Tightly linked Rps12 and Rps13 genes provide broad-spectrum Phytophthora resistance in soybean

The Phytophtora root and stem rot is a serious disease in soybean. It is caused by the oomycete pathogen Phytophthora sojae. Growing Phytophthora resistant cultivars is the major method of controlling this disease. Resistance is race- or gene-specific; a single gene confers immunity against only a subset of the P. sojae isolates. Unfortunately, rapid evolution of new Phytophthora sojae virulent pathotypes limits the effectiveness of an Rps (“resistance to Phytophthora sojae”) gene to 8–15 years. The current study was designed to investigate the effectiveness of Rps12 against a set of P. sojae isolates using recombinant inbred lines (RILs) that contain recombination break points in the Rps12 region. Our study revealed a unique Rps gene linked to the Rps12 locus. We named this novel gene as Rps13 that confers resistance against P. sojae isolate V13, which is virulent to recombinants that contains Rps12 but lack Rps13. The genetic distance between the two Rps genes is 4 cM. Our study revealed that two tightly linked functional Rps genes with distinct race-specificity provide broad-spectrum resistance in soybean. We report here the molecular markers for incorporating the broad-spectrum Phytophthora resistance conferred by the two Rps genes in commercial soybean cultivars.

. Response of differential soybean lines and PI 399073 carrying Rps12 and Rps13 genes to 21 Phytophthora sojae isolates. Plants were rated seven days after inoculation as either R (resistant, < 30% seedling death) or S (susceptible, ≥ 70% seedling death).  www.nature.com/scientificreports/ no known Rps genes 17,[42][43][44] were inoculated using the wounded-hypocotyl inoculation technique 45,46 . The experiment was conducted three times. Plants were rated seven days after inoculation as either R (resistant, < 30% seedling death) or S (susceptible, ≥ 70% seedling death). Inocula were prepared using a modified version of the protocol described by Dorrance et al. (2008) 42 . The macerated R17 and Val 12-11 cultures were mixed in equal proportion to prepare the mixed inoculum 45 that is virulent to soybean cultivars carrying Rps genes mapped to Rps1 to 7 loci and partially virulent to lines carrying Rps8. P. sojae strain V13 was also used as a separate inoculum as it is virulent to soybean lines carrying any of Rps1a, 1c, 1d, 4, 7, and 12 genes.

Rps gene
DNA preparation, bulked segregant analysis (BSA). Before inoculation, one unifoliate leaf from each of 11 random plants of individual RIL was collected, bulked and frozen in liquid nitrogen, and stored at − 80 °C. The genomic DNA was extracted from the bulked leaf samples using the CTAB (cetyl trimethylammonium bromide) method 47 49 . Seventeen SBP markers were identified for the Rps12-Rps13 region (Table S1). Simple sequence repeats (SSR) and SBP markers were used to construct a linkage map of the genomic region carrying the putative novel Rps13 gene. Molecular markers based on previously reported NBSRps4/6 sequence and SSR markers 17 , and newly developed SBP markers were used in mapping the Rps13 gene (Table S1). SSR markers linked to RpsJS were also used in mapping the Rps13 region 34 . Eleven polymorphic SSR markers, two previously reported NBSRps4/6 molecular markers along with the newly developed five SBP markers were used to map the Rps13 gene 17 (Tables S1, S2).  2,3,7,8,9,10,11,Yu25, WY, Rps1?, RpsUN1, UN2, and YD29 loci were used to evaluate for possible polymorphisms between the AR2 (susceptible), and PI399036 (resistant) parents (Table 2) in order to identify RILs that carry SSR alleles specific to the P. sojae susceptible AR2 parent.
Linkage map construction and statistical analysis. The Chi-square (χ 2 ) analysis was performed to check the phenotypic data for goodness-of-fit to a Mendelian segregation 1:1 ratio using Graphpad (http:// www. graph pad. com/ quick calcs). Mapmaker version 3.0 50 and the Kosambi mapping function 51 were used to calculate genetic distances in cM units from the recombination fractions between any given two loci. A logarithm of the odds (LOD) threshold was set as 3.0 to determine the linkages between studied loci. Mapmaker package uses the Lander-Green algorithm to calculate the "best" map order of loci 50 . The marker order was determined using the log-likelihood method 50 . The linkage map of molecular markers and the Rps genes was drawn using Map-Chart 2.3 52 . The segregating materials studied in this study were generated by author Silvia Cianzio and will be available from the Bhattacharyya lab, G319 Agronomy Hall, Iowa State University, Ames, IA 50011, USA. All plant collection methods were complied with relevant institutional, national, and international guidelines and legislation.

