Xiu Yang,Yunfeng Jiang,Xianghai Yu,Haipeng Zhang,Yuqi Wang,Fangnian Guan,Li Long,Hao Li,Wei Li,Qiantao Jiang,Jirui Wang,Yuming Wei,Jian Ma,Houyang Kang,Pengfei Qi,Qiang Xu,Meng Deng,Yazhou Zhang,Youliang Zheng,Yonghong Zhou,,Guoyue Chen,

a State Key Laboratory of Crop Gene Exploitation and Utilization in Southwest China,Wenjiang,Chengdu 611130,Sichuan,China

b Triticeae Research Institute,Sichuan Agricultural University,Wenjiang,Chengdu 611130,Sichuan,China

c College of Agronomy,Sichuan Agricultural University,Wenjiang,Chengdu 611130,Sichuan,China

Keywords: Stripe rust All-stage resistance (ASR)BSE-Seq Transcriptome analyses Candidate genes

ABSTRACT Stripe rust,caused by Puccinia striiformis f.sp. tritici (Pst),threatens wheat production worldwide,and resistant varieties tend to become susceptible after a period of cultivation owing to the variation of pathogen races.In this study,a new resistance gene against Pst race CYR34 was identified and predicted using the descendants of a cross between AS1676,a highly resistant Chinese landrace,and Avocet S,a susceptible cultivar.From a heterozygous plant from a F7 recombinant inbred line(RIL)population lacking the Yr18 gene,a near-isogenic line(NIL)population was developed to map the resistance gene.An allstage resistance gene, YrAS1676,was identified on chromosome arm 1AL via bulked-segregant exomecapture sequencing.By analyzing a large NIL population consisting of 6537 plants,the gene was further mapped to the marker interval between KA1A_485.36 and KA1A_490.13,spanning 485.36-490.13 Mb on 1AL.A total of 66 annotated genes have been reported in this region.To characterize and predict the candidate gene(s),an RNA-seq was performed using NIL-R and NIL-S seedlings 3 days after CYR34 inoculation.Compared to NIL-S plants,NIL-R plants showed stronger immune reaction and higher expression levels of genes encoding pathogenesis-associated proteins.These differences may help to explain why NIL-R plants were more resistant to Pst race CYR34 than NIL-S plants.By combining fine-mapping and transcriptome sequencing,a calcium-dependent protein kinase gene was finally predicted as the potential candidate gene of YrAS1676.This gene contained a single-nucleotide polymorphism.The candidate gene was more highly expressed in NIL-R than in NIL-S plants.In field experiments with Pst challenge,the YrAS1676 genotype showed mitigation of disease damage and yield loss without adverse effects on tested agronomic traits.These results suggest that YrAS1676 has potential use in wheat stripe rust resistance breeding.

1.Introduction

As a staple food crop,wheat(Triticum aestivumL.)provides 20%of total global grain yield (https://www.fao.org/faostat).Various diseases limit wheat production,of which stripe rust,caused byPuccinia striiformisf.sp.tritici(Pst),is the predominant wheat disease worldwide [1].In China,frequent pandemics of stripe rust cause large yield losses: respectively 6.0,3.2,1.8,and 1.3 Mt in 1950,1964,1990,and 2002[2,3].Although stripe rust can be controlled by the application of fungicides or the use of hosting resistance,host resistance has been adopted as the most preferred strategy in stripe rust control and wheat breeding [4,5].

Resistance againstPstis categorized as all-stage resistance(ASR) and adult-plant resistance (APR) [6,7].ASR is observed at the seedling stage and persists through all growth stages.However,it is easily overcome by new virulent pathogen races because it is race-specific and controlled by a single gene or a few large-effect genes [6].In contrast,APR operates only in the adult plant stage and lacks any race-specific character[8].APR is controlled by multiple minor genes,with plants showing durable resistance when several genes are present but low resistance when only a single gene is present[8].Of hundreds of disease-resistance genes or loci(Rgenes) identified in wheat [9],most are categorized as ASR genes.Among stripe rust resistance genes,six ASR genes,namelyYr5,Yr7,Yr15,YrAS2388,YrSP,andYrU1,have been cloned and found to encode either a putative kinase (Yr15) or a nucleotidebinding site and leucine-rich repeat protein[10-13].ManyRgenes are ineffective against newly prevalentPstraces or have not yet been widely incorporated in wheat cultivars.Some previously effectiveRgenes have shown loss of function when challenged by CYR34,a newPstrace prevalent throughout China [14].Identification of newRgenes and elucidation of their mechanisms of action are desirable for wheat breeding.

