Mapping of crown rust resistance gene Pc53 in oat (Avena sativa)

Crown rust disease caused by the fungus Puccinia coronata f. sp. avenae (Pca) is a major production constraint of oat in North America, Europe, and Australia. There are over 100 genes effective against one or more Pca races, but only a handful of seedling resistance (Pc) genes have been mapped to a known chromosomal location. The goal of the present study was to use linkage mapping to identify the genomic location of the Pc53 gene, and to produce a list of linked SNPs with potential as molecular markers for marker assisted breeding. The Pc53 gene was placed on the linkage group Mrg08 at 82.4 cM using F5-derived recombinant inbred lines (RILs) from a cross between the Pc53 carrier 6-112-1-15 (PI 311624) and the susceptible cultivar Otana. The map location was validated using RILs from a cross between 6-112-1-15 and the Pc50 differential line. Single nucleotide polymorphism marker GMI_ES02_c14533_567 was the closest to Pc53. A major seedling resistance gene ‘PcKM’ and QTL QcC.Core.08.1, QCr.Core.08.2, QCr.Core.08.3 and QCr.cdl9-12D were previously reported on Mrg08. QPc.Core.08.1 and PcKM were mapped to within 1 cM of Pc53; but previous virulence studies have indicated separate identities. The chromosomal location of Pc53 and SNPs linked with it will facilitate the utilization of Pc53 in oat breeding programs.


Introduction
Crown rust caused by the fungus Puccinia coronata f. sp.avenae (Pca) is a major production constraint of oats (Avena sativa L.) in North America [1].The disease is also present in most oat producing countries of the world [2,3].The virulence of the pathogen evolves continuously as it undergoes sexual recombination on the alternate host buckthorn (Rhamnus spp.) in North America [1].There are effective fungicides available to manage crown rust in oats, but their use may not be economically justifiable [4].Deployment of cultivars carrying effective genes against the pathogen is generally accepted as the best strategy for managing rust diseases in small-grain cereals including oat crown rust [4][5][6].Over 100 race-specific crown rust resistance genes have been identified in oats.Most of these genes originated from the wild oat species A. sterilis [7].'Fidler' was the first oat cultivar to be developed in the 1980s with the A. sterilis-derived resistance gene, Pc39.Since then, genes Pc38, Pc39, Pc48, Pc68 and Pc94 have been deployed in cultivars such as: Dummont, AC Medallion, AC Assiniboia, AC Pinnacle, and Leggett with widespread use [7].Unfortunately, the effectiveness of singly deployed seedling resistance genes is short-lived [8,9] due to the continuous evolution of virulence in the pathogen in North America [1].
Pc53 is one of the many Pc genes identified in oats.It was originally identified from the wild oat species, A. sterilis, in the Mediterranean region of Israel (GRIN, http://www.ars-grin.gov/npgs/).Despite the occurrence of virulence to Pc53, the frequency of virulence against this gene in North America had been consistently low compared to most Pc genes [1,10,11], suggesting that it might be a useful gene to utilize in oat breeding for crown rust resistance.
Only a handful of the nearly 100 Pc genes have been mapped to a known chromosomal location of oat [6,12,13].These include Pc38 [14], PcKM/Pc45 [13], Pc58 [12,15], Pc68 [16], Pc71 [17], Pc91 [6,13] and Pc94 [18].Absence of information about the chromosomal location of most of the Pc genes has created multifaceted challenges in oat research.First, it has limited the utilization of genomic tools in oat breeding.Second, in the continuous search for new sources of crown rust resistance, the lack of information on genomic location restricts investigations as to novelty to the conventional virulence test methods which require tremendous amounts of time and resources.For example, Gnanesh et al. [13] identified a major seedling crown rust resistance gene designated as 'PcKM' in the oat cultivar 'Morton' in field and greenhouse studies; but disease reactions of differential lines and molecular marker data suggested that PcKM may be Pc45.Another study by Martens et al. [19] and Simons [20] had suggested Pc54 could be allelic or closely linked with gene Pc35; while Leonard et al. [9] placed Pc35 and Pc54 on different linkage groups.An association mapping study of elite germplasm identified QTL conditioning crown rust resistance (Qpc.CORE.05) in the same region as Pc71 [21].The novelty of these and other QTLs that are effective against crown rust [21,22] could not be determined because of the lack of information about the map location of most Pc genes.The current study complements other studies which have mapped crown rust resistance genes, and will improve the availability to oat researchers of information on chromosomal location of Pc genes.
The objectives of the present study were to: 1) map Pc53 on the oat consensus linkage map of Chaffin et al. [23] using SNPs of the 6K Infinium gene chip; and 2) identify SNPs closely linked with Pc53.

