Relocation of Sr48 to Chromosome 2D Using an Alternative Mapping Population and Development of a Closely Linked Marker Using Diverse Molecular Technologies

The Ug99-effective stem rust resistance gene Sr48 was mapped to chromosome 2A based on its repulsion linkage with Yr1 in an Arina/Forno recombinant inbred line (RIL) population. Attempts to identify markers closely linked to Sr48 using available genomic resources were futile. This study used an Arina/Cezanne F5:7 RIL population to identify markers closely linked with Sr48. Using the Arina/Cezanne DArTseq map, Sr48 was mapped on the short arm of chromosome 2D and it co-segregated with 12 markers. These DArTseq marker sequences were used for BlastN search to identify corresponding wheat chromosome survey sequence (CSS) contigs, and PCR-based markers were developed. Two simple sequence repeat (SSR) markers, sun590 and sun592, and two Kompetitive Allele-Specific PCR (KASP) markers were derived from the contig 2DS_5324961 that mapped distal to Sr48. Molecular cytogenetic analysis using sequential fluorescent in situ hybridization (FISH) and genomic in situ hybridization (GISH) identified a terminal translocation of chromosome 2A in chromosome 2DL of Forno. This translocation would have led to the formation of a quadrivalent involving chromosomes 2A and 2D in the Arina/Forno population, which would have exhibited pseudo-linkage between Sr48 and Yr1 in chromosome 2AL. Polymorphism of the closet marker sunKASP_239 among a set of 178 wheat genotypes suggested that this marker can be used for marker-assisted selection of Sr48.


Introduction
Common wheat (Triticum aestivum L., 2n = 6x = 42, AABBDD) is among the world's most widely grown crops, with an annual acreage of 220 million ha [1]. Its acreage is ranked second, next to rice, in developing countries. Its global trade volume surpasses all other major food crops combined [2]. It provides about one fifth of the recommended daily calories, and its contribution to protein is greater than that of maize [1].
Stem rust, also known as black rust, caused by Puccinia graminis f. sp. tritici (Pgt), has historically been a devastating disease of common wheat and durum wheat in Africa, Asia, Australia, Europe, the Middle East and New Zealand [3]. Yield losses of 20-30% from stem rust were reported during the mid-20th century in many geographical regions, such as eastern and central Europe [4], Asia, Australia, Europe and the United States [3]. The favorable environment for disease development and cultivation of susceptible cultivars has the potential to lead to destruction of the whole crop within three weeks of disease onset [3,5].
The detection of Pgt pathotype Ug99 (TTKSK) in Uganda in 1998 [6] rendered a large proportion (90%) of global wheat cultivars susceptible [7]. Pathotype TTKSK rapidly migrated to neighboring wheat-growing countries, including Tanzania, Kenya, Ethiopia and Sudan, and has subsequently been detected in the Middle East and South Africa [8]. This pathotype acquired virulence for stem rust resistance genes Sr24 and Sr36 in Kenya in 2006 and 2007, respectively [9,10]. In 2013-2014, the Pgt pathotype TKTTF overcame the resistance gene SrTmp and caused an epidemic that destroyed 10,000 ha of the widely cultivated wheat variety Digalu in Ethiopia [11].
Various control measures have been used to manage stem rust epidemics, including the removal of the alternate host (Berberis vulgaris L.) of Pgt, the use of host plant resistance [7] and the use of fungicides [12]. More than 60 stem rust resistance genes have been formally named in wheat (https://shigen.nig.ac.jp/wheat/komugi/genes/symbolClassList.jsp accessed on 20 September 2022). The deployment of rust resistance genes to control stem rust was estimated to save the Australian wheat industry AUD 12 million annually [13]. The deployment of resistance genes singly has resulted in boom-and-bust cycles through the emergence of new pathogen variants [14]. Watson and Singh [15] emphasized the deployment of combinations of more than one gene following the breakdown of singly deployed genes Sr6 (Eureka) and Sr11 (Gabo) in Australia. Although it is possible for a pathogen to acquire virulence for more than one gene at once, the probability of such an event is very low. A successful breeding program for disease resistance relies on the continuous supply of genetically diverse sources of resistance [16,17]. Therefore, identification and characterization of new sources of resistance are crucial to achieve durable disease control.
The Swiss hexaploid winter wheat cultivar Arina was observed to carry an uncharacterized seedling stem rust resistance gene based on multi-pathotype tests by Pathan and Park [18]. Bansal et al. [19] mapped this resistance on the long arm of chromosome 2A based on its repulsion linkage (16.5 cM) with Yr1 using an Arina/Forno recombinant inbred line (RIL) population, and formally named it Sr48. Attempts to identify markers closely linked with Sr48 using several genomic resources in the distal region of chromosome 2AL were futile. In the current study, we used an Arina/Cezanne RIL population to identify and validate markers closely linked with Sr48 for use in marker-assisted selection (MAS).