Exome capture sequencing coupled with bulked-segregant analysis can identify wheat genes responsible for phenotypes of interest on the basis of forward genetic analyses[15,16].Transcriptome sequencing using next-generation sequencing technology has been widely performed to identity candidate genes and the mechanisms underlying wheat-pathogen interactions [17-19].By combining mapping and transcriptome sequencing analyses,candidate genes can be identified and potentially linked to a specific trait or disease resistance [20-22].

We identified a Chinese wheat landrace that was highly resistant to stripe rust at the seedling and adult stages.A nearisogenic line (NIL) population for stripe rust resistance was developed from a heterozygous line.The aims of the present study were to(1)identify and fine-map the ASR gene in the NIL population,(2)investigate the mechanisms underlying theYrAS1676response to stripe rust using histological observations and transcriptome analysis,(3) identify candidate genes by combining fine-mapping and transcriptome sequencing,and (4) evaluate the genetic effects ofYrAS1676on major agronomic traits to assess its utility for breeding.

2.Materials and methods

2.1.Plant materials

The stripe rust-resistant Chinese wheat landrace AS1676 was crossed with the susceptible cultivar Avocet S (AvS) to produce an F7recombinant inbred line (RIL) population of 212 lines.A NIL population of 241 lines was derived from a heterozygous line that was selected from the F7RILs (Fig.1A).A large NIL population of 6537 plants was derived from 10 heterozygous plants in the NIL population.Mingxian 169,SY95-71,and AvS were used as susceptible controls.

Fig.1.Construction of a mapping population and evaluation of stripe rust resistance.(A)Diagram of the pedigree for the construction of a mapping population.(B)Seedling phenotype of parents and NILs in response to infection by race CYR32 and CYR34(B).(C)Adult plant phenotype of NILs in the field.(D)Microscopic observation of Pst growth.

2.2.Stripe rust assessment

NIL-R and NIL-S plants were artificially inoculated withPstraces CYR32 and CYR34 under controlled greenhouse conditions.NIL-R,NIL-S,and Mingxian 169 were seeded in plastic pots and grown under 15-18 °C and 16-h light/8-h dark conditions.Seedlings were inoculated with urediniospores at the two-leaf stage and incubated at 10 °C for 24 h.Based on field phenotyping in the previous year,10 each resistant and susceptible RIL plants were selected and artificially inoculated withPstCYR34 in the greenhouse.Infection type (IT) was recorded using the 0-9 scale [23]at 18-21 dpi.Stripe rust reaction was designated as resistant(IT ≤6) or susceptible (IT ＞6).

Assessments of adult-plant stripe rust responses were conducted in the RIL population,NIL population,241 lines,and 6537 individual plants in experimental fields in Chongzhou (CZ;30°32′N,103°38′E) and Wenjiang (WJ;30°360′N,103°41′E),Sichuan province,China.Seeds of the RIL (212 lines) and NIL(241 plants) populations were planted in the field in 2019-2020.Seeds of the 241 lines and 6537 plants were planted in the field in 2020-2022.Approximately 20 seeds of each accession were sown in rows that were 2.0 m long with 0.30 m spacing between rows.Plants were artificially inoculated with a mixture of uredinio spores from the predominant ChinesePstraces (namely CYR32,CYR33,CYR34,G22-14,Su11-4,Su11-5,and Su11-7).The highly susceptible lines SY95-71 and AvS were planted every 20 rows to ensure uniform disease development.Adult plant stripe rust reactions were assessed as IT scores,recorded three times at weekly intervals when the susceptible checks were fully covered with uredinia.

2.3.Assessment of agronomic traits under stripe rust stress conditions

Agronomic traits of 20 NIL-R and 20 NIL-S plants were evaluated in four year-location environments in CZ and WJ during the 2020-2022 growing seasons.Eight traits were scored for five spikes from five plants: grain number per spike (GNPS),spike length (SL),spike number (SN),thousand-kernel weight (TKW),plant height (PH),tiller number (TN),flag leaf length (FLL),and spike extension length (SEL).Kernel length (KL) and kernel width(KW) were also measured for 15 randomly selected NIL-R and NIL-S samples.The methods for evaluating agronomic traits were described previously [24,25].

2.4.Histological observations

Leaf samples from NIL-R and NIL-S plants infected withPstrace CYR34 were collected at 1,3,5,and 7 day to analyze the fungal hyphae.The leaves were segmented and submerged in an ethanol:acetic acid (1:1) solution to remove the chlorophyll,followed by autoclaving at 121°C for 5 min[26].The fragments were then treated with chloral hydrate,stained with wheat germ agglutinin Alexa-488 solution (Thermo Fisher Scientific,Waltham,MA,USA),and observed under a fluorescence microscope (Olympus BX63) equipped with a Photometric SenSys DP-70 CCD camera(Olympus,Tokyo,Japan).Adobe Photoshop (Abobe Systems Incorporated,San Jose,CA,USA) was employed for photomicrograph processing.