Plant material
A population of 149 F 5 -derived recombinant inbred lines (RILs) was developed from a cross between the Pc53 differential line, Clinton/6-112-1-15 (Pc53) and the susceptible cultivar Otana, developed by ARS-USDA, Aberdeen, ID (CI 5345/Zanster (CI 5345/2 � Overland)).The map location of Pc53 was validated using 125 F 5 -derived RILs from a cross between the Pc53 and the Pc50 differential line (Pc50), (CAV 2643/4 � Pendek).Both mapping populations were developed by single seed descent at the Small Grains and Potato Germplasm Research Facility of the USDA-ARS in Aberdeen, Idaho.

Evaluation for crown rust reaction
Seedling tests were carried out on both RIL populations (Pc53 x Otana and Pc53 x Pc50) at the Small Grains and Potato Germplasm Research Facility of the USDA in Aberdeen, Idaho in 2017-2018.Pca race LGCG designated according to Chong et al. [24] was used to evaluate populations Pc53 x Otana.The race was virulent against Otana producing infection type (IT) of 3 and 4, but avirulent against Pc53 (infection type; 1).As race LGCG produced similar IT on PC53 and Pc50, a different race 'NBTG' was used to evaluate the Pc53 x Pc50 population as it produced distinct IT characterized as small flecks with small uredia (;1) on Pc50, but an immune response (IT of 0) on Pc53.Races 'LGCG' and 'NBTG' were obtained from the UDA-ARS Small Grains and Potato germplasm Research Facility, Aberdeen, ID and the USDA-ARS Cereal Disease Laboratory, St. Paul, MN, respectively.A single pustule isolate from each of the races was propagated on Otana plants and used to derive inoculum for resistance testing.Approximately, five seedlings from each F 5:6 RIL, and the parents, were sown in 3.8 cm diameter 'containers' containing a 3:2:2 (v:v:v) mix of sand, peat moss, and vermiculite and maintained in a greenhouse adjusted to 22˚C and 18 hour photoperiod.Each RIL was planted in three containers, for a total of 15 seed evaluated per line.Two weeks after planting, seedlings were inoculated with uredinia of the respective race suspended in Soltrol170 isoparaffin (Chevron Phillips, The Woodlands, TX) using a motorized sprayer (GAST Manufacturing inc., Benton Harbor, MI) and a small atomizer (G-R Manufacturing Manhattan, KS).Plants were left in the open for one hour to dry, and then transferred to a dew chamber set at 20˚C with no light.After 18 hours in the dew chamber, plants were placed in a growth chamber adjusted to 20-22˚C with an 18-hour photoperiod.Disease reactions on the first and second leaves were recorded 14 days after inoculation as IT on a 0 to 4 scale where 0-2 are considered resistant reactions, while 3-4 are susceptible reactions [24].

Genotyping
High quality DNA was isolated from F 5 plants of the two populations (Pc53 x Otana and Pc53 x Pc50) and parents following a protocol by Anderson et al. [25] with modifications, including grinding freeze dried leaves with beads in a Mixer Mill MM 300 shaker (Retsch, Hannover, Germany) for 10 min at 25 strokes per second.The concentration and quality of DNA were estimated using the BioTek plate reader (BioTek Instruments Inc., Winooski, VT).Genotyping was performed with an Illumina Infinium iSelect oat SNP chip containing 4975 SNPs at the Cereal Crops Research Unit of ARS-USDA in Fargo, ND.Genotype calling for each RIL and parental line was performed automatically using the DBSCAN procedure in GenomeStudio, v. 2.0 (Illumina, San Diego, CA, 2016), and was manually inspected for call accuracy.

Statistical analysis
Individual RILs in each population were classified based on the IT produced by the F 5:6 families.The goodness-of-fit of the observed disease reaction to the expected segregation ratio of 1:1 for a single gene was tested using Pearson's Chi-squared (X 2 ) distribution analyses.

