Nucleotide-binding site leucine-rich repeat (NLR) resistance genes typically confer resistance to a single pathogenic population, and multiple R genes are required to confer persistent resistance to a single pathogen. The authors report the NLR gene Lr/Yr548 from Aegilops sharonensis and Aegilops longissima, which has a positive effect on Puccinia triticina (Pt) and P. longissima, which cause leaf rust and stripe rust, respectively. striiformis tritici (PST) is specifically resistant. Lr/Yr548 could prevent disease occurrence in both wheat introgressive lines and transgenic wheat lines. Comparative analysis of Lr/Yr548 and all cloned NLR resistance genes of cereals showed that Lr/Yr548 had a unique helix-helix domain, which was only present in Salon goatgrass and tall goat grass. Phylogenetic analysis showed that Lr/Yr548 underwent multiple gene flow events between the two species and suggested loss of resistance in susceptible races. The specific presence of Lr/Yr548 in tall goat grass and Sharon goat grass, as well as its cross-resistance to Pt and Pst in wheat, highlight the potential of these species as a source of novel disease resistance genes for wheat improvement.
Outcome:
Different methods revealed the same leaf rust and stripe rust resistance genes in Sharon goat grass and tall goat grass
Sharon goat grass
Leaf rust and stripe rust resistance genes in Salon goatgrass were previously introduced into the wheat cultivar Galil. Resistance is localized to a 17 Mb region of chromosome 6B-6Ssh of the D42 introgressive line (Figures 1a1-3). To isolate candidate genes, the authors used ethylmethanesulfonate (EMS) mutagenesis to create mutant alleles of leaf rust and stripe rust resistance loci in 3,086 seeds of the D42 line, yielding 1,158 M2 families. By stepwise screening for Pt #526-24 and Pst #5006小种 that are not virulent to D42, the authors identified 16 independent M3 lineages that are susceptible to both races (Extended Data Figure 1). The authors confirmed the presence of imported fragments in all 16 M3 lines by PCR and diagnostic primers (Extended Data Figure 2). The crosscrossing of the susceptible M3 mutant with its resistant plants yielded a 1:3 segregation ratio (Extended Data Table 1) and the crossing with the susceptible wheat parent variety Galil produced homogeneously susceptible progeny, suggesting that resistance to both pathogens is conferred by a single genetic locus.
扩展数据图1. M3 D42渐渗系易感突变体对Puccinia triticina小种#526-24（顶部）和P. striiformis小种#5006（底部）感病的反应。
Extended Data Figure 2. Diagnostic primers were used to verify representative results of the imported fragments in the EMS M3 susceptible mutant mut_23-2.
Extended Data Table 1. F2 progeny of M3 EMS mutants with their resistant brothers or susceptible parents cv. After Galil hybridization, pairs of Pt #526-24 and Pst #5006的感染类型 (IT) were tested at the seedling stage.
To isolate this gene, the authors employed a mutant transcriptomic approach, using six EMS mutagenic lines (Extended Data Table 2), the D42, and R gene donors Sharon goat grass AEG-548-4. The authors #5006小种接种植物 with Pt #526-24 and Pst, and extracted RNA from leaf samples at 0 h post-inoculation (hpi), 24 hpi, and 48 hpi. After RNA sequencing and processing, the authors searched the data for candidate genes that were expressed in D42 and AEG-548-4 at one or more time points and contained EMS-derived mutations in at least three of the six mutant lines. Analysis yielded several potential candidate genes, the most promising being the transcript TRINITY_DN6116_c0_g1 (Supplementary Table 1): this transcript was expressed at all time points, contained mutations in five mutant lines, and encoded nucleotide binding sites (NBS) and leucine-rich repeat (LRR) domains. Single nucleotide polymorphisms (SNPs) in all mutant lines are located in the same coding sequence (CDS) and result in amino acid substitutions (Figure 1a4 and Extended Data Table 3).