2.5.Bulked-segregant exome-capture sequencing (BSE-seq) analysis

Genomic DNA was extracted from leaf samples by the CTAB method [27].Resistant and susceptible pools were produced by combining equally the DNA from 30 NIL-R (IT ≤3) and 30 NIL-S(IT ≥7) plants of the NIL population (Table S1).DNA pools were analyzed using the wheat exome-capture sequencing protocol described by Dong et al.[15].Briefly,4.5 mg genomic DNA(35 ng mL-1) was sonicated using a Bioruptor UCD-200 sonicator(Diagenode,Denville,USA) to generate 300-bp fragments.The KAPA Hyperprep kit for Illumina HiSeq (Hoffmann-La Roche Ltd.,Basel,Switzerland) was used to construct pre-libraries.The 2100 Bioanalyzer (Agilent Technologies,Santa Clara,CA,USA) was used to assess library quality.The libraries were quantified with a quantitative real-time polymerase chain reaction (qRT-PCR) assay and then sequenced using the Illumina HiSeq Nova platform to generate 150-bp paired-end reads.The sequence files were deposited in the National Genomics Data Center Genome Sequence Archive(https://ngdc.cncb.ac.cn/) under BioProject ID PRJCA014185.

Using Fastp (v.0.12.4) [28],the raw sequencing data were refined by discarding low-quality reads and adapters.Paired reads were then aligned to the IWGSC Chinese Spring wheat reference genome assembly (v.1.0) (https://wheatgenome.org/) using the default parameters of the BWA(v.0.7.16)program[29].Alignments were processed with SAMtools(v.1.9)[30].Variant calling was performed with GATK (v.4.0.10.1) [31],and reads were filtered by quality(＞30)and depth(≥5)with BCFtools(v.1.9)[32].Gene variants were annotated by SnpEff [33] (IWGSC RefSeq v.1.1,https://wheatgenome.org/).The information of predicted genes is available in the Triticeae-GeneTribe database (https://wheat.cau.edu.cn/TGT/) [34].

The QTL-seq method [35],smoothed G statistics (G′) [36],and Euclidean distance(ED)[37]were used to identify candidate genomic regions associated with stripe rust resistance.Using the average ΔSNP index,G′,and ED4(i.e.,ED to the fourth power) of the single-nucleotide polymorphism (SNP) markers were smoothed across the whole genome using a 10-Mb sliding window and 1-Mb steps.The WheatGmap web-based platform (https://183.223.252.63:3333/) [38] was used for these analyses.

2.6.KASP marker development and genetic map construction

The SNPs from the exome capture sequencing data on chromosome arm 1AL with BSE-seq enrichment interval were converted into Kompetitive Allele Specific PCR (KASP) markers using the Polymarker tool [39].KASP assays were performed on a Bio-Rad CFX96 (Bio-Rad,Hercules,CA,USA) as previously described [40].

The NIL populations (241 and 6537 plants) were used for recombination screening and fine mapping.JoinMap (v.4.0) software [41] with default parameters was used for linkage analysis and genetic map construction.The Kosambi mapping function was used to convert recombination fractions to centimorgans.MapDraw (v.2.1) [42] was used to draw the linkage map.

2.7.Transcriptome analysis

Leaf samples (three biological replicates each) from NIL-R and NIL-S seedlings inoculated withPstrace CYR34 or mockinoculated were collected 72 h post-inoculation,frozen immediately in liquid nitrogen,and stored at -80 °C.Total RNA was extracted with TRIzol reagent (Invitrogen,Carlsbad,CA,USA).The NanoDrop 2000 spectrophotometer(Thermo Fisher Scientific,Waltham,USA) was used to determine the RNA purity and concentration and the 2100 Bioanalyzer was used to assess RNA quality and integrity.RNA sequencing was performed by Beijing Berry Genomics (Beijing,China),which generated 125-and 150-bp pairedend reads.The sequence files were deposited in the National Genomics Data Center Genome Sequence Archive under BioProject ID PRJCA014185.The CLC Genomics Workbench (v.12.0) [43] was used to analyze the clean reads.Differentially expressed genes(DEGs) were identified using the following criteria: |log2(foldchange)| ＞1 and a false discovery rate (FDR) ＜0.05.The Gene Ontology Resource (https://geneontology.org/) was used to identify enriched (Gene Ontology) GO terms among the DEGs with an FDR ＜ 0.05 as the significance threshold.The candidate gene sequences obtained based on RNA-seq were aligned to the 10+wheat reference genomes (https://wheat.cau.edu.cn/TGT/).

