The Gene Sr33, an Ortholog of Barley Mla Genes, Encodes Resistance to Wheat Stem Rust Race Ug99

See allHide authors and affiliations

Science  16 Aug 2013:
Vol. 341, Issue 6147, pp. 786-788
DOI: 10.1126/science.1239028

Resistance May Not Be Futile

Recently, Ug99, a particularly devastating strain of wheat stem rust fungus, has emerged, which could potentially threaten food security. Now, two genes have been cloned that offer resistance to Ug99. Saintenac et al. (p. 783, published online 27 June) cloned Sr35 from Triticum monococcum, a diploid wheat species not often cultivated. Periyannan et al. (p. 786, published online 27 June) cloned Sr33 from Aegilops tauschii, a diploid wild grass that contributed to the hexaploid genome of cultivated wheat. The genes both encode proteins that show features typical of other disease resistance proteins and offer opportunities to slow the pace of Ug99 progression.


Wheat stem rust, caused by the fungus Puccinia graminis f. sp. tritici, afflicts bread wheat (Triticum aestivum). New virulent races collectively referred to as “Ug99” have emerged, which threaten global wheat production. The wheat gene Sr33, introgressed from the wild relative Aegilops tauschii into bread wheat, confers resistance to diverse stem rust races, including the Ug99 race group. We cloned Sr33, which encodes a coiled-coil, nucleotide-binding, leucine-rich repeat protein. Sr33 is orthologous to the barley (Hordeum vulgare) Mla mildew resistance genes that confer resistance to Blumeria graminis f. sp. hordei. The wheat Sr33 gene functions independently of RAR1, SGT1, and HSP90 chaperones. Haplotype analysis from diverse collections of Ae. tauschii placed the origin of Sr33 resistance near the southern coast of the Caspian Sea.

Stem rust [Puccinia graminis f. sp. tritici (Pgt)] of wheat is a major threat to global food security. Continued adaptation of the pathogen necessitates continued development of new Pgt-resistant wheat varieties (1). A Pgt race, Ug99 or TTKSK, identified in Uganda in 1999, was virulent on 90% of wheat cultivars grown globally, including those carrying the stem rust resistance (Sr) 31 resistance gene, which hitherto had been widely deployed and effective for more than 30 years (2, 3). Ug99 and subsequent mutational derivatives that overcame additional resistance genes raised concerns of a disease epidemic that could devastate wheat crops, which provide 20% of the world’s caloric intake. More than 50 Pgt resistance (R) loci, including those introgressed from wild relatives such as Aegilops tauschii, have been cataloged in wheat.

The Pgt R gene, Sr33, discovered from Ae. tauschii, the diploid progenitor of the D genome in hexaploid wheat (4, 5), was introgressed into common wheat (Triticum aestivum, genomes AABBDD). There, Sr33 provides a valuable, intermediate level of resistance to diverse Pgt races, including the Ug99 lineage (6). To isolate the Sr33 gene, we used a single-chromosome substitution genetic stock, CS1D5405, which has chromosome 1D of wheat cv Chinese Spring (CS) replaced by the corresponding chromosome bearing Sr33 from Ae. tauschii accession RL5288, to generate a recombinant inbred family segregating for Sr33 (5). We identified 30 progeny, from an estimated 2850 gametes of this family, that contained recombination events between expressed sequence tag markers BE405778 and BE499711, which flank the 1 cM region that contains Sr33.

We used an amplified fragment-length polymorphism (AFLP) marker derived from a dehydrin gene located 0.04 cM from Sr33 to initiate the construction of a physical map of the Sr33 locus by screening a bacterial artificial chromosome (BAC) library of Ae. tauschii accession AL8/78 (7). A BAC contig, ctg4713, was identified that contained several genes, including a coiled-coil nucleotide-binding leucine-rich repeat (CNL) gene designated as Ae. tauschii resistance gene analog 1a (AetRGA1a) (Fig. 1). We rescreened the BAC library with an AetRGA1a probe to identify contig ctg5455, which contained three additional RGA1 members (designated as AetRGA1b-d) that cosegregate with Sr33 (Fig. 1). We screened a second BAC library from Ae. tauschii accession AUS18913 (8) to identify two additional cosegregating RGA1 members (AetRGA1e and AetRGA1f) and two other RGAs unrelated to the RGA1 family (designated as AetRGA2a and AetRGA3a). The three RGA families (RGA1 through 3) present at the Sr33 locus are syntenic and orthologous to three CNL genes (RGH1 through 3) present at the barley Mla locus. The RGH1 family includes the Mla gene, which provides resistance to Blumeria graminis f. sp. hordei (Bgh), the causal agent of barley powdery mildew (9). More than 30 variants of the Mla gene exist, each conferring specific resistance to different races of Bgt (10). The cDNA of AetRGA2a carries an unusual C terminal, absent from the related barley gene (RGH2a), that has 80% identity to the Exo70 subunit of exocyst complex present in T. urartu (11) (fig. S1).