Genetic mapping
SNP markers that were polymorphic between parents of each population, and with less than 10% missing data were selected.Of these, only SNPs that were assigned a location on the oat consensus map of Chaffin et al. [23] were used for analyses.A total of 706 and 569 mapped polymorphic markers were used in the Pc53 x Otana and Pc53 x Pc50 populations, respectively (Table 1).JMP genomics, v. 8.1 (SAS Institute, Cary, NC, 2016) was used to construct genetic linkage maps in both populations.The initial number of linkage groups and marker order in each linkage group were determined using the recombination and linkage groups, and the linkage map order functions of the software, respectively.SNPs were placed into linkage groups using a maximum distance between markers of 30 cM.Genetic distances between markers were calculated in centiMorgans (cM) using the Kosambi map function [26].Graphical linkage groups were generated using the linkage map viewer function.Linkage groups were assigned to the oat consensus map of Chaffin et al. [23].

Sequence homology
The sequences of the single nucleotide polymorphism markers linked with the Pc53 gene were compared with the rice (Oryza sativa L.) genome (https://www.ncbi.nlm.nih.gov) using the search function 'sequence(BLAST) search' to identify candidate genes for Pc53.

Phenotypic analysis
Race LGCG produced compatible ITs of 3 to 4 on the susceptible parent (Otana) and incompatible ITs of; and; 1 were observed on the resistant parent Pc53.Race NBTG produced ITs of 0 and; 1 on the Pc53 and Pc50 differential lines, respectively.Phenotyping of F 5:6 families showed that the Pc53 x Otana and Pc53 x Pc50 progeny segregated at a ratio of 79R:71S and 66R:59S.Based on Chi-square tests both populations fit 1R:1S Mendelian ratio with X 2 of 0.32 and 0.39, and P-value of 0.572 and 0.531, respectively.The segregation ratio shows that the single-seed descent procedure worked as expected without introducing any selection bias.

Genetic mapping of Pc53
The 706 polymorphic SNP markers were assigned to 18 linkage groups.Linkage analysis revealed that the Pc53 gene was localized at 2.4 cM distal to the SNP markers GMI_ES01_c28412_66 and GMI_ES02_c12737_306 (Fig 1).Other single nucleotide polymorphism markers closely linked to Pc53 were GMI_DS_LB_9591 (2.8 cM), GMI_ES02_c3359_447 and GMI_ES02_c3359_678 (3.2 cM), GMI_ES02_c8737_267 and GMI_ES02_c24327_173 (7.9 cM), and GMI_ES02_c8277_506 (9.1 cM).All the SNP markers linked to Pc53 were located on the linkage group Mrg08 of the oat consensus map (23) (Table 2).

Validation
RILs from the Pc53 x Pc50 population were used to validate the chromosomal location of Pc53.
The 569 polymorphic markers were assigned to 8 linkage groups.Linkage analysis revealed that Pc53 was on the same map location as GMI_ES02_c14533_567 at 46.6 cM of the linkage group (Fig 2).SNP markers GMI_ES15_c12065_114 and GMI_ES17_c19330_272 flanked Pc53 at 9 cM proximal; while GMI_ES15_c3200_563, GMI_GBS_821 and GMI_ES_LB_11757 were 13.8 cM distal to Pc53.Like the mapping population the SNP markers linked to Pc53 were located on linkage group Mrg08 of the oat consensus map [23] (Table 2).

Sequence homology
According to Oliver et al. [15] Mrg08 of the oat chromosome is homologous to rice chromosome 4.Comparison of the sequences of SNPs closely linked to Pc53 on Mrg08 of the oat chromosome with the rice genome identified three oat SNPs (GMI_ES02_c14533_567, GMI_ ES02_c3359_678, and GMI_ES02_c3359_447) with �80% sequence homology to rice loci.Rice chromosome 4 loci LOC107276041 (31764063), LOC_Os04g0661600 (33740410), and LOC_Os04g56620 (33761378) had sequence identity of 90, 80, and 84%, respectively to these three oat SNPs.The former rice locus encodes an uncharacterized protein, and the latter two loci encode the molybdopterin biosynthesis protein CNX1.The oat SNPs closest to the Pc53 gene in the Pc53 x Otana population (GMI_ES01_c28412_66, GMI_ES02_c12737_306 and GMI_DS_LB_9591) did not have significant sequence homology to a rice locus.