To validate the RNA-seq results,eight DEGs were randomly selected for qRT-PCR analysis.Gene-specific primers were designed with Primer3Plus (https://www.primer3plus.com/)(Table S2).The SYBR Premix pro Taq HS qPCR Kit (Accurate Bio Co.,Ltd.,Changsha,Hunan,China) was used for qRT-PCR with a reaction volume of 10 μL containing 100 ng cDNA,10 μmol L-1of each primer,and 2× SYBR Premix pro Taq HS Premix.A CFX96 RealTime PCR System (Bio-Rad) was used according to the manufacturer’s instructions.Relative gene expression levels were calculated by the 2-ΔΔCt method [44],usingTaEFas the internal reference gene.qRT-PCR analysis of each sample was performed in triplicate using three biological replicates per treatment.

2.8.Agronomic trait analysis

The base package in R was used to fit an analysis of variance to the agronomic traits.The ggplot2 package in R was used to prepare box and bar plots.

3.Results

3.1.Assessment of stripe rust resistance

The KASP analysis results showed that AS1676 carriedYr18,which conferred adult-stage resistance(Fig.S1).A functional marker forYr18was previously described by Fang et al.[45].AS1676 showed high seedling resistance (IT=1) to CYR32 and CYR34(Fig.1B).The IT scores varied among the selected RILs in response to inoculation with CYR34 (Fig.S2),suggesting that AS1676 might carry additional ASR gene(s).

To avoid any interference fromYr18,a stripe rust resistance NIL population was developed from a heterozygous line lackingYr18,which was selected from the F7RIL population.NIL-R plants were highly resistant to CYR32 and CYR34 at the seedling stage(Fig.1B).In the field,they were also resistant (IT=1) to mixedPstraces at the adult plant stage.In contrast,NIL-S plants were susceptible at both the seedling and adult stages (Fig.1C).Histological examination showed no differences in hyphal development between NIL-R and NIL-S plants at 1-5 dpi (Fig.1D).However,by 7 dpi,exponential growth of the pathogen was observed only in the NIL-S plants (Fig.1D).

Of the 241 plants in the NIL population that were included in the field experiment,177 were resistant (IT=0-6) and 64 were susceptible (IT=7-9),fitting the expected 3:1 ratio (χ2=0.60,P＞0.05) for segregation at a single locus.The segregation in the 241 family lines also fitted the expected 1:2:1 ratio for a single codominant gene (60 resistant,128 segregating,and 53 susceptible;χ2=1.16,P＞0.05) (Table S3).Accordingly,a single ASR gene was isolated from AS1676 using the NIL population.

3.2.BSE-seq analysis and fine mapping of YrAS1676

Exome capture sequencing generated respectively 51.6 and 60.8 Gb of raw sequencing data for the resistant and susceptible bulks.After filtering,51.2 and 60.4 Gb(99.3%)of clean sequencing data remained,with an average Q20 of 96.2% and Q30 of 88.3%,indicative of the high quality of the sequencing data.The clean reads were aligned to the reference genome sequence (IWGSC RefSeq v.1.0) (Table S4).Of 34,599 SNPs identified in the resistant and susceptible bulks,20,970 SNPs were unique to the resistant bulk.After the analysis involving the ΔSNP index (Fig.2A),G′(Fig.2B),and ED4(Fig.2C) with a 10-Mb sliding window,a major peak significantly associated with stripe rust resistance was identified on chromosome arm 1AL.This peak(in a 350-500 Mb interval) (Fig.2) was designated asYrAS1676.

Fig.2.Identification of genomic regions associated with stripe rust resistance by bulked-segregant analysis coupled with exome capture sequencing analysis.ΔSNP index(A);G′ (B);and ED4 (C) plots generated by bulked-segregant analysis coupled with exome-capture sequencing analysis using a 10-Mb sliding window.G′,smoothed G statistic;ED4,Euclidean distance raised to the fourth power.

To validateYrAS1676,20 new KASP markers were developed for the 400-500 Mb interval on chromosome 1A based on the exome capture sequencing data(Table S5).A genetic map was constructed spanning 14.8 cM using the 20 KASP markers (Fig.3A;Table S5).YrAS1676was preliminarily located between KASP markersKA1A_484.34andKA1A_493.22,an interval corresponding to a physical interval of 484.3-493.2 Mb (Fig.3A;Table S6).Using 6537 plants and nine KASP markers,YrAS1676was more precisely localized to a 1.7 cM region betweenKA1A_485.36andKA1A_490.13,corresponding to a region from 485.3 to 490.2 Mb on chromosome 1A (Fig.3B;Table S7) that cosegregated with six markers (KA1A_486.42,KA1A_487.09,KA1A_487.40,KA1A_487.48,KA1A_488.22,andKA1A_488.62).