Extended Data Table 2. Purity in chromosome fractions of 6B-6Ssh sorted from wheat-goat grass introduction line D42 and six EMS mutants (muts) of the same line, and total DNA prepared from chromosome fractions sorted by flow cytometry.
Extended Data Table 3. SNP mutations in TRINITY_DN6116_c0_g1 transcripts detected in EMS mutant systems.
To analyze the non-coding regions, the authors performed chromosome sorting and sequencing and obtained high-quality chromosome enrichment and sequences of five mutant lines. Aligning the reads to the D42 assembly of deep sequencing (Extended Data Table 4) revealed that three genomic fragments covered the entire sequence of the candidate gene.
To verify the function of the candidate gene, the authors used virus-induced gene silencing (VIGS) technology to inhibit its expression in the D42 line by targeting two non-overlapping gene-specific sequences (Figure 1a4). VIGS plants showed a more than 75% reduction in the expression of the candidate gene (Figure 1b) and #5006小种易感 to Pt #526-24 and Pst (Figure 1c), confirming that this gene is required for leaf rust and stripe rust resistance. The authors tentatively named the candidate gene Lr/Yr548.
Figure 1. The Lr/Yr548 gene of Sharon goat grass was cloned by MutChromSeq and functionally validated by VIGS.
Extended Data Table 4. Statistical analysis of de novo genome assembly using chromosome sorting sequence data from D42 lead-in lines using three different assemblers.
To visualize the Lr/Yr548 protein, the authors generated a three-dimensional model of its predicted structure (Figure 1a5). The predicted Lr/Yr548 protein has a similar structure to helix-helix (CC) NLRs, including a horseshoe-shaped structure of a typical NBS domain and an unusual C-terminal LRR domain containing 36 repeats (Extended Data Figure 3a). All amino acid substitutions found in the EMS mutant are mapped to the functional domain of Lr/Yr548 (Figure 1a, 4-5 and Extended Data Table 3), with two mutations located in the outer part of the LRR domain (Extended Data Figure 3b), a diverse hotspot that may facilitate pathogen identification.
Extended Data Figure 3. Structural analysis of Lr/YR548.
Tall goat grass
To look for additional rust resistance genes, the authors also screened a diversity panel of 244 tall goat grass strains, inoculated with two Pt races #12460 and #12337 at seedling stage. The authors obtained similar results with a resistance frequency of 30% for these racelets in tall goat grass panels (Figure 2a and Supplementary Table 2). To identify candidate leaf rust resistance genes, the authors used the AgRenSeq method using a whole-grain NLR capture library Tv_315 combined with phenotypic results from plants infected with two Pt races. The authors generated de novo assemblies from Illumina short reads, filtered sequences using NLR-parser, and performed k-mer-based association genetic analysis of 138 strains.
The authors found that two fragments, 71128 and 71798, were significantly associated with these two races, which were located in an unlocalized region of the reference genome ("chromosome unknown"). To pinpoint the R gene, the authors used PacBio to generate additional genome sequence data for the AEG-6782-2 reference line of Goat Grass and generated an improved genome assembly (Extended Data Figure 4) with a contig N50 of 16 Mb. Using the Hi-C data, the assembly was sequenced into 7 pseudo-molecules consisting of 5.9 Gb (98.6% of the assembled sequence) containing 67.8 Kb of voids, while the first assembly based on short-read data contained 931.4 Mb of voids. The improved assembly allowed the authors to map two AgRenSeq NLR candidate fragments to positions 51,756,423–51,772,629 (16,207 bp) on chromosome 6S (Figure 2b).
Extended Data Figure 4. Comparison of chromosome positions of tall goat grass genome assembly version v1.069 and v2.0 (in this study).