Discussion
Pc53 was originally identified from the A. sterilis accession 6-112-1-15 collected in northern Israel.This gene has not been used in cultivar development apart from some component lines of the Iowa multilines [1].Virulence to Pc53 is present in the Pca populations of North America [1,10,11].However, the proportion of isolates collected in a decade (2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009) in North America with virulence to Pc53 has never been greater than 5% of all isolates collected.In contrast, 65-95% of isolates collected in the same period were virulent against the widely deployed genes Pc38, Pc39, Pc63, Pc67, and Pc71 [1].This result suggests that the Pc53 gene may have sufficient effectiveness to be useful in variety development, particularly when deployed in combination with other Pc genes.QTL conditioning crown rust resistance in oats have been previously mapped to the Mrg08 linkage group.A major seedling resistance gene 'PcKM', and the QTL QCr.cdl9-12D were reported on Mrg08, as were the GWAS QTL QcC.Core.08.1, QCr.Core.08.2, and QCr.Core.08.3 [13,21,22].QPc.Core.08.2 and QPc.Core.08.3 were proximal to GMI_ES02_c3359_447 and GMI_ES15_c6639_318, while Pc53 was distal to both markers, suggesting that the two QTL may not be related to Pc53.Similarly, QCr.cdl9-12D was proximal to GMI_DS_LB_9591, GMI_ES02_c3359_678 and GMI_ES17_c19330_272, and distal to GMI_ES17_c3134_489, GMI_ES01_c14013_124 and GMI_ES17_c12269_176, while Pc53 was distal to the former three and proximal to the latter three SNPs.
Pc53 was mapped to within 1 cM of the reported locations of QPc.Core.08.1 and PcKM, suggesting that they may be linked or allelic.Based on crown rust virulence profiles over the period of their field trials, Esvelt-Klos et al. [21] suggested Pc45 and Pc53 as possible candidate genes for the QTL QPc.Core.08.1.The Pc45 gene has been widely used in North American oat breeding and is expected, therefore, to be abundantly represented in the CORE panel.It is less likely that the QTL identified in the CORE germplasm originated from Pc53 as this gene is not known to be in the pedigrees of released cultivars apart from the Iowa multi-lines ( [1], https:// triticeaetoolbox.org/),which were not included in the CORE association mapping panel.In addition, virulence analyses of differential lines and molecular marker data have suggested that PcKM might be Pc45 [13].To understand the relationship between Pc45/PcKM and Pc53 we examined previous virulence analyses reports.Admassu-Yimer et al. [27] reported that Pca race 'MGBH' produced an IT of 0; on Pc45 and an IT of 3-4 on Pc53.However, the present study placed Pc53 at the same linkage group and map location as PcKM [13].We hypothesize that Pc45 and Pc53 may be components of a complex of genes in this region conditioning resistance to crown rust.Such phenomena are not uncommon in oats.A good example is the resistance of TAM-O-301 to crown rust, which is conferred by complex of Pc58 genes [12].Gregory and Wise [28] and Wise et al. [29] also reported two and five linked genes respectively conditioning crown rust resistance in diploid oat.In another study by Chong and Brown [30] two to five clustered genes were reported as responsible for crown rust resistance in A. sativa.
There is currently no sequence available for the Avena sativa genome, limiting our ability to exhaustively catalog positional candidate genes for Pc53.However, sequence homology of SNPs linked to Pc53 are consistent with synteny to a region of rice chromosome 4 spanning over 10 Mbp (https://www.ncbi.nlm.nih.gov).A smaller region of synteny corresponding to the region between GMI_ES02_c14533_567 and GMI_ES02_c3359_678 spans close to 2 Mbp and contains over 600 gene annotations.Genes of interest in that region include LOC9269204 disease resistance protein RGA2, and the innate immune system associated LOC4336849 secretory carrier-associated membrane protein 6.In the future, access to an annotated oat genome sequence may enable the identification of positional candidate genes for Pc53 with sufficient confidence to advance as targets for cloning to further study the mechanisms of crown rust resistance in oats.
In summary, we found that Pc53 conferring race-specific crown rust resistance in oat is linked with SNP markers placed on the oat consensus linkage group Mrg08.This is the first report of the chromosomal location of Pc53.The SNPs linked with it may facilitate the use of this gene in oat breeding programs.In the long term, mapping of oat crown rust resistance genes will contribute towards a more complete understanding of oat genomics and enable researchers to utilize molecular tools to develop new oat cultivars with desirable traits.

Table 2 . Single nucleotide polymorphism markers closely linked with the crown rust resistance gene Pc53 in two oat mapping populations.
[23]groups and positions are based on the consensus map developed by Chaffin et al.[23].