Results
Identification of putative RILs carrying the Rps12 gene. It was proposed that the Phytophthora resistant PI399036 line contains multiple Rps genes 40 . Earlier we mapped Rps12 of this line using a mixture isolates that overcome most known Rps genes 17 . To investigate the utility of Rps12 against a set of P. sojae isolates collected from Iowa soybean field, we looked for RILs that carry only Rps12. We have investigated 60 Phytophthora resistant RILs generated from the cross between PI399036 × AR2 17 for SSR markers linked to the known Rps regions as described below.
We used 23 SSR markers that were published earlier (  Table 2). The 23 SSR markers were investigated for polymorphisms between the resistant PI399036 and susceptible AR2 parents. Of the 23 SSR markers, 10 SSR markers were polymorphic between the two parents ( Fig. 1) and applied initially in evaluating all 60 RILs homozygous for Rps12; and subsequently, 60 Phytophthora susceptible RILs (rps12rps12). The ten polymorphic SSR markers considered for this study include Satt510 for Rps3 locus, Satt663 for Rps8, Satt440 for Rps2 and Satt631, Satt152 and Satt009 for Rps1, 7, 9, Yu25, WY, and Rps1?, Sattwd15-24 for Rps10, SSR_07_0286 for Rps11, and Satt159 and SSR_03_0250 for RpsUN1 (Table S3). From screening of the 60 resistant RILs, we identified RILs 12 and 14 that carry AR2-specific SSR alleles for nine and eight SSR markers, respectively. For RIL12, SSR marker linked to Rps11 is heterozygous; and for RIL14, two SSR markers linked to Rps8 and Rps11 are heterozygous. These two lines were selected to determine the efficacy of Rps12 to a set of P. sojae isolates.
Identification of Rps13. We have obtained 17 P. sojae isolates from the Robertson lab, Iowa State University, collected earlier from the Iowa soybean fields. The isolates were characterized for their pathotypes by inoculating a set of 14 soybean lines that are considered to be differential lines for 14 individual Rps genes ( Table 1). All these isolates were used to infect the differential cultivars and selected RIL12 and RIL14 and two parents, PI399036 and AR2. RIL12 and RIL14 contain Rps12 and confers resistance against the isolate mixture of R17 and Val 12-11 17 . Surprisingly, RIL12 is not resistant against seven of the 17 new P. sojae isolates and Val 12-11 (Table 1). On the contrary, RIL14 is resistant against these seven isolates. Based on the genetic make-ups of RIL12 and RIL14 for molecular markers of the Rps12 region, we deducted that there is recombination breakpoint in between the NBSLRR533 and Sat_064 in RIL12. We hypothesized that there could be a novel Rps gene named Rps13 located in between Rps12 and telomere on Chromosome 18.
To further support our hypothesis that there is an Rps gene next to Rps12, we evaluated 60 Phytopthora resistant RILs (Rps12Rps12) and 60 Phytopthora susceptible RILs (rps12rps12) for molecular markers of the genomic region containing the Rps12 gene (Table S4). From molecular mapping of the 120 RILs, we were able to identify two additional RILs, RIL9 and RIL81, that carry recombination breakpoints in the Rps12 region, and were evaluated for their responses to 17 new P. sojae isolates, and mixture of R17 and Val 12-11 isolates. The RIL12 contains Rps12, but not the putative Rps13 region; whereas, RIL81 contains the putative Rps13 region but not Rps12. RIL9 contains the putative Rps13 region, but not Rps12. RILs that carry Rps12, but not Rps13, were susceptible to the P. sojae isolates, V13, IV 6b and Val 12-11, resistant to R17 (Table 3). On the contrary, RIL81 carrying Rps13 but not Rps12 was susceptible to P. sojae isolate R17 (Table 3). Our results established that Rps12 is overcome by several P. sojae isolates, against which Rps13 provides immunity.  Table 3). www.nature.com/scientificreports/ Analysis of Rps gene-linked SSR markers revealed that alleles of Satt009 and Satt510 markers specific to Rps1c and Rps3a alleles, respectively, are present in PI399036, but not in AR2. We hypothesized that most likely PI399036 contains Rps1c and Rps3a, in addition to Rps12 and Rps13. P. sojae isolate V13 overcomes the resistance conferred by Rps1c, but not Rps3a. We therefore classified the RILs into two groups based on Satt510: (i) The RILs which carry Satt510 allele specific to rps3a and AR2 parent; and (ii) RILs carry Satt510 allele specific to Rps3a. Both groups segregated for resistance to susceptibility in a 3:1 ratio, as observed for single Mendelian genes, following infection with P. sojae V13 isolate that overcomes the resistance governed by Rps12 and Rps1c. This confirms that there is a novel Rps gene in PI399036.