Construction of DArTSeq Linkage Map
DArTSeq markers were filtered and markers with allele-calling ≥ 98% were selected. SNP markers with a high proportion of missing data (>10%) and rare SNPs with <5% minor allele frequency were discarded. The most informative SNPs were selected based on a threshold PIC value ≥ 0.2. Extra steps; such as (i) imputation of missing data, (ii) genotype correction using the R tool 'ABHgenotypeR', (iii) removal of co-located SNPs and (iv) repeated marker ordering were employed for genetic map construction. Finally, 7115 DArTseq markers were used for linkage map construction.
A genetic linkage map was generated using the ASMap package in the R program at a LOD value of 6. The Arina/Cezanne linkage map had 54 linkage groups. The A, B and D genomes had 1429.0, 1298.6 and 925.9 cM map length, respectively (Supplementary Table S1). The D genome was the least saturated of the three genomes. The Arina/Cezanne linkage map covered a genetic distance of 3653.5 cM, with a density of at least 1 marker per 0.52 cM.

Mapping of Sr48
Seedling IT data for the Arina/Cezanne RIL population was converted into genotypes. The resistant phenotype was converted to 'A' and the susceptible to 'B', and they were then incorporated into the DArTSeq-based linkage map. The stem rust resistance gene Sr48 showed co-segregation with 12 markers located on the short arm of chromosome 2D (Figure 2b). A BlastN search of the DArTseq marker sequences linked with Sr48 identified four CSS contigs ( Table 1) that were used to develop PCR-based STS and SSR markers. We were unable to obtain a good hit (100% similarity) for markers 1211448, 1094958, 1148632, 1000216, 1064890, 1093453, 1104800 and 1246029 when BlastN searched for CSS contigs. Eighteen markers were developed and tested for polymorphism on the parents. Two markers, 'sun590 and 'sun592 , from the CSS contig 2DS_5324961 were polymorphic between the parents. Markers xib59, xib58 (published polymorphic markers from chromosome 2DS), sun590 and sun592 were genotyped on the entire RIL population (172 lines). The sun590 marker was co-dominant and amplified 161 bp and 157 bp fragments in Arina and Cezanne, respectively. The sun592 marker was dominant and amplified a 159 bp amplicon in Arina in addition to a 153 bp amplicon common to both parents. Sanger sequencing of amplicons with STS markers sun846, sun847, sun848 and sun849 (Table 1) from parental lines was performed. Six SNPs were identified when sun849 parental sequences were compared. These SNPs were converted to KASP assays and named sunKASP_238 to sunKASP_243 (Table 2). Primer sequences of these KASP markers are given in Table 2. A genetic map was constructed using polymorphic SSR, STS and KASP markers. Consequently, the Sr48 map ( Figure 2c) included seven markers with sunKASP_239, which located 0.9 cM distal to Sr48, and xib59 was placed 3.5 cM proximal to Sr48 (Figure 2c). Sr48 flanking markers (sunKASP_239 and xib59) were also tested on the Arina/Forno RIL population (200 lines). The sunKASP_239 marker mapped 16.7 cM from Sr48 and xib59 segregated independently.

Molecular Cytogenetic Analysis
Sequential FISH and GISH were performed on the cultivars Arina, Forno and Cezanne in order to study whether there were any translocations between chromosomes. The 4AL-5AL-7BS cyclic translocation found in all hexaploid wheats [20,21] was also observed in all three cultivars (Figure 4b

Molecular Cytogenetic Analysis
Sequential FISH and GISH were performed on the cultivars Arina, Forno and Cezanne in order to study whether there were any translocations between chromosomes. The 4AL-5AL-7BS cyclic translocation found in all hexaploid wheats [20,21] was also observed in all three cultivars (Figure 4b

Polymorphism of the Sr48 Linked Marker sunKASP_239
The closest marker, sunKASP_239, for Sr48 was tested on a set of Australian and European wheat genotypes to determine the polymorphism for its use in the markerassisted selection. One Australian cultivar (Naparoo) and eight European cultivars (Avle, Blanka, Boru, Dacke, Kadett, Vitus, William and WW 20299) amplified the Arina allele (A:A), and the remaining 169 cultivars amplified the Cezanne allele (G:G). These results indicated 94.9% polymorphism among the Australian and European wheat cultivars, respectively.

