The wheat-Leymus mollis introgression line M8664-3 used in this study was provided by Professor Jie Fu, College of Agronomy, Northwest A&F University, Yangling, China.

Chinese predominate Pst race CYR33 was used for seedling tests. After the identification on differential hosts of Chinese Pst, CYR33 was increased in susceptible variety Mingxian169.

Seedling tests were conducted under controlled greenhouse conditions according to Wan et al. (2007) and Bansal et al. (2017). M8664-3 and Mingxian 169 were planted in 7 7 7 cm pots with 15–20 seeds per pots. When the first leaves fully expanded, fresh CYR33 urediniospores were inoculated onto wheat leaves by the smear method, and sterile water was used as MOCK-inoculation control (Roelfs et al., 1992). Approximately 14–16 days post inoculation (dpi), when obvious uredinia were observed on the leaves of Mingxian169, the types of infection types were recorded according to a 0–9 scale. Also, leaf samples were collected at 0, 12, 24, 48, 72, and 96 h post inoculation (hpi). RNA extraction was performed using a Magen plant total RNA extraction kit (Magen Biotech, Guangzhou, China). The concentration and purity of each RNA sample were evaluated using a micro-ultraviolet spectrophotometer (NanoDrop 2000; Thermo Fisher Scientific, Wilmington, DE, United States), and the integrity was determined by 1% agarose gel electrophoresis. The first strand of DNA was synthesized with HiScript II Q-RT SuperMix for qPCR (+ gDNA wiper) (Vazyme Biotech, Nanjing, China).

The sequence of the linked markers of YrM8664-3 was blasted against the genome sequence of Chinese Spring IWGSC RefSeq v1.0 Genome (IWGSC et al., 2018) (https://wheat-urgi.versailles.inra.fr/Seq-Repository/Assemblies), and the gene was located in the range of 41.6 × 10⁷-63.9 × 10⁷ bp (base pair) on wheat chromosome 4AL (Chao et al., 2018). Referring to the IWGSC RefSeq v1.0 annotation (https://urgi.versailles.inra.fr/download/iwgsc/IWGSC_RefSeq_Annotations/v1.0/), genes near the linked markers containing conserved domains of disease resistance, such as NBS, LRR, and PK, or genes hit by the linked markers were selected.

To measure the transcriptional expression level of selected genes by qRT-PCR, specific primers (Supplementary Table 1) were designed using the Primer Premier 5.0 software (Li et al., 2011). Using TaEF-1α (GenBank accession number Q03033) as an internal reference gene, the relative transcription expression level of target genes was determined. All the qRT-PCR reactions were performed in a 20-μl reaction mixture containing 10 μl Cham Q^TM SYBR® qPCR Master Mix (Vazyme Biotech, Nanjing, China), 0.2 μl each of the forward and reverse gene-specific primers (10 μM), and 2 μl of diluted cDNA (1:10). A Bio-Rad iQ5 Real Time PCR (Bio-Rad, Hercules, CA, United States) system was used to generate cycle threshold (CT) values for the quantification of relative gene expression using the comparative 2^−ΔΔCt method (Livak and Schmittgen, 2001). All the samples were analyzed in three biological replications, and all the PCR analyses were replicated three times.

According to the primers (Supplementary Table 2) at the positive and negative ends of the open reading frame (ORF) region, the cDNA of M8664-3 was used as the template for amplification with gene-specific primers as follows: pre-denaturation at 94°C for 2 min, followed by 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 1 min, with a final incubation at 72°C for 2 min. The PCR-amplified products were examined by 1.2% agarose gel electrophoresis. The specific target bands were recovered and sequenced by Tsingke (Xi'an, China).

The online BLAST¹ program from the National Center for Biotechnology Information (NCBI) was used to analyze the cDNA sequence. The amino acid sequence was analyzed with ProtParam² and TMPred³ to detect the primary structure and position information of the polypeptide transmembrane region, respectively. TargetP-2.0⁴ and Plant-mPLoc⁵ in Cell-PLoc 2.0 were used to predict the subcellular location of the possible protein structure of the amino acid sequence. The conserved domain was identified by Pfam⁶ and simple modular architecture research tool (SMART⁷). Multiple sequence alignments were performed using DNAMAN8.0 (Lynnon BioSoft, San Ramon, CA, United States). To reveal phylogenetic relationships and potential functional characteristics of targeted proteins from different species, corresponding proteins were collected from different species according to the domain, namely Arabidopsis thaliana, Oryza sativa, Zea mays, Brachypodium virgatum, and Hordeum vulgare. The phylogenetic relationship was inferred with the neighbor-joining (NJ) method, and a midpoint rooted base tree was drawn in MEGA 7.0 with 1,000 bootstrap iterations (Kumar et al., 2016).

