confirm this; however, some of the lines tested containing the same segments did not give the same resistance results. The most notable of these are lines 124 and 355 and line 202 compared with lines 191,192,195,196,and 198. Line 124 was susceptible to all races tested while, as outlined above, line 355 shows useful resistance to leaf and stem rust. Lines 191,192,195,196,and 198 all show adult resistance to P. triticina, while line 202 was susceptible. There are a number of possible explanations for these apparently conflicting results. Firstly, neither the molecular nor cytogenetic characterization would have revealed small differences in the size of the segments. The markers used for the characterization of the lines were part of the Axiom Wheat Wild Relative Array  and while they give good coverage of the chromosomes, gaps do exist. Secondly, additional very small segments might have been present in some lines but not detected. It has become clear that the level of recombination seen in the wheat/Am. muticum introgression lines is very extensive and the recombination occurs in the gametes of all generations and not just the F 1 hybrids as originally expected.
The hexaploid heritage is also contributing some adult plant resistance in the introgression lines. Chinese Spring has previously been shown to contain Lr34 (McIntosh et al., 2008). Line 17 exhibited LTN and was indeed positive for the active Lr34 allele. Chinese Spring was the first cross to Am. muticum for several of the introgression lines and therefore likely to be the source of Lr34 in line 17. However, it was not maintained in the other lines developed from Chinese Spring, that is, lines 15, 16, 19, 20, and 21. Pavon76, Paragon, and 24 of the introgression lines were also found to be heterozygous for the Lr46 allele. However, the heterozygous result for the introgression lines probably is due to the functionality of the marker as these lines were produced via a doubled-haploid procedure and were therefore expected to be homozygous at  all loci. The races used for all three pathogens were all virulent on Pavon76 and Paragon with the exception of the P. triticina field composite, which exhibited an adult Lr12-like reaction on Paragon (McIntosh et al., 1995). Paragon is listed as having moderate resistance to P. triticina and P. graminis without specific resistance gene designations (Cereals Variety Handbook, 1997). Lr12 is known to be present in Chinese Spring, but introgression lines without Chinese Spring in their parentage also expressed this phenotype, suggesting Paragon may be the source (McIntosh et al., 2008). The infection phenotype of lines 19, 86, 161, 192, 195, 196, 198, and 355 was different than that of Paragon suggesting the presence of different genes from chromosomes 1T, 2T, and 4T or 7T.
In this work we have evaluated a group of Am. muticum introgression lines as a new source for resistance to cereal rusts. The screening has identified several lines with resistance to rust races that are representative of current field populations. Five lines-86, 92, 96, 121, and 355-are now a useful source of germplasm with introgressions supported by genomic in situ hybridization (GISH)  and are supported by markers to follow the introgressions. From 237 Am. muticum-specific SNPs, 137 KASP markers have been derived that span the genome Supplemental Table S1 includes KASP markers that can be used for each line). Most importantly, two lines have small fragments that should have less linkage drag and can be useful as germplasm in breeding strategies. The germplasm is available through the John Innes Centre Germplasm Resource Unit or by request from USDA-ARS.

AUTHOR CONTRIBUTIONS
J. Fellers conceived and conducted the screening experiments, provided funding, and co-wrote the manuscript. A. Matthews and A.K. Fritz provided marker analysis and co-wrote the manuscript. M. Rouse conducted the Ug99 screening and revised the manuscript. S. Grewal, J. King, and I.P. King provided the lines, verified the introgressions with GISH and markers and co-wrote the manuscript.