Discussion
Rust diseases, including stem rust, have historically caused severe yield losses in wheat [23,24]. The identification and use of stem rust resistance genes has been an important strategy to effectively control this disease since the 1950s [7,25]. Sr48 is one of the race-specific resistance genes among over 63 formally designated resistance genes to date [26] https://shigen.nig.ac.jp/wheat/komugi/genes/symbolClassList.jsp, accessed on 20 September 2022) that is effective against Ug99 (H.S. Bariana, unpublished), and can, therefore, be useful when combined with other stem rust resistance genes. Markers linked with this gene can facilitate its rapid selection and pyramiding with other marker-tagged stem rust resistance genes. Bansal et al. [19] located Sr48 on the long arm of chromosome 2A based on its repulsion linkage with the stripe rust resistance gene Yr1, which was previously located on chromosome 2AL [27]. Several unsuccessful attempts were made to find closely linked markers in the Arina/Forno RIL population using various genomic resources.
Repetitive testing of the RIL population at low and high temperatures demonstrated that Sr48 expresses better at lower temperatures (17 • C) compared to the optimum stem rust incubation temperature of above 25 • C (Figure 1; Table 3). This resulted in robust phenotyping of the Arina/Cezanne RIL population. Temperature sensitivity of stem rust resistance genes, including Sr6 [28,29], Sr21 [30], Sr38 [27] and Sr52 [31], has been reported in wheat. The Arina/Cezanne RIL population was originally produced to Mendelize the QTL (Qsr.sun-5BL) that was permanently named Sr56 Bansal et al., 2014 to determine genomic locations of leaf rust and stripe rust resistance genes present in both parents. Sr56 is an adult plant stem rust resistance gene; it is located in the long arm of chromosome 5B and it is ineffective at the seedling stage. Therefore, no interference was observed while phenotyping the population for Sr48. A set of 92 RILs was genotyped using DArTseq markers, and this information was used for the purpose of developing markers closely linked with Sr48. To our surprise, linkage analysis mapped Sr48 to the short arm of chromosome 2D, contrary to the previous report about the location on the long arm of 2A in the Arina/Forno RIL population [19]. The PCR-based markers (sun590 and sun592) developed from the CSS contig 2DS_5324961, showing homology with the DArTSeq marker 1141405, also confirmed the location of Sr48 on 2DS in the Arina/Cezanne RIL population. Further, SNP-based KASP markers were developed by sequencing the parental amplicon produced by STS markers. This observation prompted us to genotype Arina and Cezanne parental cultivars using the previously reported markers on chromosome 2DS [32][33][34] and Cao et al., 2012. This resulted in the construction of a chromosome 2DS linkage map that included sun592, sunKASP_238, sun590, sunKASP_239, Sr48, xib59, xib58 and cfd36 markers (Figure 2c). Flanking markers sunKASP_239 and xib59, when tested on the Arina/Forno population, did not show the expected linkage, authenticating our previous results.
In order to confirm the discrepancy in the location of Sr48 in the Arina/Forno and Arina/Cezanne populations, we performed sequential FISH and GISH analyses on the parental cultivars. The objective behind this cytological study was to find out if there was any translocation that could explain the pseudo-linkage between Yr1 and Sr48. Interestingly, GISH results identified a translocation of what was most likely chromosome 2A at the distal end of the long arm of chromosome 2D of cultivar Forno (Figure 4f; arrow points to the breakpoint). We concluded that this translocation might produce quadrivalents to suggest a pseudo-linkage between Yr1 and Sr48, as was observed by Bansal et al. [19].
The relocation of Sr48 to the short arm of chromosome 2D prompted us to prove the uniqueness of this gene. It is important to examine the relationship of Sr48 and other stem rust resistance loci reported in this region. Two other stem rust resistance genes, Sr6 (Tsilo et al., 2009(Tsilo et al., , 2010 and Sr46, which originated from diploid wheat Aus18913 (Arora et al., 2019), have been mapped to chromosome 2DS (7.9 Mb of the physical map). Sr6 expresses seedling IT 0- [29,34], contrary to the intermediate IT 22 − displayed by Sr48 at low temperatures. In addition, Sr6 is linked with marker wmc453, which was mapped at the 69 cM position [33]. The most distal marker cfd36 in this study was mapped at the 7 cM position [33], which is 62 cM distal to wmc453. Therefore, Sr6 and Sr48 represent different loci.