Candidate sequence alignment of the resistance gene to the tall goat grass leaf rust and Sharon goat grass Lr/Yr548 showed that the two genes were almost identical: they were identical in size (16,207 bp) and contained three highly similar introns, with almost identical exons and differing only in three synonymous nucleotide substitutions, yielding a 4,533 bp CDS encoding a 1,510 amino acid predicted protein. Therefore, the authors independently isolated a direct homologous rust resistance gene from Sharon goat grass and tall goat grass by two different methods, respectively.
To validate the function of Lr/Yr548 from Goat Wort in resistance to Pt and Pst, the authors created a transgenic wheat variety Fielder expressing Lr/Yr548 cDNA or a synthetic genomic DNA (gDNA) clone containing four exons separated by shortened introns, both with a primitive promoter and termination sequence (described in detail in the method and Figure 2c and the extended data in Figure 5). The authors generated independent primary transgenic events (T0), selected three events each containing cDNA or synthetic gDNA clones with different copy numbers, produced homozygous T1 and T2 progeny, and tested their responses to infection with Pt and Pst races. Three cDNA transgenic lines showed varying degrees of resistance to Pt #526-24 and #12460 and Pst #5006小种均敏感, while three gDNA transgenic lines showed varying degrees of resistance (Figure 2d). Reverse transcription quantitative PCR (RT-qPCR) analysis showed that the expression level of Lr/Yr548 in cDNA transgenic plants was at least 2.5-fold lower than in gDNA plants and about 6-fold lower than in tall goat grass (Figure 2e). The relatively low expression levels of the cDNA transgenic lines (Extended Data Figure 6a) suggest that introns within the coding sequence are important for proper Lr/Yr548 expression. In addition, there was no difference in the expression of Lr/Yr548 between infected and uninfected plants (Extended Data Figure 6b), although it increased with plant age (Extended Data Figure 6c). These results support the function of Lr/Yr548 in resistance to Pt and Pst, and highlight the importance of introns for their proper expression.
Figure 2. Lr/Yr548 was cloned by AgRenSeq and functionally validated in the transgenic wheat variety Fielder.
Extended Data Figure 5. Construction strategy of Lr/YR548 genome clonal transformation vector.
Extended Data Figure 6. Lr/Yr548 gene expression
Lr/Yr548基因在沙伦山羊草和高大山羊草中特有
The authors searched for orthologous genes of Lr/Yr548 in other species. A local BLAST search was performed in the v2.0 tall goat grass genome assembly using the CDS of Lr/Yr548 as a query, and the results showed that there was only one copy of Lr/Yr548 in the tall goat grass genome. The published genome of Salon goat grass does not contain Lr/Yr548. A global BLAST search using CDS of Lr/Yr548 was used as a query in the NCBI and specific cereal databases, and the results showed sequence similarities in the plant species of three families: wheat, grass, and breviflora (Supplementary Table 3). The two genes with the highest sequence similarity to Lr/Yr548 were the hypothetical resistance gene RGA1 (87% similarity to the predicted CDS of Lr/Yr548) and the RGA2-like putative resistance gene of Urartu wheat (83% similarity). All other predicted genes had sequence similarity of less than 80% or coverage of less than 80%. A global BLAST search using the Lr/Yr548 protein sequence as a query showed a low similarity with G. japoides RGA1 (79%), and the similarity of the remaining sequences was less than 65% (Supplementary Table 4).
To examine whether RGA1 is a functional homolog of Lr/Yr548, the authors examined the responses of three R. arthrium lines with or without RGA1 to Pt #526-24 race. All three strains showed moderate resistance to Pt #526-24 (Extended Data Figure 7). Similarly, wheat varieties containing multiple RGA2 homologs with 73-79% similarity were susceptible to Pt #526-24. The authors concluded that there are no functional homologs of Lr/Yr548 from Sharon goat grass and tall goat grass in the current database.
Extended Data Figure 7. Reaction of Arthrium japodium strain to Pt #526-24 and PCR detection of RGA1 homologs.