Virus-induced gene silencing, mediated by barley stripe mosaic virus (BSMV), was performed to reveal the function of nine candidate genes during the interaction between wheat M8664-3 and CYR33 (Fitzmaurice et al., 2002; Scofield et al., 2005). The fragment of target gene with PacI and NotI was derived from its coding sequence and amplified by RT-PCR (Supplementary Table 1) to construct the BSMV:γ plasmid for gene silencing. After amplification of the target gene fragment and vector linearization, the target gene and linearized vector were recombined, connected, linearized with a restriction enzyme, and transcribed in vitro (RiboMAX TM Large-Scale RNA Production System-T7 and Ribo m7G Cap Analog; Promega, Madison, WI, United States) to form a recombinant virus containing the target gene (Petty and Jackson, 1990). BSMV:γ was used as a blank control and γ-PDS as a positive control by friction inoculation, and the second leaf of wheat seedlings was infected with BSMV, with three biological replicates for each gene.

After 24-h incubation in the dark in an artificial climate incubator, the seedlings were placed in a growth chamber at 25°C with 60–80% humidity. When the photo-bleaching phenotype was observed on the BSMV: γ-PDS plants at ~10 dpi, the fourth leaves of γ-gene silencing plants were inoculated with CYR33. Leaf samples were collected at 0, 24, 48, 96, and 120 hpi for qRT-PCR analysis and histological observation (Li et al., 2011). The infection phenotype of Pst was observed at ~14 dpi.

Inoculated and control leaves of both Pst seedling test and VIGS-induced gene silencing were sampled at different time points for histological observation. For each time point, ~15 blades were sampled (Ayliffe et al., 2011). The H₂O₂ burst for ROS was observed by DAB (MP Biomedicals, Solon, OH, United States) staining (Xiao et al., 2003; Zou et al., 2018). WGA-Alexa 488 (Invitrogen, Carlsbad, CA, United States) was used to fluorescently stain the Pst infestation structure in wheat leaf tissue. The infestation site was determined by the production of germ tubes by Pst and the formation of substomatal vesicle in the stomata. For each treatment, 50 infection sites were randomly examined using an Olympus BX-53 microscope (Olympus Corporation, Tokyo, Japan), and the area of ROS, hypha branches, hypha length, and necrotic areas were observed and measured.

The standard error of deviation was calculated using Microsoft Excel. The statistical significance was evaluated by Student's t test (P < 0.05) using the SPSS software (SPSS, Inc., Chicago, IL, United States).

Agrobacterium tumefaciens strain EHA105 harboring binary vector, pCAMBIA3301, was used to optimize the transformation system, with ubiquitin promoter and targeted gene replacing cauliflower mosaic virus 35S promoter and GUS gene encoding β-glucuronidase. The ORF of TaFBN and Ta_Pes_BRCT was amplified separately and inserted into the frame of an expression cassette within the T-DNA region of the pCAMBIA3301 vector digested with BamHI and SpeI. The construct was verified by DNA sequencing and introduced into Agrobacterium EHA105. Then, the Agrobacterium-mediated transformation method was applied to genetically transform the target gene (Li et al., 2019). The seeds of 14-day-pollinated immature wheat variety Fielder (scutellum size 1 mm) were treated with 75% alcohol for 30 s and 0.1% HgCl for 10 min, and the immature embryos were then removed with a dissecting needle on an aseptic work table. The immature embryos were infected with the obtained Agrobacterium, and then placed on the screening medium and co-cultured in darkness at 23°C for 3 days. Under a microscope, hypocotyls were excised from the contact between the hypocotyl and scutellum of the seeds, and the obtained scutellums were cultured on the screening medium for 14 days. After cutting the callus, a second selection was performed, and the tissue was cultured for 14 days. The healthy callus was transferred to a regeneration medium, and regenerated plantlets were obtained after 7 days. The 786-bp genomic fragment of TaFBN and the 1,791-bp genomic fragment of Ta_Pes_BRCT were introduced into wheat cultivar Fielder, and the T₀ generation plants were screened with gene-specific primers by PCR amplification under the follow conditions: 1 min pre-denaturation at 94°C, followed by 30 cycles of denaturation at 94°C for 1 min, annealing at 60°C for 1 min, elongation at 72°C for 45 s, and a final extension step of 10 min at 72°C. Then, the T₂ generation plants derived from the positive T₀ plant progeny were inoculated with CYR33 for functional analysis of the target genes.

