The wheat stripe rust resistance gene YrNAM is Yr10

arising from F. Ni et al. Nature Communications https://doi.org/10.1038/s41467-023-39993-2 (2023)

The wheat stripe rust resistance gene Yr10 is defined by a unique array of responses to multiple isolates of the stripe rust pathogen Puccinia striiformis f. sp. tritici (Pst), yet has been attributed to two different genes in the wheat host. Sequencing of putative Yr10 mutants by Ni et al. identified changes in a gene named YrNAM that was closely linked to gene Yr10_CG, reported earlier as the Yr10 candidate. Independent analysis of mutants in a universal reference Yr10 stock by us identified changes in YrNAM. An agreed simplification of two separate gene identities conferring a common specificity to a single Yr10 designation will benefit the wider cereal research community.

Yr10 was originally identified in Turkish wheat landrace PI 178383 and is located on chromosome arm 1BS¹. It was first deployed in the cultivar (cv.) Moro (CI 013740) in the Pacific Northwest, USA². Moro and a Moro-derived near isogenic line (NIL) Avocet S + Yr10 (Avocet S*6/Moro)³ were globally distributed and continue to be included in differential sets for phenotyping Pst in Australasia, Europe, India and North America. Two publications reported the putative cloning of Yr10 as homologs of two closely linked genes, TraesCS1B03G0000200, an NLR designated Yr10_CG, Genbank accession AF149112⁴, and more recently homologs of TraesCS1B03G0003600 LC.1 and TraesCS1B03G0003500 LC.1, Genbank accession OP490604, was designated YrNAM⁵. We believe this situation is untenable and that one of these linked genes cannot be Yr10. Here, we present evidence that YrNAM is Yr10.

The CC–NBS–LRR candidate gene Yr10_CG was first identified following RAPD analysis and genetic mapping, and was used in genetic analysis, gene silencing and transgenesis to demonstrate putative functionality⁴. The identified sequence was present in Moro and several backcross derivatives predicted to have Yr10, along with cv. Jacmar and Triticum spelta 415 shown earlier to carry Yr10⁶. Of 46 plants recovered from scutella bombarded with the Yr10_CG partial genomic clone 4B, only one transgenic individual produced a Moro-like stripe rust response with a Yr10-avirulent Pst isolate, and the same plant was susceptible to a Yr10-virulent isolate. Reliance on one out of 46 independent transgenic events as a basis for identifying Yr10_CG as the causal gene for Yr10 is however inadequate and does not allow for any effects of somaclonal variation. Subsequent tests from an independent study, carried out on additional Yr10_CG cDNA transgenic events, failed to detect the Yr10 resistance phenotype⁷. Virus induced gene silencing of Yr10_CG, while suggesting the possible requirement of Yr10_CG or similar sequence in stripe rust resistance in the specific host used, does not address the sufficiency of Yr10_CG as the causal gene of Yr10-mediated resistance.

In a subsequent study⁷, the Yr10_CG sequence AF149112 was mapped and shown to be located 1.2 cM from Yr10, hence indicating that it was not Yr10. Of 62 wheat lines identified with an AF149112-specific marker, only 10 gave stripe rust responses typical of Moro. In a further experiment, all T1 progeny of cv. Bobwhite transformed with the AF149112 sequence were susceptible to a Pst isolate avirulent on Moro. This publication provided strong evidence that the AF149112 sequence was not the causal gene for Yr10.

More recently, Ni et al.⁵ generated 12 putative YrNAM mutants in two homozygous lines identified in a cross of Moro and a susceptible line. Seven of these mutants were subjected to RNA-seq and transcriptome analysis, and mutations were identified in a gene designated YrNAM, which was located 1.2 cM from the Yr10_CG sequence. YrNAM encodes a non-canonical resistance protein with a NAM (No Apical Meristem) domain and a ZnF (Zinc Finger)-BED domain at the N and C termini, respectively⁵. It is structurally similar to Rph7, a leaf rust resistance gene in barley containing NAC transcription factor (TF) and ZnF-BED domains⁸. Together YrNAM and Rph7 represent a different class of R proteins, which we propose be designated as Triticeae NAC-BED domain (TNB) proteins. Although only two resistance genes have been reported with this structure to date, NAC TF domains feature widely as regulators of plant immunity⁹, and several other cloned resistance genes possess ZnF-BED domains (e.g. Rph15 in barley¹⁰ and Yr5a, Yr5b, and Yr7 in wheat¹¹) albeit in a different clade to that of TNB proteins. It thus appears likely that additional TNB resistance genes remain to be discovered.