Fig.3.Fine mapping of YrAS1676 on wheat chromosome arm 1AL.The genetic linkage map of YrAS1676 constructed with KASP markers genotyped in 241 lines derived from the cross NIL-R/NIL-S and 6537 F2 plants generated from heterozygous F7 RILs;recombinants were classified by marker analysis.The red box indicates candidate gene regions.Green,orange and blue bars represent respectively homozygous susceptible(HS),heterozygous resistant(HP),and homozygous resistant(HR)genotypes.The letters a to w represent 23 recombinants.B,A and H represent respectively susceptible,resistant,and homozygous phenotypes.The numbers 0,1,4,5,8,and 9 represent IT scores.

3.3.Transcriptome analysis

A total of 61.55 million 150-bp paired-end reads were obtained for 12 samples (Table S8).Removal of adapters and low-quality reads left 59.76 million clean reads with average Q20 of 97.43%,Q30 of 93.52%,and GC content of 58.46% (Table S8),reflecting the high quality of the sequencing data.Of the clean reads,99.35% were mapped to the Chinese Spring wheat reference genome (IWGSC RefSeq v.1.1) (Table S8).

A principal component analysis was performed for RNA sequencing.Replicates from the same group clustered together,reflecting the satisfactory repeatability of each treatment.The samples from the infected NIL-R and NIL-S groups (YRPst and YSPst) clustered separately from the mock control samples,indicating that stripe rust infection induced marked changes in gene expression (Fig.S3).Respectively 4636 (2114 upregulated and 2522 downregulated) and 7128 (4017 upregulated and 3111 downregulated) DEGs were identified in the mock control vs.NIL-R and mock control vs.NIL-S comparisons (Fig.4A,B).These DEGs were functionally annotated by GO enrichment analysis(Fig.4C;Tables S9,S10).Most of the 20 most highly enriched GO terms were related to plant disease resistance.Several plant immunity-related GO terms were commonly enriched among the DEGs in the NIL-R and NIL-S plants: response to chitin(GO:0010200),MAP kinase kinase kinase activity (GO:0004709),stress-activated protein kinase signaling cascade (GO:0031098),signal transduction by protein phosphorylation (GO:0023014),phenylalanine ammonia-lyase activity (GO:0045548),abscisic acid-activated signaling pathway (GO:0009738),and ethyleneactivated signaling pathway (GO:0009873).The 1997 DEGs (752 upregulated and 1245 downregulated)unique to the mock control vs.NIL-R comparison were subjected to a GO analysis.Several plant immunity-related GO terms were enriched among these DEGs,including response to chitin(GO:0010200),regulation of salicylic acid biosynthetic process (GO:0080142),and immune response-regulating signaling pathway (GO:0002764).These GO terms were also assigned to the DEGs (2655 upregulated and 1834 downregulated) unique to the mock control vs.NIL-S comparison.The significantly enriched GO terms among the DEGs(3112 upregulated and 529 downregulated) revealed by the NILR vs.NIL-S comparison included ncRNA metabolic process(GO:0034660),ribonucleoprotein complex biogenesis(GO:0022613),ncRNA processing(GO:0034470),and RNA processing(GO:0006396),indicating that the enriched GO terms were not associated with plant immunity.The first layer of plant innate immunity,namely pathogen-associated molecular patterntriggered immunity (PTI),was activated by thePstinfection of NIL-R and NIL-S plants.

Fig.4.Comparison of transcriptome data from NIL-R and NIL-S.YR/YRPst,DEGs in inoculated and mock-inoculated samples of NIL-R.YS/YSPst,DEGs in inoculated and mockinoculated samples of NIL-S.YR represents mock-inoculated NIL-R.YS represents mock-inoculated NIL-S.YRPst represents inoculated NIL-R.YSPst represents inoculated NILS.Figures describing the DEGs in NILs(A);Venn diagram of DEGs with upregulation and downregulation in NILs(B);GO enrichment analysis of upregulated DEGs in NILs(C);Scatterplot of gene expression,with each dot representing the fold change(log2)value of a gene in NILs(D);Detailed expression profiles of pathogenesis-related protein,color gradient of low(blue)to high(red)(E);Overview of key gene expression and signal transduction in salicylic acid pathway at 72 h post-inoculation.Each box represents a DEG,blue and red colors denote down-and up regulated DEGs.CML,calmodulin-like proteins;MAPK,mitogen-activated protein kinase;NPR1,non-expresser of pathogenesis related;TGA,TGACG motif-binding factor(F);Expression levels of several WRKY family transcription factors in NILs.Values used for the bar chart are fold change(log2)(G).