In order to evaluate the expression level of TaFBN in wheat M8664-3 under hormone perception and environmental stress conditions, seedlings of M8664-3 were divided into eight groups. The first four groups were sprayed with methyl jasmonate (MeJA), ethylene (ET), salicylic acid (SA), or abscisic acid (ABA), at 100 mM each as a hormone treatment and cultivated at 16°C. The last four groups were treated with low temperature (4°C), high temperature (37°C), salt (200 mM NaCl), and drought (15% PEG6000) to experience environmental stress, and sterile water was sprayed at 16°C as a blank control. Leaf samples were collected for the expression level analysis of TaFBN 0, 1, 3, 6, 12, and 24 h after the different stimulus treatments.

The infection type (IT) of wheat M8664-3 to CYR33 was 2 with large areas of necrosis on the leaves, whereas the IT of susceptible wheat Mingxian169 was 9 with fully expanded uredinium covered on the leaves at 15 dpi (Figure 1A).

Histological observation showed that significant DAB staining appeared at the leaf infection sites at 12 and 96 hpi (Figure 1B). At 12 hpi, strong DAB staining appeared in the guard cells directly contacted by the substomatal vesicle, and there was a significant difference in the H₂O₂ staining area of the infestation sites between M8664-3 and Mingxian 169. Between 24 and 48 hpi, the H₂O₂ staining area of the M8664-3 and Mingxian169 infection sites continuously increased. Between 72 and 96 hpi, the staining area of the M8664-3 infection site expanded to include guard cells and surrounding areas. In the mesophyll cells of Mingxian169, the DAB staining expansion area was far smaller than that of M8664-3 (Figure 1).

Through annotation analysis, 19 genes near the linked markers and containing disease resistance-related domains, or hit by the linked markers of YrM8664-3 were selected (Supplementary Figure 1; Supplementary Table 2). Among these, nine genes were significantly upregulated after inoculation with CYR33, as determined by qRT-PCR (Figure 2), namely, TraesCS4A01G305600 (Leucine-rich repeat receptor-like protein kinase family protein), TraesCS4A01G319000 (PGR5-like protein 1A, chloroplastic), TraesCS4A01G319300 (disease-resistance protein (NBS-LRR class) family), TraesCS4A01G365200 (Sn1-specific diacylglycerol lipase alpha), TraesCS4A01G272000 (plastid-lipid associated protein PAP/fibrillin family protein, which encodes TaFBN), TraesCS4A01G272900 (beta-glucosidase, which encodes Ta_Pes_BRCT), TraesCS4A01G276400 (Pescadillo homolog), TraesCS4A01G298300 (nonspecific phospholipase C), and TraesCS4A01G181300 (AP2-like ethylene-responsive transcription factor) (Table 1).

Unique fragments were designed to knock down these nine candidate genes using primers specified in Supplementary Table 1. All of the BSMV-inoculated plants displayed chlorotic mosaic symptoms at 10 dpi, but there were no obvious defects in further leaf growth, while the leaves inoculated with BSMV: TaPDS exhibited photobleaching (Figure 3A), indicating that the BSMV induced gene silencing system functions well. After knocking down these genes with the VIGS method, M8664-3 became susceptible in the TaFBN- and Ta_Pes_BRCT-silenced system (Figure 3B), which indicated that these two genes may be involved in the resistance of M8664-3 to Pst. To determine the efficiency of VIGS, qRT-PCR was performed to examine the relative transcript levels of TaFBN and Ta_Pes_BRCT in the fourth leaves of infected plants. Compared with control inoculations, transcript levels of TaFBN-knockdown plants were reduced by 49, 39, 42, 43, and 42% at 0, 24, 48, 96, and 120 hpi, and Ta_Pes_BRCT knockdown plants also showed a stable efficiency by reducing to 49, 44, 36, 44, and 44% at 0, 24, 48, 96, and 120 hpi with CYR33, respectively (Figure 3C).

