Plant materials
A total of 198 wheat recombinant inbred lines (RILs) of the Yanzhan 1×Neixiang 188 mapping population (obtained from the Chinese Academy of Agricultural Sciences [CAAS]) were used for linkage analysis. The mapping population for this study was planted at the experimental station of the Chengdu Institute of Biology, Chinese Academy of Sciences, in Shifang (SF) and Maerkang (Ma) during the growing seasons of 2015 to 2018, according to local legislation in Sichuan Province (2015-2016, 2016-2017, 2017-2018 at SF; 2016-2017 and 2017-2018 at Ma). Twenty seeds of each accession were planted in a row. The stripe rust-susceptible wheat line Minxian 169 (obtained from the Chengdu Institute of Biology, Chinese Academy of Sciences), a control, was inserted after every 9 rows. Each experiment was arranged in a randomized complete block design, with two replicates, at Shifang and Maerkang from 2015-2018.
Evaluations of lesion-mimic phenotypes and agronomic traits
In total, 198 RILs and 2 parents (Yanzhan 1 and Neixiang 188) were evaluated for their lesion-mimic (LM) phenotype at SF (104°17’ E, 31°13’ N) and Ma (102°11’ E, 31°92’ N) in Sichuan Province from 2015 to 2018. The lesion-mimic phenotypes were arbitrarily subdivided into 5 scores based on flag leaf symptoms according to the methods of Yao et al. [17], with modifications. No visible lesions (specks) were recorded as 0 (the parental phenotype); few specks and low severity (<25%) were recorded as 1; some specks and moderate severity (25%-50%) were recorded as 2; large specks and high severity (50%-75%) were recorded as 3; and a large number of specks and very high severity (>75%) were recorded as 4. Plants with scores of 0 or 1 were considered normal, and those with scores of 2 or higher were classified as having lesion-mimic phenotypes.
Agronomic traits of the RILs were investigated by our team at the Chengdu Institute of Biology, Chinese Academy of Sciences, during the Shifang cropping seasons. These parameters included plant height (PH), spikelet number (SPI), number of sterile spikelets per spike (SSNS), grain number per spike (GNS), 1000-grain weight (TGW), and spike length (SL). Three to five plants of each wheat line were evaluated, and their means were used for analysis.
Evaluation of stripe rust resistance
All 198 RIL lines and the 2 parents were evaluated for stripe rust at Shifang and Maerkang from 2016 to 2018. Mixtures of Pst spores from races Pst-CYR32, Pst-CYR33, Pst-SU11, Pst-Hybrid46 and Pst-G22 (provided by SAAS) were suspended in 0.05% Tween 20 and sprayed onto four-leaf-stage wheat seedlings.
In the adult stage, stripe rust response types (ITs) were identified, and each environment was evaluated at least twice, mainly from 20 weeks to 23 weeks after sowing. Stripe rust infection types (ITs) were evaluated based on typical 0-4 classification systems [38].
Lesion-mimic gene mapping
Seedling leaves of the 198 RIL lines and two parents (Yanzhan 1, Neixiang 188) were collected, and genomic DNA was extracted from each sample using the CTAB method [39]. The quality and quantity of the DNA were determined using 1.0% agarose gel electrophoresis and a spectrophotometer (NanoDrop ND-1000, Thermo Scientific, Wilmington, DE). Two hundred and fifty-two polymorphic SSR markers covering 21 wheat chromosomes were used to genotype the mapping population to identify the chromosomal location of the LM gene (Additional file 1). Information about the SSR markers is available on the Grain Genes website (http://wheat.pw.usda.gov).
Based on the phenotypic evaluations, 10 wheat lines with an LM score of 0 and 10 RILs with an LM score of 4 were used to prepare two bulks representing extreme phenotypes. The DNA of these lines along with the parental lines was genotyped by 660K SNP arrays at China Golden Marker Corporation (Beijing; http://www.cgmb.com.cn). Various SNP markers located on 2DS associated with lesion mimics were identified from the SNP typing results (Additional file 1). Whole wheat genome sequences were searched by SNP-tagged probe sequences (https://www.ncbi.nlm.nih.gov/) or according to the possible physical intervals on 2DS obtained from the SNP analysis search of the Chinese Spring genomic intervals (https://urgi.versailles.inra.fr/jbrowseiwgsc/gmod_jbrowse/?data=myData%2FIWGSC_RefSeq_v1.0&loc=chr2D%3A1..651852609&tracks=DNA&highlight=). A matched scaffold sequence was obtained, and repeated a DNA analysis was performed using the SSR Hunter 1.3 program (Li Qiang and Wan Jianmin 2005). The DNA sequences of both ends of the repeats were obtained, and primers were designed using Primer Premier 6.0 software (Canada). These primers (Table S1) were used for PCR- and electrophoresis-based analyses, and primers suitable for polymorphism were selected as molecular markers to obtain genotypes in the genetic population (Table S2).
PCR was conducted in a total volume of 20 μl comprising 200 ng of DNA template, 10 μl of 2× Es Taq MasterMix (Kangwei Century, China), 0.6 μl of 10 μM forward primer and 0.6 μl of 10 μM reverse primer. The amplification procedure was as follows: 94°C for 5 min; 35 cycles of denaturation at 94°C for 30 s, 45-60°C (adjusted according to the primers) for 30 s, and 72°C for 45 s; and then 72°C for a total extension of 10 minutes. The separation of the PCR products was carried out by 1% agarose gel electrophoresis or 8% nondenaturing polyacrylamide gel electrophoresis. The agarose gel electrophoresis was performed with ethidium bromide (EB), and the polyacrylamide gel electrophoresis was performed with silver nitrate [40-41].
Data analysis
All phenotypic data were recorded in Microsoft Office Excel 2013 for statistical analysis. One-way analysis of variance (ANOVA) was conducted to evaluate the variance and significance between groups by using SPSS 20.0 and GraphPad Prism 5.0. The genetic segregation ratio of normal (LM0-1) and lesion-mimic phenotypes (LM2-4) was tested by the chi-square test. Mean phenotypes of LMs and stripe rust scores for each RIL collected from each individual experiment were used for QTL analysis. The inclusive composite interval mapping of additive (ICIM-ADD) QTL method was used, and a walking speed of 1.0 cM with a stepwise regression probability of 0.001 was chosen for QTL detection. The threshold for declaring a significant QTL was determined by 1,000-permutation tests. The LOD score to determine significant QTLs was 3.5 in all environments, and a LOD threshold of 3.5 was the criterion selected for a significant QTL. Linkage map construction and QTL mapping were performed using QTL IciMapping V4.1 software, and the genetic distance between markers was measured using centimorgans (cM) [42]. The threshold of the logarithm of odds value was set to 3.0 to determine linkage between markers, with a maximum recombination fraction at 0.4.