Phylogenetic analysis using the predicted full-length Lr/Yr548 protein sequence (Figure 3a) or only the LRR and NBS domains (extended data Figure 8a-b) and the protein sequence of the previously cloned NLR gene showed that the predicted Lr/Yr548 protein was most closely related to the Tsn1 ToxA-sensitive protein of wheat. However, when the authors analyzed using only the CC domain, they found that the sequence similarity was low (Figure 3b), suggesting that this domain differs from the CC domain of other NLRs.
Extended Data Figure 8. Phylogenetic relationship between Lr/Yr548 and other disease-resistant NLR genes in wheat, barley, rye and their wild relatives.
Next, the authors generated an SNP matrix of Lr/Yr548 CDS using the original AgRenSeq reads from 244 tall goat grass strains and 193 sharon goat grass lines in this study, as well as the new tall goat grass reference genome (v2.0). After filtering the missing data, the final matrix contained 57 tall goat grass genotypes and 71 sharon goat grass genotypes (Extended Data Table 6) for a total of 300 SNPs. The guide-maximal likelihood tree obtained by the genotype matrix shows a highly mixed branching of both species (Figure 3C). In addition, many clades contain genotypes from geographically distant collection sites, and genotypes from the same collection site are often located in different clades. This pattern suggests that Lr/Yr548 experienced multiple gene flow events between the two species. In addition, many clades contain genotypes from geographically distant collection sites, and genotypes from the same collection site are often located in different clades, a trait that is common in the R gene in natural populations.
The authors overlaid known leaf rust resistance data onto the phylogenetic tree and found that 57 genotypes were resistant, 18 were susceptible, and 5 had mixed reactions. Most susceptible genotypes clustered on a single clade with a high rate of evolution, suggesting a loss of root resistance in that clade and a removal of selection pressure.
Figure 3. Origin and evolutionary dynamics of Lr/YR548.
Lr/Yr548 confers quantitative resistance in seedlings and adult plants
To gain insight into the resistance patterns conferred by Lr/Yr548, the authors inoculated wild-type Fielder and transgenic lines containing the Lr/Yr548 gene in tall goat grass with Pt and Pst subraces, and then stained leaf samples with FITC-tagged wheat germ lectin (WGA), which specifically stained the fungal cell wall. Pt and Pst races are able to penetrate and grow in the leaf tissues of resistant plants, but the fungus develops significantly more slowly, and the fungus occupies a much smaller area than the susceptible cultivar Fielder (Fig. 4a–c). The authors observed similar results on susceptible and resistant tall goat grass plants (Extended Data Figure 9).
Figure 4. Lr/Yr548 confers resistance in seedlings and adult plants.
Extended Data Figure 9. Seedling reaction 7 days after infection with Pt Racee #526-24.
In addition, detection of reactive oxygen species (ROS) with diaminobenzidine (DAB) staining showed that the transgenic plants accumulated lower levels of ROS than wild-type Fielder plants, however DAB staining occupied more space, indicating the presence of a quantitative defense response (Extended Data Figure 10). Although the expression of Lr/Yr548 increased with plant age (extended data Fig. 6c) and resistance varied with expression levels (extended data Fig. 6d), there was no difference in disease response between seedlings and adult plants (flag leaf stage) (Fig. 4a). In addition, transgenic Fielder plants were susceptible to stem rust (Pgt) and powdery mildew (Blumeria graminis) pathogens (not shown). The authors concluded that Lr/Yr548 hinders the infection of Pt and Pst by slowing the development of pathogenic bacteria (Figure 4b-c and extended data Figure 9), and while it does not completely prevent the progression of the fungus, it can prevent disease development in seedlings and mature plants. In addition, the resistance conferred by Lr/Yr548 was specific to the pathogens of leaf rust and stripe rust.
Extended Data Figure 10. Seedling response to infection with Pt race #526-24 and Pst race #5006.
Wheat multi-omics website: http://wheatomics.sdau.edu.cn submission, cooperation, etc. E-mail: [email protected] WeChat group: Click on the exchange group of the wheat research alliance to join the group