To map the novel gene, the 120 RILs from the AX20925 population were infected with a mixture of P. sojae R17 and V13 isolates that together overcome all known Rps genes including Rps12, but not the novel Rps13 gene. Of the 120 RILs, 52 RILs showed resistance against the isolate mix and 67 showed susceptibility. The observed segregating 0.867:0:0.08:1.117 genotypic ratio of resistance to susceptibility among the 120 RILs fits to the expected 0.984:0.032:0.984::RR:Rr:rr ratio, where R is Rps13 and r is rps13 for single gene segregation among the RILs in F 7 generation with an estimated 98.4% of the genes homozygous (χ 2 = 0.104).
We conducted bulked segregant analysis (BSA) to identify molecular markers linked to the novel Rps13 resistance gene and confirm that Rps13 is mapped next to Rps12 48 . In this BSA study, we used SSR markers of the Rps12 region to test our hypothesis that Rps13 is linked to Rps12. The results of BSA suggested that indeed the Table 3. Response of four RILS and their parents to 23 P. sojae isolates along and a few P. sojae isolate mixtures. Plants were rated seven days after inoculation as either R (resistant, < 30% seedling death) or S (susceptible, ≥ 70% seedling death). www.nature.com/scientificreports/ gene is co-segregated with the markers mapped in between Rps12 and telomere. To develop a high-resolution map of the Rps13 region, we investigated 19 putative SBP markers for polymorphisms. Five of the 19 putative SBP markers are polymorphic between resistant and susceptible parents and were used for mapping the Rps12-Rps13 region (Fig. 4). The Rps13 gene co-segregated with the Sat_064, BARCOSOYSSR_18_1859, and BAR-COSOYSSR_18_1860 markers. The genetic distance between Rps12 and Rps13 genes is 4 cM (Fig. 5, Table S4). To identify homologues of the candidate Rps13 genes, we investigated the annotated soybean genes in the 92.7 kb Rps13 region between the two markers, SBP56.32 and BARCSOYSSR_18_1861, in the Williams 82 genome sequence located at the soybean genome browser (SoyBase; https:// www. soyba se. org) 53 . Sixteen genes including an NB-ARC domain-containing disease resistance-like gene, Glyma.18g283200, are present in this region (Table S5). Williams 82 does not carry the Rps13 gene.

Discussion
This study was designed to investigate the usefulness of the Phytophthora resistance governed by the Rps12 gene 17 . In mapping Rps12, we had to use a mixture of P. sojae isolates to mask the effect of previously known Rps genes that were in the PI399036 line, the source of Rps12 17 . To determine the utility of Rps12 genes against a set of uncharacterized P. sojae isolates, we must identify an RIL that contains only Rps12. We therefor first examined a set of 60 Phytophthora resistant RILs for possible absence of other known Rps genes by studying the Figure 2. Two RILs differing alleles at the linked Rps12 and Rps13 loci showed distinct responses to P. sojae V13 isolate and the mixture of R17 and Val12-11 isolates. P. sojae isolate V13 failed to defeat resistance mediated by Rps13 gene but could overcome that by Rps12; whereas, the mixture of R17 and Val12-11 isolates could overcome Rps13, but not Rps12. www.nature.com/scientificreports/ polymorphisms of Rps gene-linked SSR markers. Linked SSR markers co-evolved with linked Rps genes and SSR alleles can be used to predict alleles of the linked Rps genes. A total of 210,990 SSRs were identified from the soybean genome. Of these, 61,458 SSRs contain repeat units of di-, tri-, and tetranucleotide with (AT)n, (ATT)n and (AAAT)n as the most abundant motifs 54 . A genetic linkage map consisting of 20 linkage groups with approximately 1500 SNP, 1000 SSR markers, 700 RFLP, and 73 RAPD markers and 46 classical trait loci is available in soybean [55][56][57] . Information of genetic markers has been used to map Rps1, Rps2, Rps3, Rps4, Rps5, Rps6, Rps7, and Rps8 loci to Chromosomes 3, 16, 13, 18 and 13, respectively 24,25,31,32,35,[57][58][59] . While the Rps4 locus was mapped close to the Rps6 region, the Rps8 locus mapped close to the Rps3 region 31,32,35 . The RFLP marker pT-5 was shown to be linked to the Rps5 locus 36 . SSR markers mapped to the Rps5 locus are yet to be identified 24 . Thus, SSR markers linked to each Rps gene except Rps5 have been reported 24 .
