TaTLP1 interacts with TaPR1 to contribute to wheat defense responses to leaf rust fungus

Thaumatin-like proteins (TLPs), which are defined as pathogenesis-related protein family 5 (PR5) members, are common plant proteins involved in defense responses and confer antifungal activity against many plant pathogens. Our earlier studies have reported that the TaTLP1 gene was isolated from wheat and proved to be involved in wheat defense in response to leaf rust attack. The present study aims to identify the interacting proteins of TaTLP1 and characterize the role of the interaction between wheat and Puccinia triticina (Pt). Pull-down experiments designed to isolate the molecular target of TaTLP1 in tobacco resulted in the identification of TaPR1, a pathogenesis-related protein of family 1, and the interaction between TaTLP1 and TaPR1 was confirmed by yeast two-hybrid experiments (Y2H), bimolecular fluorescence complementation (BiFC), and co-immunoprecipitation (Co-IP). In vitro, TaTLP1 and TaPR1 together increased antifungal activity against Pt. In vivo, the disease resistance phenotype, histological observations of fungal growth and host responses, and accumulation of H2O2 in TaTLP1-TaPR1 in co-silenced plants indicated that co-silencing significantly enhanced wheat susceptibility compared to single knockdown TaTLP1 or TaPR1 plants. The accumulation of reactive oxygen species (ROS) was significantly reduced in co-silenced plants compared to controls during Pt infection, which suggested that the TaTLP1-TaPR1 interaction positively modulates wheat resistance to Pt in an ROS-dependent manner. Our findings provide new insights for understanding the roles of two different PRs, TaTLP1 and TaPR1, in wheat resistance to leaf rust.