Histological observations showed that the number of haustorial mother cells, hypha length, and Pst growth area in TaFBN- and Ta_Pes_BRCT-silenced leaves all slightly increased as compared with that of the unsilenced treatment 48 hpi, but the DAB-reactive oxygen staining area did not significantly change, and no necrotic cells were found. At 120 hpi, the area of Pst growth, length of hyphae, and number of hyphae branches, haustorial mother cells, and haustorium in TaFBN- and Ta_Pes_BRCT-silenced leaves were all increased, and the difference was extremely significant. DAB-reactive oxygen staining decreased to 0, and the area of necrotic cells also significantly decreased compared with the unsilenced treatment at 120 hpi (Figure 4). In summary, the silencing of TaFBN and Ta_Pes_BRCT reduced disease resistance and made plants more susceptible to Pst.

TaFBN encodes a protein composed of 261 amino acids, with a molecular weight of 28.59 kDa, an isoelectric point (PI) of 9.34, and an average hydrophobicity of −0.330, which suggested that it may be a hydrophilic protein. There was a PAP_fibrillin domain at positions 90–251 of the amino acid; therefore, it was temporarily named TaFBN. The phylogenetic analysis of TaFBN with H. vulgare (HvFBN4, KAE8820120), Brachypodium distachyon (BdFBN4, XP_003560711), O. sativa (OsFBN4, XP_015632312), A. thaliana (AtFBN3a, NM_113511; AtFBN3b, BT020596), and Z. mays (ZmFBN4, ACG27798) resulted in the clustering of TaFBN with HvFBN4, BdFBN4, OsFBN4, and ZmFBN4, all of which are members of FBN4 proteins in monocotyledons (Figure 5A). A nucleic acid sequence analysis revealed that TaFBN shared 96.55% identity with HvFBN4 from H. vulgare. Multiple amino acid sequence alignments of TaFBN with HvFBN4, BdFBN4, OsFBN4, and ZmFBN4 showed that TaFBN4 is predicted to encode proteins with the unique conserved domains of PAP complex FBN4 (Figure 5B). Therefore, it was determined that TaFBN and FBN4 were clustered together, and that the functional annotations of the latter on the UniProt website were related to resistance to bacterial diseases and ozone.

Ta_Pes_BRCT encodes a protein composed of 596 amino acids, with a molecular weight of 68.32 kDa, PI of 7.68, and an average hydrophobicity of −0.615, which indicated that it may be a hydrophilic protein. The amino acid coded by Ta_Pes_BRCT has a Pescadillo (PES) domain at positions 9–277, and the 339–417 amino acids contain a BRCT domain. The predicted function may be related to ribosomes; therefore, it was temporarily named Ta_Pes_BRCT. The phylogenetic analysis and multiple amino acid sequence alignment of Pes_BRCT proteins indicated that Ta_Pes_BRCT is predicted to encode proteins with conserved domains of the Pes_BRCT complex (Supplementary Figure 2). The functional annotations of BRCT on the UniProt website were related to DNA repair under stress. As predicted by the TMpred program, TaFBN and Ta_Pes_BRCT have no transmembrane domain. As predicted by TargetP-2.0 and Plant-mPLoc, the location of the protein encoded by Ta_FBN was predicted in Chloroplast; the location of the protein encoded by Ta_Pes_BRCT was predicted in Nucleus.

In order to further analyze the function of TaFBN and Ta_Pes_BRCT in stripe rust resistance, the two genes were introduced into the susceptible bread wheat variety Fielder by an Agrobacterium-mediated transformation method. The individuals of T₀ generation were identified by PCR analysis (Figure 6A), and seven independent T₂ lines obtained from the positive T₀ progeny were further used to conduct disease resistance tests and PCR detection procedures. Three replicates were tested for each line. The T₂ generation of TaFBN transgenic plants was resistant to CYR33, while all of the Ta_Pes_BRCT transgenic plants were susceptible to CYR33 (Figure 6B), which indicated that TaFBN confers more important resistance to CYR33.

Under different hormone treatments (MeJA, ET, SA, and ABA), the transcription level of TaFBN in the leaves of M8664-3 wheat seedlings was determined (Figure 7). The expression level of TaFBN significantly increased under SA and ABA treatments, and peaked (more than twice) at 6 and 12 hpi, respectively. The expression of TaFBN also slightly increased after MeJA and ET treatment. Under abiotic stress, the expression of TaFBN significantly increased after NaCl, PEG6000, and 4°C treatment, but its expression in 37°C-high temperature treatment did not significantly increase. In summary, TaFBN might be induced by SA, ABA, high salt, drought, and low temperature to increase its expression.