Sr46 was first identified from Ae. tauschii accession Aus18913 and was cloned by Arora et al. (2019). It exhibits IT 1 + to 2 − under optimal conditions. We tested Arina (Sr48) and a resistant derivative of Aus18913 (donor source of Sr46) 7.1a (kindly provided by Dr. Evan Lagudah) against Pgt 34-1,2,7+Sr38, which is virulent on Sr48 at both low and high temperatures ( Figure 5). The line carrying Sr46 exhibited resistance against this pathotype. Sr46 was placed at 7.9 Mb of the physical map of Chinese Spring (RefSeq v1.0), whereas the marker sunKAP_239 was located close to Sr48 at 2.69 Mb. This comparison indicated that DArTseq marker 1141405 varied from 2.3 to 6.3 Mb in different genomes, but the sunKASP_239 was placed at 3.88 Mb position in Arina 2D pseudomolecules (Figure 3a-d).
The comparison of infection type, phenotypic specificity and genomic locations of Sr46 and Sr48 led us to conclude that these genes are different. All these results demonstrated the location of Sr48 on the short arm of chromosome 2D. Breeder-friendly KASP markers were developed in this study. The marker sunKASP_239 mapped 0.9 cM distal to Sr48 and showed a high level of polymorphism (>94%) in diverse genetic backgrounds. An Australian wheat panel was tested against the different Pgt pathotypes under the Australian Cereal Rust Program, genes were postulated and information was provided to breeders (Bariana per comm). Based on multipathotype tests, it was determined that none of the lines carried Sr48. The marker sunKASP_239 showed the Sr48-specific allele in winter wheat cultivar Naparoo, which also carried Sr24. Pathotypes that were virulent on Sr24 are not available in Australia; therefore, the phenotypic data were not sufficient to confirm the presence of Sr48 in Naparoo. The European wheat panel was tested with five stem rust isolates (Randhawa et al. 2016) and the presence of Sr7b, Sr8a, Sr12, Sr15, Sr17, Sr23 and Sr30 was predicted. Out of nine lines that possessed the Sr48-specific allele, Blanka, Boru and Kadett carried Sr15, whereas Avle carried Sr15+Sr17. The cultivars Dacke and Vitus were susceptible, and the cultivar Kota carried Sr7b+Sr28. Therefore, we concluded that sunKASP_239 showed false positive amplification in these lines.
Due to its high level of polymorphism, the sunKASP_239 marker will be useful in marker-assisted selection of Sr48 and its pyramiding with other marker-tagged genes. Markers for many stem rust resistance genes have been developed (https://maswheat.ucdavis.edu/protocols/stem_rust_protocols, accessed on 20 September 2022), and these include adult plant resistance genes Sr2, Sr55, Sr56, Sr57, Sr58 and Sr63. The deployment of combinations of all stage resistance and adult plant resistance genes in new wheat cultivars through MAS can ensure durable stem rust control [17].
In  Breeder-friendly KASP marker was developed in this study. The marker sunKASP_239 mapped 0.9 cM distal to Sr48 and showed a high level of polymorphism (>94%) in diverse genetic backgrounds. An Australian wheat panel was tested against the different Pgt pathotypes under the Australian Cereal Rust Program, genes were postulated and information was provided to breeders (Bariana per comm). Based on multipathotype tests, it was determined that none of the lines carried Sr48. The marker sunKASP_239 showed the Sr48-specific allele in winter wheat cultivar Naparoo, which also carried Sr24. Pathotypes that were virulent on Sr24 are not available in Australia; therefore, the phenotypic data were not sufficient to confirm the presence of Sr48 in Naparoo. The European wheat panel was tested with five stem rust isolates and the presence of Sr7b, Sr8a, Sr12, Sr15, Sr17, Sr23 and Sr30 was predicted. Out of nine lines that possessed the Sr48-specific allele, Blanka, Boru and Kadett carried Sr15, whereas Avle carried Sr15+Sr17. The cultivars Dacke and Vitus were susceptible. Therefore, we concluded that sunKASP_239 showed false positive amplification in these lines.
Due to its high level of polymorphism, the sunKASP_239 marker will be useful in marker-assisted selection of Sr48 and its pyramiding with other marker-tagged genes. Markers for many stem rust resistance genes have been developed (https://maswheat. ucdavis.edu/protocols/stem_rust_protocols, accessed on 20 September 2022), and these also include adult plant resistance genes Sr2, Sr55, Sr56, Sr57, Sr58 and Sr63. The deployment of combinations of all stage resistance and adult plant resistance genes in new wheat cultivars through MAS can ensure durable stem rust control [17].
In conclusion, the development of a new population (Arina/Cezanne) helped us in relocating Sr48 from the long arm of chromosome 2A to the short arm of chromosome 2D. The presence of a chromosome 2D/2A translocation in Forno was reported for the first time, and it was the cause of the initial location of Sr48 on the long arm of chromosome 2A based on the pseudo-linkage with Yr1. The marker sunKASP_239 is very closely linked with Sr48 and thus it can be used for MAS in breeding. The Sr48-linked SNP can also be useful in genomic selection-based wheat improvement.