Introduction
Plenty of studies have demonstrated that pathogenesis-related (PR) proteins play key roles in plant disease-resistance responses and specifically systemic-acquired resistance (SAR) [1]. Half a century ago, PRs were found to play a pioneering role in the discovery of plant innate immunity as different gene families were discovered to be responsible for the accumulation of PRs upon pathogen challenge. PRs and similar proteins have been reported to be induced by various types of pathogens in many plants. So far, a total of 17 gene families designated as PR genes were identified from plant species in response to various pathogens [2][3][4]. Among different groups of PRs, PR1, PR2, and PR5 are the most common and accumulate locally and systemically suggesting an involvement in SAR that represents a form of plant immunization [2]. PRs generally encode small antimicrobial proteins and their expression levels increase quickly when stimulated by biotic or abiotic stress, and therefore serve as a marker of plant immune signalling. However, the biochemical activity and mode of action of PR proteins has remained elusive.
PR5 family proteins, which are also referred to as thaumatin-like proteins (TLPs), are able to rapidly accumulate to high levels in response to biotic or abiotic stress and confer antifungal activity in various plant species [5][6]. Most TLPs respond to some biotic or abiotic stresses and some resistance genes function together with TLPs based on genome-wide and functional network analyses in Arabidopsis, Oryza, Populus, Zea, Physcomitrella and Chlamydomonas [7][8]. The PR5 family protein osmotin has been identified as an interacting partner for BcIEB1, a glycoprotein abundantly secreted by Botrytis cinerea, and this interaction protects B.cinerea against the antifungal action of osmotin and also modulates the elicitor activity of BcIEB1 in plants [9]. GhPR5, a partner protein of PevD1 which is an Alt a 1-like protein, was isolated from cotton (Gossypium hirsutum) plants by co-purification assays, and the PevD1-GhPR5 interaction was determined to occur at the C-terminus of PevD1 by pull-down and yeast twohybrid techniques [10]. A new osmotin-like protein (OLP) gene of the PR5 family from Solanum nigrum, SindOLP, was cloned and exhibited in vitro antifungal activity against Macrophomina phaseolina [11]. Transgenic sesame lines overexpressing SindOLP showed enhanced resistance against drought and salinity stresses and biotic stress caused by a fungal pathogen [12].
PR1 proteins represent a large protein family (23 in wheat) that are upregulated early in the defense response. Findings from a recent study provided genetic and biochemical evidence that indicates that PR1 binds to sterols and inhibits pathogen growth [13]. A dimeric PR1 type protein (PR-1-5) was found to interact with Stagonospora nodorum ToxA and potentially mediate ToxA-induced necrosis in wheat. Subsequent mutational analysis identified several residues on both ToxA and TaPR-1-5 that are required for their interaction as well as the induction of necrosis, which showed that SnToxA interacts specifically with TaPR-1-5 [14]. Interestingly, analysis of the SnTox3-TaPR-1 interaction demonstrated that SnTox3 can interact with a broad range of TaPR-1 proteins [15]. Recently, PR1 was found to interact with SsCP1, a Cerato-platanin protein (CP), in the apoplast to facilitate infection by the necrotrophic phytopathogen Sclerotinia sclerotiorum using various methods including yeast twohybrid, Glutathione S-transferase (GST) pull-down, co-immunoprecipitation (CoIP), and bimolecular florescence complementation (BiFC). Overexpressing PR1 enhanced resistance to the wild-type strain, but not to the Sscp1 knockout strain of S. sclerotiorum [16]. Collectively, the finding that both ToxA and SsCP1 interact with PR1 proteins is suggestive of the critical nature of these interactions and implies that PR1 proteins play a key role in mediating disease outcomes in the plant-fungus interaction.
Wheat leaf rust, caused by the obligate biotrophic plant pathogen Puccinia triticina (Pt), occurs world-wide and is one of the most destructive diseases of wheat in major wheat-growing regions [17]. Wheat leaf rust can reduce the quality of harvested grain and cause significant yield losses. Thus, it is crucial to understand the defense mechanisms of wheat against Pt and to search for new ways to control wheat leaf rust disease. Resistance of wheat cultivars against Pt is usually divided into two genetic categories, seedling resistance genes and adult plant resistance (APR) genes also referred to as minor genes. Seedling resistance genes are referred to as major resistance genes and they confer resistance at both the seedling and adult plant stages. Major resistance genes are critical for expression of a hypersensitive reaction (HR), and are often race-specific and mediated by gene-for-gene resistance [18][19]. APR genes are commonly detected at post-seedling stages, are quantitatively inherited with partial effects on disease severity (DS), and are not involved in express of a HR [20][21]. As we know, regardless of seedling resistance genes or APR genes, our findings strongly suggest that PRs are involved in wheat rust defense responses [22][23][24].
However, the mechanistic understanding of how rust can interact directly with host proteins remains limited because of the large genome size, polyploid nature, and high content of repetitive DNA of hexaploid wheat (T. aestivum L.) obfuscating functional gene characterization efforts. Recently Yang et al. reported that an effector, Pst18363, was identified from Puccinia striiformis f. sp. tritici (Pst) and shown to interact with and stabilize wheat Nudix hydrolase 23, which suppresses reactive oxygen species (ROS) accumulation to facilitate Pst infection [25]. Huai et al. found that Pst stimulates ABA biosynthesis in host cells and thereby upregulates TaSTP6 (a sugar transporter in wheat) expression, which increases sugar supply and promotes fungal infection [26]. In our previous study, a thaumatin-like protein gene, TaTLP1, was cloned from wheat in response to the infection of Pt race PHNT (GenBank accession number KJ764822). TaTLP1 contained a functional secretion peptide and was secreted into the apoplast. BSMV-induced gene silencing and pathology tests on the silenced plants revealed that TaTLP1 is involved in wheat resistance to the wheat leaf rust pathogen [24]. In this study, we aim to screen for interacting proteins of TaTLP1 using pull-down experiments and to characterize the role of the protein interactions between wheat and Pt. Our findings provide new insights for understanding the roles of different PRs, especially TaTLP1 and TaPR1, in wheat defense to Pt infection.