In this study, we selected 23 SSR markers that have been shown to be linked to most of the reported Rps genes ( Table 2). Out of 21 SSR markers, 10 were polymorphic between the two parents, PI399036 and AR2 (Fig. 1). These 10 SSR markers were applied in evaluating all 60 RILs homozygous for Rps12. These polymorphic markers included Satt510 linked to the Rps3 locus, Satt663 to Rps8, Satt440 to Rps2, and Satt631, Satt152 and Satt009 to Rps1, 7,9,Yu25,WY,Rps1?, to Rps10, SSR_07_0286 to Rps11, Satt159 and SSR_03_0250 to RpsUN1. PI399036, the source of Rps12, exhibitted the alleles of the Satt009 and Satt510 linked to linked to Rps1c and Rps3a alleles, respectively, suggesting that PI399036 most likely contains Rps1c and Rps3a genes as well.
From PCR assays of 60 Phytophthora resistant RILs with 10 SSR markers polymorphic between PI399036 and AR2, we identified RIL12 and RIL14 carrying rps alleles-specific SSR alleles for nine and eight SSR markers, respectively. For RIL12, SSR marker linked to Rps11 is heterozygous and for RIL14, two SSR markers linked to Rps8 and Rps11 are heterozygous. These two lines were selected to determine the efficacy of Rps12 to a set of 17 P. sojae isolates collected in Iowa soybean fields. RIL12 was susceptible to seven of the 17 P. sojae isolates; whereas, RIL14 was resistant to these seven isolates. Earlier both lines were shown to carry Rps12 17 . We hypothesize that RIL12 lacks an unknown Rps gene that is present in RIL14. In the absence of this unknown gene the RIL12 failed to provide immunity against four of the 17 isolates studied ( Table 3). The putative unknown gene is named as Rps13. BSA revealed that the gene is linked to Rps12. Genetic mapping using 18 molecular markers placed the gene on the south arm of Chromosome 18, at a 4 cM genetic distance from Rps12. We observed that due to the absence of Rps13 in RILs 6,9,42, and 49 resulted in susceptibility to the P. sojae isolate V13. However, the four lines contain Rps12 and resistant to the mixture of the isolates, R17 and Val 12-11 that cannot overcome resistance encoded by Rps12. On the contrary, RIL81 contains Rps13 but not the Rps12 gene. Therefore, this RIL is resistant to V13 and susceptible to the mixture of R17 and Val 12-11 isolates (Fig. 2). These results established that there is a novel gene next to Rps12 that is essential for immunity of the RILs against four of the 17 P. sojae isolates collected in Iowa. Two linked functional Rps genes provide broad-spectrum resistance against P. sojae isolates tested in this study.
Plant activates defenses against pathogen attacks, determined by a corresponding pair of genes, a gene for avirulence in the pathogen and a gene for resistance (R) in the host. Such resistance mechanisms function in both   69 and rice 70,71 . The clustered distribution of R-genes provides a reservoir of genetic variation from which new pathogen specificity can evolve through gene duplication, ectopic recombination, unequal crossing-over and diversifying selection 72 . These clusters frequently comprise tandem arrays of genes that regulate resistance to multiple pathogens and to multiple variants of a single pathogen. The clusters may be tight with a little intervening sequence as 20 kb between two functional Rps1-k genes in soybean 26,73 , the RPP5 cluster in Arabidopsis thaliana spans 91 kb 64 , or be spread over several megabases as the Resistance Gene Candidate2 (RGC2) locus in lettuce (Lactuca sativa) 74 . In rice, also Chromosome 11 is highly enriched in R-genes, mostly in clusters; up to 201 loci encode the domains of NBS-LRR and LRR-receptor-like kinase (LRR-RLK) or wall-associated serine/threonine protein kinase (WAK) 70 . The Rps12-Rps13 region is rich in Rps genes. As of now, Rps4, 6,12,13 and JS are mapped to the same genomic region spanning probably less than 5 cM in different soybean haplotypes ( 17,34,35 this work). Earlier we demonstrated that Rps4 and Rps6 are allelic and Rps4 co-segregates with Sat_064 35 . Therefore, most likely Rps13 is allele to Rps4 and Rps6. The PI399036, the donor of Rps12 and Rps13, does not carry Rps4 or Rps6 and therefore Rps13 is distinct from the two Rps genes 17 , this study. www.nature.com/scientificreports/