To further analyze the main difference between resistant and susceptible plants,we compared the NIL-R and NIL-S plants in terms of the expression of key genes involved in disease resistance pathways.Genes encoding serine/threonine-specific protein kinases,pathogenesis-related (PR) proteins,and diseaseresistance proteins were substantially activated in the NIL-R and NIL-S plants.However,genes encoding serine/threonine-specific protein kinases and disease-resistance proteins were more highly expressed in NIL-S plants,whereas genes encoding PR proteins were more highly expressed in NIL-R plants(Fig.4D).Among these PR genes,five PR1 genes (TraesCS7A02G558500,TraesCS5A02G183300,TraesCS5B02G181500,TraesCS7D02G161200,andTraesCS5A02G018200)were expressed at significantly higher levels in NIL-R than in NIL-S plants (Fig.4E).TraesCS7A02G558500andTraesCS5A02G018200belong to homoeologous groups andTraesCS7D02G161200 and TraesCS5A02G183300belong to homoeologous groups,butTraesCS7D02G181500is not homologous to them.The PR genes,which are commonly used as markers for the salicylic acid-mediated activation of systemic acquired resistance,are regulated by WRKY family transcription factors and non-expressor of pathogenesis-related gene(NPR)(Fig.4F).WRKY family transcription factor genes were differentially expressed in NIL-R and NIL-S plants (Fig.4G).TraesCS1B02G440300,TraesCS4B02G170100,TraesCS4D02G172200,andTraesCS1D02G418000were more highly expressed in NIL-R than in NIL-S plants (Fig.4G).In contrast,the expression of three NPR1 genes (TraesCS3A02G105400,TraesCS3B02G123800,andTraesCS3D02G107500) was significantly upregulated only in the NIL-S plants.

To verify the transcriptome sequencing results,four PR genes and four WRKY family transcription factor genes were selected for a qRT-PCR analysis of their expression levels at 0 and 72 h post-inoculation.The qRT-PCR and expression levels based on RNA-seq were correlated (0.58,P＜0.001),indicating that the expression trends for these genes were consistent with the RNAseq result.(Fig.S4).

3.4.Expression profiles of the genes in the candidate region

Based on the annotated genes in the Chinese Spring reference genome (IWGSC RefSeq v.1.1 and IWGSC RefSeq v.2.1),54 lowconfidence genes and 66 high-confidence genes were detected in theYrAS1676candidate region on chromosome 1A,but none of them encoded nucleotide-binding site and leucine-rich repeat domain receptors (Table S11).Of these 54 low-confidence genes,the expression of 33 was undetectable,suggesting that they were unlikely to be associated with stripe rust resistance.The expression of one gene (TraesCS1A02G435600LC) encoding a polynucleotidyl transferase was significantly upregulated in NIL-S plants.The expression ofTraesCS1A02G435000LC[DNA-(apurinic or apyrimidinic site) lyase] was significantly downregulated only in NIL-R plants.TheTraesCS1A02G435200LC[RNA-directed DNA polymerase(reverse transcriptase)-related family protein] expression was significantly upregulated only in NIL-R plants.Of these 66 highconfidence genes,the expression of 33 was undetectable,suggesting that they were unlikely to be associated with stripe rust resistance(Fig.5A,B).Among the 33 expressed genes,the expression of one(TraesCS1A02G290400)encoding uclacyanin 1 was significantly upregulated in NIL-R and NIL-S plants (Fig.5A).The expression ofTraesCS1A02G289500(putative serpin-Z12) was significantly upregulated only in NIL-S plants(Fig.5A).The expressions of bothTraesCS1A02G291000(NETWORKED 2A) andTraesCS1A02G291100(nicotianamine aminotransferase A) were upregulated in NIL-R plants but downregulated in NIL-S plants (Fig.5A).In contrast,the expression ofTraesCS1A02G291800(uncharacterized chloroplastic protein At4g08330) was upregulated in NIL-S plants but downregulated in NIL-R plants (Fig.5A).TheTraesCS1A02G291700(histone H4)expression level was significantly downregulated only in NIL-R plants(Fig.5A).The functions of the remaining expressed genes are unknown (Fig.5A).We analyzed the sequences of these 33 expressed genes in the NIL-S and NIL-R plants(Table S11).Fourteen genes showed the same sequence in NIL-S and NIL-R plants,whereas sequence variations were detected in 15 other genes.However,the alignment with IWGSC RefSeq v.1.0 suggested that variations were exclusive to NIL-S.Examination of theTraesCS1A02G290800,TraesCS1A02G291100,andTraesCS1A02G290500coding regions in NIL-R detected respectively one,two,and one SNPs,but these SNPs did not alter the amino acid sequence.A SNP (G-to-A) was found in the 5′untranslated region of the calcium-dependent protein kinase gene.Because comparison with the 10+wheat reference genomes showed that this variation was unique to NIL-R,this gene was assigned as a candidate gene forYrAS1676.