Identifying interacting proteins of TaTLP1 by GFP-Trap and mass spectrometry (MS)
The ChromoTek GFP-Trap tool is based on a GFP-binding protein derived from an alpaca single variable domain antibody. It has been demonstrated to be an excellent tool to for purification of GFP-fused proteins and their interacting partners [27][28]. To identify interaction protein(s) targeted by TaTLP1, pull-down assays using the GFP-Trap kit were carried out by mixing Agrobacterium tumefaciens GV3101 containing a construct encoding recombinant TaTLP1-pEarlyGate103 (c-GFP tagged) with Nicotiana benthamiana leaf extracts, followed by purification of proteins with anti-GFP gel beads. After elution, purified proteins were separated by SDS-PAGE. Several different bands were observed between samples from TaTLP1-pEarlyGate103 and controls (Fig 1). A total of 165 proteins were identified by mass spectrometry in differential bands from plant extraction with or without TaTLP1. S1 Table summarizes all 7 candidates uniquely found in the TaTLP1-pEarlyGate103 expressing samples according to their biological function which have at least 15 different peptides detected in MS (S1 Table). This result provided us great confidence in the reliability of identifying the interacting proteins of TaTLP1 via GFP-Trap and MS. Among the identified proteins, a PR1-like protein was co-purified with TaTLP1-pEarlyGate103 as an 18-kD protein band was co-purified with TaTLP1, whereas no additional band appeared in the control assays with extracts alone (Fig 1). Since the expression profiles and functional analysis of TaPR1 were performed in our previous study [29], these results inspired our hypothesis that TaTLP1 and TaPR1 are likely to interact with each other and are associated with resistance of wheat against Pt infection.

PLOS GENETICS
TaTLP1-TaPR1 is required for wheat resistance to Pt

TaTLP1 and TaPR1 are induced by leaf rust and share the same expression profiles
To further examine whether the identified proteins functionally interact with and regulate TaTLP1, TaPR1 was selected for further biochemical experiments on the basis of it being identified in all three independent mass spectrometry results. First, we quantified the relative expression levels of both defense-related genes (TaTLP1 and TaPR1) in response to leaf rust infection using Thatcher which is a susceptible material to Pt as control. In the compatible interaction (Thatcher response to Pt), TaTLP1 and TaPR1 did not show significant changes at either time point after inoculation with Pt race 07-10-426-1, named as PHNT according to the international code [30]. In contrast, in the incompatible interaction (TcLr19 response to Pt), the transcript levels both of TaTLP1 and TaPR1 increased to their highest levels from 72 hours post inoculation (hpi) to 168 hpi. The transcript levels of TaTLP1 and TaPR1 were much higher during the incompatible interaction compared to the compatible interaction at 0-168 hpi. Based on these observations, the expression levels of TaTLP1 and TaPR1 in TcLr19 plants after inoculation by leaf rust were similar throughout the time course (Fig 2), suggesting that the expression of TaTLP1 and TaPR1 are similarly induced by leaf rust.

TaTLP1 physically interacts with TaPR1 in vitro
To further assess whether TaPR1 could directly interact with TaTLP1, we conducted in vitro yeast-two hybrid (Y2H) assays using TaPR1 without a signal peptide as bait fused to the GAL4 DNA-binding domain (BD) and TaTLP1 without a signal peptide as prey fused to the GAL4-activation domain (AD). AvrLm1 and BnMPK9 served as a positive control with serially diluted cell suspensions that were used to determine the strength of the interaction by blue coloration derived from α-galactosidase activity on selective agar plates. Similar to positive controls (Fig 3, rows 1), the yeast transformant co-expressing the full-length mature TaTLP1 and TaPR1 proteins produced blue coloration (Fig 3, rows 5). In contrast, replacement of either the TaTLP1 prey or the TaPR1 bait with the yeast proteins AD or BD from empty vectors caused a loss of blue coloration (rows 2-4). These results support the hypothesis that TaTLP1 and TaPR1 are likely to interact with each other.
Then this interaction was further studied in vivo using co-localization techniques. Overlapping fluorescence was detected in the apoplast of agroinfiltrated cells when TaTLP1-pGWB454 and TaPR1-pEarlyGate103 were co-expressed ( Fig 4B). These results indicate that TaTLP1 interacts with TaPR1 in the plant cell apoplast. Co-immunoprecipitation assays were performed upon transient expression in N. benthaminana to further validate the interaction between TaTLP1 and TaPR1 in vivo. ΔspTaTLP1-pGWB414 co-immunoprecipitated with ΔspTaPR1-pEarlyGate103, and vice versa ( Fig 5A). Taken together, these observations support that TaPR1 and TaTLP1 interact in vivo. Yeast two-hybrid analysis of the interaction between TaPR1 and TaTLP1. Yeasts expressing the indicated combinations of bait and prey were spotted on the synthetic dropout medium without leucine and tryptophan (SD-WL) and SD medium without leucine, tryptophan, histidine, and adenine supplemented with X-α-Gal and Aureobasidin A (SD-HAWL + X-α-Gal). Only those co-expressing TaPR1 and TaTLP1 grew on SD-HAWL plates and developed blue coloration, indicative of an interaction between both proteins. AvrLm1 (bait) and BnMPK9 (prey) are the positive controls; TaPR1 and pGADT7, TaTLP1 and pGBKT7, pGADT7 and pGBKT7 are used as the negative controls. https://doi.org/10.1371/journal.pgen.1008713.g003