All 12 YrNAM mutants were confirmed to encode non-synonymous amino acid changes in the same gene; six in each of the NAM and ZnF-BED protein domains⁵. Concurrent with that study, we produced four EMS-derived Yr10 mutants in Avocet S + Yr10 (Fig. 1). All four mutations were in YrNAM (Fig. 1b, c) and not in Yr10_CG, with supporting evidence from mutational genomics of enriched NLR sequences (MutRenSeq) including AF149112 previously reported to encode Yr10_CG. MutRenSeq analysis did not detect any mutations in Yr10_CG. This corroborated the previous high resolution mapping study distinguishing Yr10 from Yr10_CG⁷, and subsequent validation by YrNAM transgenesis⁵. Interestingly, three of the susceptible mutants coincided with those reported by Ni et al.⁵ at amino acid positions G193E (independent mutants M6225 & M6227) and D155N (M6211) whereas a fourth mutation (M6220) is a putative intron splice junction mutation presumably affecting the last exon with the ZnF-BED domain. Amino acid conservation values were calculated using the EVcouplings server (https://evcouplings.org)¹² on 6268 NAM domain-containing sequences retrieved and aligned with Yr10. The G193E mutation occurs at a very highly conserved amino acid, G193 (Fig. 1d), with 97.2% conservation across the set of 6282 homologs, and this together with the independent generation of loss-of-function mutations in the study by Ni et al.⁵ for this residue indicate that G193 plays a critical role in NAM domain function.

Fig. 1: YrNAM mutations support identity as Yr10.

a Seedling leaves of cv. Moro, YrVav7089 (backcross derivative of Triticum vavilovii AUS22498), AvS+Yr10, M6225 (AvS+Yr10 mutant), and Avocet S (AvS) infected with Yr10 virulent- (culture no. 598, race 150 E16 A +) (left) and Yr10-avirulent (culture no. 674, race 239 E237 A-17 + 33 +) (right) Puccinia striiformis f. sp. tritici isolates. All five lines were susceptible to the Yr10-virulent isolate. Moro, YrVav7089, and AvS+Yr10 were resistant whilst mutant M6225 and AvS were susceptible to the Yr10-avirulent isolate, respectively. b Schematic of the protein structure of Yr10 (redrawn from Ni et al.⁵), which includes NAM and ZnF-BED protein domains. Our EMS-induced AvS+Yr10 mutation line numbers are in blue and non-synonymous amino acid substitutions are in red, while the intron splice junction mutation base substitution is given in italicized red. Underlined mutations were also independently generated and reported by Ni et al.⁵ Mutations in black are non-synonymous mutations reported for YrNAM⁵. The mutation sites given are relative to ‘ATG’ for nucleotides, or methionine for amino acids. c AlphaFold-predicted structure of Yr10 and positions of mutations D155N (blue residue) and G193E (yellow residue). NAM and ZnF-BED domains are colored magenta and green, respectively. d EVcouplings output across a window centered on highly conserved amino acid G193 (marked by asterisk) showing sequence conservation, sequence logo and consensus sequence for a set of 6268 putative NAM-domain containing proteins. G193 is conserved across 6090 of the 6268 sequences retrieved. Source data are provided as a Source Data file.

There were apparent problems concerning the phenotyping of transgenic plant stripe rust responses, as wildtype Fielder was not fully susceptible as expected for an isolate of Pst race CYR34⁵. We suggest that was due to technical problems associated with the experiments rather than suggested combined effects of defeated genes Yr6 in Fielder and Yr9 in CB037 as proposed by Ni et al.⁵.

Ni et al.⁵ observed that YrNAM was rare in the wheat gene pool. Our results, based on rust response phenotyping with the recent and differentiating Yr10-avirulent (culture no. 674, race 239 E237 A-17 + 33 +) and Yr10-virulent (culture no. 598, race 150 E16 A +) Pst isolates and YrNAM gene-specific sequence assays, confirmed that Yr10 is present in Iranian accession T. spelta 415 (Plant Breeding Institute accession C89.19)⁶, and T. vavilovii AUS22498¹³ and four backcross derivatives (Fig. 1a). Ni et al.⁵ also suggested that the widespread absence of YrNAM in contemporary wheat cultivars was due to its close linkage with the brown glume color, considered as an undesirable trait. There is no evidence that gene Rg1 for red/brown glumes causes negative yield or grain quality effects, rather its absence appears attributable to cosmetic factors of breeder and/or farmer choice. Due to the close linkage of Rg1 and Yr10 in accession PI 178383, most cultivar derivatives of this accession will have brown glume color; however, Rg1 is not present in T. spelta 415 and T. vaviloii AUS22498.

In conclusion, the current evidence indicates that Yr10 is YrNAM, a gene with Triticeae NAC-BED domain (TNB) domain architecture similar to that reported for barley leaf rust resistance gene Rph7⁸ Together, these genes represent a different class of R proteins. We propose the simplification of two separate gene identities conferring a common specificity to a single designation, Yr10, for the benefit of the wider winter cereal research community.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.