Fig.5.Expression levels of candidate genes in the candidate region of YrAS1676.DEGs between inoculated and mock-inoculated samples against one set of the NILs.NIL-R and NIL-S;the color key represents the fold change(log2)values(A);A heat map showing the expression levels of candidate genes in NIL-R and NIL-S;the color key represents the RPKM normalized values (B).

3.5.Agronomic trait evaluation

During the 2020-2021 and 2021-2022 growing seasons in CZ and WJ,the agronomic traits of thePst-inoculated NIL-R and NILS lines were compared.The presence ofYrAS1676resulted in higher values(P＜0.001)for GNPS,TKW,KL,and KW in NIL-R than in NILS plants(Fig.6;Table S12).But SL,SN,PH,TN,FLL,and SEL did not differ between NIL-R and NIL-S plants (Fig.6;Table S12).

Fig.6.Agronomic traits of NIL-R and NIL-S plants shown as bar plots.2021WJ,2021Wenjiang;2021CZ,2021Chongzhou;2022WJ,2022Wenjiang;2022CZ,2022Chongzhou.TN,tiller number;SN,spike number;GNPS,grain number per spike;PH,plant height;FLL,flag leaf length;SL,spike length;SEL,spike extension length;TKW,thousand kernel weight;KL,kernel length;KW,kernel width.

4.Discussion

Chinese wheat landraces,which were the dominant cultivars in China until the 1950s,adapted to disease stress conditions in stripe rust epidemic zones via long-term natural and artificial selection.These landraces contain uncharacterized stripe rust resistance genes.However,identification of new resistance genes in these landraces is often hindered by the widespread presence ofYr18[46].The Chinese wheat landrace AS1676 was highly resistant to the predominantPstraces in field trials conducted over multiple years.It was also highly resistant toPstraces CYR32 and CYR34 in seedling tests.To characterizeYrAS1676,a stripe rust resistance NIL population was derived from a heterozygous line lackingYr18selected from the F7RIL population.After the BSE-seq analysis and fine-mapping,YrAS1676was finally localized to a 1.7-cM interval corresponding to the region from 485.36 to 490.13 Mb on chromosome 1A.The cosegregation of six markers in the large NIL population suggests that recombination events are suppressed in theYrAS1676region.Accordingly,it may be difficult to further narrow the candidate region by expanding the population.

Several QTL and ASR genes for stripe rust resistance have been identified on chromosome arm 1AL (Table S13):QYr.hebau-1AL[47],QYr.tam-1AL[48],QYr.caas-1AL[49],QYrPI197734.wgp-1A[50],QYr.cim-1AL[51],QYr-1A_IWA5754[52],QYr.wsu-1A.1[53],QRYr1A.1[54],QYr.wsu-1A.2[53],QYr.inra-1A[55],QYrst.orr-1AL[56],YrHA[57],andYrXH-1AL[58].They are located at the distal end of 1AL except forQYr.tam-1AL,YrHA,andYrXH-1AL.QYr.tam-1AL,identified in cultivar TAM112,was mapped to the marker interval betweenXwPt-5167andXwPt-1011corresponding to a physical interval of 382.79-583.32 Mb[48].This QTL is associated with stripe rust resistance in adult plants.YrHAis located close to markersXwmc469andXgwm497(corresponding to the interval from 242.83 to 551.15 Mb) in the translocation line H9014-121-5-5-9 [57].TheYrAS1676region overlaps the physical intervals ofQYr.tam-1ALandYrHA,but the genes are likely different,in view of the diversity in the type and origin of resistance.Jiang et al.[58]recently identified an ASR gene (YrXH-1AL) in the Chinese wheat landrace Xiaohemai.This gene was mapped to almost the same region asYrAS1676.Xiaohemai carryingYrXH-1ALalso showed resistance to CYR34 and CYR32.The markersKA1A_485.36,KA1A_486.42,andKA1A_487.40shared byYrAS1676andYrXH-1ALsuggest that they may be a gene or an allele.

Plants have evolved innate immunity,which is a unique defense system that provides protection against pathogens.This system consists of PTI and effector-triggered immunity (ETI).PTI,which represents the first layer of immunity,is activated when plant cell-surface pattern recognition receptors perceive conserved molecules or structures from the pathogen.However,pathogens can secrete effector proteins inside host cells to evade PTI.The effector proteins disrupt plant defense responses,thereby activating ETI,the second layer of immunity.In ETI,pathogen-secreted effectors are perceived by plant proteins encoded byRgenes[59].Most of the clonedRgenes in crops belong to the nucleotide-binding site and leucine-rich repeat (NBS-LRR) gene family [60] and typically mediate disease resistance via the ETI pathway.The finding of the present study that several plant immunity-associated pathways were induced in NIL-R and NIL-S plants implies that PTI was activated in the NIL-R and NIL-S plants infected withPst.NPR1 gene expression levels were higher in NIL-S plants than in NIL-R plants,whereas the opposite trend was detected for the expression of the downstream WRKY transcription factor genes.PR gene expression was higher in NIL-R than in NIL-S plants.The accumulation of PR proteins is closely associated with the induction of the hypersensitive response,the accumulation of the plant hormone salicylic acid,and systemic acquired resistance,which may explain the difference in stripe rust resistance between NIL-R and NIL-S plants.