Fig. 1 Synteny between (C) wheat (CS1D5405) 1DS region carrying Sr33 and (A) Ae. tauschii (AL8/78) BAC contigs, (B) Ae. tauschii (AUS18913) BACs, (D) barley, (E) Brachypodium, and (F) rice.

The ovals in orange represent the members of AetRGA1 class, in yellow for AetRGA2, and in brown and gray the pseudo genes AetRGA3a and AetRGA1f, respectively. The Sr33 gene AetRGA1e is indicated by the red oval. Numbers in blue indicate the number of recombinants among 2850 gametes and in red the physical distance in kilobases. Numbers in black indicate the Ae. tauschii BACs. The order of AetRGA1d through AetRGA1d in CS1D5405 was inferred from the mapping information on the small interstitial deletion mutant E5 and BAC sequences from accessions AL8/78 and AUS18913.

We mutagenized CS1D5405 with ethyl methane sulfonate (EMS) and identified nine mutants that had lost Sr33 resistance to Ug99 (table S1). Five mutants (E1 to E5) contained deletions that removed AetRGA1b, AetRGA1c, AetRGA1e, AetRGA2a, and AetRGA3a (Fig. 2). We found no evidence for deletions of these five genes in the remaining four mutants (E6 to E9), but those four mutants did show nucleotide changes relative to the wild-type sequence in AetRGA1e. Mutants E6 and E7 had single, unique missense mutations in the P-loop domain of the predicted protein, whereas mutants E8 and E9 had missense mutations in RNBS-B and GLPL motifs of the nucleotide-binding site (NBS) domain, respectively (fig. S2). To further confirm that AetRGA1e is the Sr33 gene, we produced 20 independent AetRGA1e transgenic lines in the Fielder wheat cultivar, which is susceptible to the Australian Pgt race 98-1,2,3,5,6 (Fig. 2). When challenged with Pgt, each transgenic line exhibited a partial-resistance phenotype characteristic of Sr33. In subsequent generations, this resistance phenotype cosegregated with the AetRGA1e transgene, which we hereafter refer to as the Sr33 gene (Fig. 2).

Fig. 2 (A) Groups of Sr33-susceptible mutants generated through EMS treatment.

Open bars indicate the length of chromosome 1DS segment lost due to mutation, and the stars in red represent the single-nucleotide polymorphism (SNP) change. Group I consists of mutants with large chromosome segments; group II is mutants with a small interstitial deletion; and group III represents the four mutants carrying SNP changes. (B) Stem rust response of Sr33 carrying line (CS1D5405, designated as CS-Sr33) and the EMS mutants against Ug99 (TTKSK). Mutant lines with the short interstitial deletion (E5) and with SNP changes (E6, E7, E8, and E9) show a clear susceptible reaction as seen on CS. (C) Rust infection response of transgenic wheat cv Fielder against the Australian stem rust race 98-1, 2, 3, 5, and 6. Fielder sib lines without Sr33 show clear susceptibility, and lines carrying the Sr33 genomic clone show the typical Sr33-mediated stem rust resistance.

RNA-based polymerase chain reaction analyses indicated that Sr33 encodes six exons (fig. S2), and the predicted protein contains motifs that are conserved in barley and T. monococcum mildew A (MLA) proteins and other CNL proteins (10, 12). The SR33 protein shows greatest similarity (86%) to the MLA-like protein present in T. monococcum (TmMLA) but shows little homology to previously characterized wheat leaf rust (P. triticina) resistance proteins leaf rust resistance (LR) 1, LR10 and LR21, which are also members of the CNL protein class (fig. S3). Among the barley MLA family, SR33 is most similar to MLA34 (80% similarity).

Some MLA proteins of barley require the chaperone proteins RAR1, SGT1, and HSP90 for function (13), as does the wheat LR21 protein (14). No compromised Sr33 resistance was apparent when these three chaperone genes were each independently silenced in CS1D5405 using a viral-induced gene silencing (VIGS) assay (fig. S4) that reduced the transcript accumulation of these genes by 50 to 80% (table S2). In contrast, targeted degradation of Sr33 transcripts by VIGS resulted in increased susceptibility to Pgt infection. In addition, no evidence of direct interaction between SR33 and these chaperones was observed by yeast two-hybrid assays (fig. S5), providing genetic and biochemical evidence that SR33 immunity function is independent of this chaperone complex. Yeast hybrid assays also showed that the coiled-coil (CC) domain of SR33, unlike MLA10, do not homodimerize and do not interact with the WRKY transcription factors that interact with both MLA10 and TmMLA1 (fig. S5). Thus, the apparent structural similarities between SR33 and MLA proteins do not predict functional requirements.