Host Materials
A Swiss winter wheat cultivar, Arina (Zenith/Moisson), was crossed with stem rustsusceptible French winter wheat cultivar Cezanne (Thesee/87B29). F 2 seeds derived from a single F 1 plant were harvested and advanced to develop an F 5:7 RIL population of 172 lines using the single seed descent method. A panel of 178 wheat cultivars were used to test the polymorphism of the marker(s) linked with Sr48.

Greenhouse Screening
Parental lines and the Arina/Cezanne RIL population were raised in 9 cm diameter pots filled with potting mix consisting of pine bark and coarse sand (2:1 ratio). The stem rust-susceptible cultivar Morocco was included as a control. The material was fertilized at sowing and six days after germination with Aquasol ® at a rate of 20 g/10 L of water. Inoculations were performed according to Bariana and McIntosh (27) with the Pgt pathotype 98-1,2,3,5,6,7 (PBI culture no. 580). Inoculated seedlings were incubated in the greenhouse at 18-20 • C for 48 h in water-filled steel trays covered with hoods made from translucent polythene sheets to allow natural light to enter and to maintain 100% relative humidity. After incubation, infected seedlings were transferred to two microclimate growth rooms maintained at low (17 ± 2 • C) and high (27 ± 2 • C) temperature regimes, respectively, to identify the ideal temperature for better expression of Sr48. All Arina/Cezanne RILs were tested three times at a low temperature (17 ± 2 • C) and two times at a high temperature (27 ± 2 • C) with pathotype 98-1,2,3,5,6,7 (using 8-10 seeds of each RIL). Rust response assessments were performed 14-16 days post-inoculation using a 0 to 4 infection type (IT) scale [35].

Genotyping of Arina/Cezzane RIL Population
Genomic DNA was extracted from eight seedlings from each RIL and parental line following the procedure described in Bansal et al. [36]. The quality and quantity of DNA were measured on a 1% agarose gel and using a NanoDrop-ND1000 spectrophotometer (Nanodrop Technologies), respectively. The DNA samples of 92 RILs and parents were sent to Diversity Arrays Technology Pty Ltd., Canberra, Australia for genotyping by sequencing (GBS) with the wheat DArTseq ® platform (http://www.diversityarrays.com, accessed on 20 September 2022).
The STS amplification products of Sr48 flanking markers were Sanger sequenced at the Australian Genome Research Facility (www.agrf.org.au, accessed on 20 September 2022) and compared for nucleotide variation using Sequencher software (www.genecodes.com, accessed on 20 September 2022). The SNPs identified in STS sequences were used to design Kompetitive Allele-Specific PCR (KASP) markers using PolyMarker (http://www. polymarker.info/, accessed on 20 September 2022). The STS and KASP markers were named using the prefix sun (Sydney University) followed by the consecutive numbers.
After stripping off the oligo probes, the same slides were used in GISH. The total genomic DNA from T. urartu (AA genome) and Aegilops tauschii (DD genome) was labeled with Fluorescein-12-dUTP and Rhodamine-5-dUTP (Roche Diagnostic Australia, Castle Hill, NSW, Australia), respectively, using nick translation. Unlabeled total genomic DNA of Ae. speltoides (SS genome) was used as a blocker. The probe to blocker ratio was~1:10. Chromosomes were counterstained with DAPI and pseudo-colored blue.

Statistical Analysis
Chi-squared (χ 2 ) test of goodness-of-fit was used to compare the observed stem rust response segregation against the expected genetic ratio for a single gene. A linkage map was constructed using the ASMap function in R software [42]. Genetic distances were determined using the Kosambi mapping function [43] to transform recombination frequencies into centiMorgans (cM).The 2D genetic linkage maps of the chromosomes were generated in MapChart [44].