It is likely thatYrAS1676is anRgene that confers resistance via the ETI pathway.However,none of the candidate genes were identified as NBS-LRR genes in the Chinese Spring reference genome or the 10+wheat reference genomes.With the exception of a calcium-dependent protein kinase gene,specific sequence variations were not detected in the candidate genes following an examination of the 10+wheat reference genomes and the Chinese Spring reference genome.Among the candidate genes,the gene encoding 40S ribosomal protein S23 was more highly expressed in NIL-R than in NIL-S,but the comparison of this gene between NIL-S and NIL-R plants based on transcriptome data revealed a lack of sequence variation.In plants,the 40S ribosomal protein S23 contributes primarily to the synthesis of proteins [61].A calciumdependent protein kinase gene is currently the most promising candidate gene.Previous studies [62-64] confirmed that this type of protein kinase can mediate disease resistance by inducing the accumulation of salicylic acid and the expression of PR genes in plants.In wheat,calcium-dependent protein kinases can trigger a hypersensitive response followingPstinfection [19].This candidate gene awaits further validation.Because the candidate genes described in this study were identified on the basis of the released wheat genomes,it is possible that there areRgenes other than AS1617 that were not considered.

The expression of genes mediating disease resistance often results in agronomically undesirable traits [65].Although the spring wheat accession PI 178759 is resistant to stripe rust under field conditions,it is agronomically undesirable [66].Chen et al.[67] reported that lines carrying theLr34/Yr18resistance allele were taller,matured earlier,and yielded less grain with lower test weights than lines lackingLr34/Yr18.In the present study,the yield-associated traits GNPS,TKW,KL,and KW were significantly higher in lines withYrAS1676than in those without,but there were no differences in SN,PH,TN,FLL,and SEL,a finding consistent with those of previous studies [68-71].In a recent study by Ye et al.[71],the GNPS and TKW values in stripe rust-resistant accessions decreased by respectively 1.5% and 5.6%,whereas the corresponding values decreased by 2.7% and 10.2% in susceptible accessions.Thus,the differences in these agronomic traits may be becauseYrAS1676confers high stripe rust resistance,minimizing disease damage and stabilizing yield.YrAS1676may be useful for selecting and breeding new wheat varieties that are highly resistant to stripe rust and exhibit ideal agronomic traits.

CRediT authorship contribution statement

Xiu Yang:Conceptualization,Data Curation,Investigation,Methodology,Formal analysis,Visualization,Validation,Writingoriginal draft preparation.Yunfeng Jiang:Conceptualization,Formal analysis,Methodology,Visualization,Investigation,Writing-Reviewing and Editing.Xianghai Yu:Data curation,Formal analysis,Investigation,Visualization,Validation.Haipeng Zhang:Investigation,Software,Visualization.Yuqi Wang:Investigation,Formal analysis,Software,Visualization.Fangnian Guan:Formal analysis,Software,Visualization.Li Long:Formal analysis,Methodology,Software.Hao Li:Formal analysis,Methodology,Software.Wei Li:Methodology,Software.Qiantao Jiang:Methodology,Software.Jirui Wang:Methodology,Software.Yuming Wei:Methodology,Software.Jian Ma:Methodology,Software.Houyang Kang:Methodology,Software.Pengfei Qi:Methodology,Software.Qiang Xu:Methodology,Software.Meng Deng:Resources.Yazhou Zhang:Resources.Youliang Zheng:Resources.Yonghong Zhou:Supervision,Conceptualization.Writing-Reviewing and Editing.Guoyue Chen:Supervision,Writing-Reviewing and Editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgments

This study was supported by the Major Program of National Agricultural Science and Technology of China (NK20220607),the National Natural Science Foundation of China (32272059 and 31971883)and the Science and Technology Department of Sichuan Province (2022ZDZX0014,2021YFYZ0002,2021YJ0297,and 23NSFTD0045).We are also grateful to Prof.Qiu-Zhen Jia (Plant Protection Research Institute,Gansu Academy of Agricultural Sciences,Lanzhou,Gansu,China)for providing the stripe rust races.

Appendix A.Supplementary data

Supplementary data for this article can be found online at https://doi.org/10.1016/j.cj.2023.04.011.