CNL proteins frequently induce cell death in response to infection with an avirulent pathogen, which can be observed microscopically as autofluorescence of dead cells. Our microscopic analysis, however, did not identify an extensive hypersensitive cell death response associated with the partial resistance conferred by Sr33 compared with the response elicited by the more effective stem rust resistance gene Sr45. The majority of Pgt infection sites on the Sr45-containing genotype CS1D5406 were surrounded by autofluorescent cells, indicative of cell death, whereas very little autofluorescence was observed on the Sr33-containing genotype (fig. S6).

We used the Sr33 sequence to search for similar sequences from a collection of 368 Ae. tauschii accessions from various geographical locations spanning the natural distribution of the species. Gene sequences were obtained from 36 accessions with no amplification product from the remaining 332 lines, which suggests that the latter accessions either encode highly divergent variants or lack the gene. Of the 36 accessions, 7 contained sequences identical to the original Sr33 source (sequence I). A second sequence variant, present in accession PI603225, differed from sequence I by a single amino acid at position 588. A third sequence variant, identified in 20 accessions, contained six amino acid substitutions in the leucine-rich repeat (LRR) region (fig. S7). A fourth variant, identified in three Russian accessions, contained several amino acid substitutions in both the NBS and LRR regions. A fifth sequence, present in five Iranian accessions, encoded a truncated protein. Sequence variants I to III were present in accessions collected in the southern coastal region of the Caspian Sea. Only sequences I and II out of the five variants conferred Pgt resistance against multiple races (tables S3 and S4). In allelism tests involving PI603225 and the original Sr33 donor, no susceptible plants were obtained in the progeny, further indicating that sequence variants I and II constitute allelic forms (15).

The wheat Sr33, barley, and T. monococcum Mla genes are examples of orthologous CNL genes evolving to recognize divergent pathogen species. Previous genetic studies have mapped Pgt R genes Sr31 and SrR (now designated Sr50) to a locus in rye chromosome 1S where a Mla-related gene family is also present. These rye chromosomal segments have been introgressed into wheat and serve as additional sources of Pgt resistance (16). It is possible that Sr33, Sr31, and Sr50 constitute a homeologous set of Mla-like Pgt R genes. Ug99 inflicts heavy damage to plants with only Sr31 but is avirulent on plants with Sr33 and Sr50. Thus, the isolation of these Pgt R genes may lead to better understanding of how the plant immune response adapts to virulence changes in Pgt.

Combining resistance genes to develop durable resistance is the prevailing strategy for gene deployment in wheat. We have previously demonstrated additive resistance when Sr33 was combined with Sr2, an adult plant, partial, race-nonspecific Pgt gene that has provided durable resistance for more than 70 years (17). Preferably, Sr33 should be deployed together with genes like Sr2 to maintain its longevity. No Pgt race is virulent on plants containing both Sr33 and the Sr35 gene identified in the companion paper (18), because collectively these two genes provide resistance to all known Pgt races when codeployed. This feature, coupled with the strong immune response of Sr35 against the Ug99 group of races, makes it very attractive to combine Sr33 and Sr35 as cisgenes at a single locus, an approach greatly facilitated by recent advances in Agrobacterium-mediated transformation of this species, and transferred into a wheat background carrying the Sr2 gene.

Supplementary Materials

Materials and Methods

Figs. S1 to S7

Tables S1 to S6

References (1939)

References and Notes

  1. Acknowledgments: This work was supported in part by funds provided through a grant from the Borlaug Global Rust Initiative (BGRI) Durable Rust Resistance in Wheat (DRRW) Project (administered by Cornell University with a grant from the Bill & Melinda Gates Foundation and the UK Department for International Development); Office of the Chief Executive (OCE) Post Doctoral Fellowship of CSIRO, Australia; Grains Research and Development Corporation, Australia; Endeavor International Postgraduate Research Scholarship of Department of Education, Science and Training (DEST) and Sydney University, Australia; Australian Centre for International Agricultural Research project (CIM2007/084). X.W. acknowledges support from the Basic Research to Enable Agricultural Development Program of the National Science Foundation (NSF) of the United States (IOS-0965429) and partial support from NSF Plant Genome Research Program grant DBI-0701916 (J.D. is principal investigator). We are grateful to M. Rouse (USDA-ARS Cereal Disease Laboratory) for the tests with Ug99 and derived races. We thank N. Upadhyaya, A. Ashton, S. Chandramohan, L. Viccars, T. Richardson, K. Newell, H. Miah, and X. Xia for technical assistance. Sequences of Resistance Gene Analog (RGA) genes from Ae. tauschii and hexaploid wheats were deposited in GenBank under accession numbers KF031279 to KF031303.
View Abstract

Stay Connected to Science

Navigate This Article