TaPR1 stabilizes the accumulation of TaTLP1
ΔspTaTLP1-pGWB414 (HA-tagged TaTLP1), ΔspTaPR1-pEarlyGate103 (GFP-tagged TaPR1), and a Myc-tag negative control were co-expressed in N. benthamiana using agroinfiltration. Most protein degradation events are accomplished via the ubiquitin/26S proteasome pathway [32]. MG132, a proteasome inhibitor, was used to examine and compare the stability and accumulation of TaTLP1 to test whether it was regulated by this pathway. In the absence of MG132, ΔspTaTLP1-pGWB414 accumulated to a higher level in the presence of  ΔspTaPR1-pEarlyGate103 compared to the Myc control. However, N. benthamiana leaves coexpressing ΔspTaTLP1-pGWB414 andΔspTaPR1-pEarlyGate103 or Myc that were pretreated with MG132 resulted in increased accumulation of TaTLP1 in the leaves of control treatments (co-expressing ΔspTaTLP1-pGWB414 and Myc) to a similar level of accumulation of TaTLP1 that was observed in the leaves co-expressing ΔspTaTLP1-pGWB414 and ΔspTaPR1-pEarly-Gate103 ( Fig 5B). Taken together, these results suggest that TaPR1 stabilizes TaTLP1 by preventing its degradation mediated by the 26S proteasome.
It has been reported that cysteines are important for TaTLP1 structure [33]. To further define the roles of disulfide bonds in the stabilization of the TaPR1-TaTLP1 interaction or in determining individual protein structures that confer interactions of TaTLP1, we generated two mutated TaTLP1 proteins with five cysteine to alanine mutations at the N-terminal region and another five cysteine to alanine mutations at the C-terminal region, and used these variants to perform Y2H experiments with TaPR1, respectively. Interestingly, one of the C-terminal cysteine-mutated regions weakened the growth of yeast on selective plates, indicating that mutation of C-terminal cysteines from TaTLP1 affects its interaction with TaPR1. Furthermore, we mutated the individual cysteines at the C-terminal, and we found that the effects are similar to the mutation of all cysteines in combination (S3 Fig), indicating that cysteines affect the interaction between TaTLP1 and TaPR1, but are not indispensable for the interaction of TaPR1 and TaTLP1.

Antifungal activity of TaTLP1-TaPR1 is higher compared to either TaTLP1 or TaPR1 alone
A PR5-like protein previously has been reported to inhibit the growth of yeast and filamentous fungi, such as Botrytis cinerea [34]. In our previous study, we found that TaTLP1 was required for Lr35-mediated resistance against leaf rust fungus via BSMV-induced gene silencing analysis [24]. However, it remains challenging to examine antifungal activity in vitro as leaf rust is an obligate biotroph, growing and reproducing exclusively on living host tissue. As an initial test to establish a possible role for TaTLP1-TaPR1 in protecting wheat against Pt, growth-inhibition assays were conducted. The pGEX-6P-3 vector system was used to express TaTLP1 in Escherichia coli with a protein tag (GST) for purification and subsequent antibody recognition (S4 Fig). We analyzed the germination of urediniospores, the length of hyphae, and the number of hyphal branches in the presence of pure TaTLP1-TaPR1, TaTLP1 alone, and TaPR1 alone, respectively, using sterile water, elution buffer and pGEX-6P-3 vector as a control. Microscopic observation showed that the germination of urediniospores and the growth of germtubes were significantly restricted with pure proteins (Fig 7, S2 Table). The antifungal activity of TaTLP1 or TaPR1 alone is significantly higher compared to controls, and the antifungal activity of TaTLP1-TaPR1 is higher compared to either TaTLP1 or TaPR1 alone, indicating that the proteins alone and especially in combination protect wheat by inhibiting the growth of leaf rust fungus and contribute to resistance against Pt.

Co-silencing of TaTLP1 and TaPR1 significantly enhances wheat susceptibility
To further confirm the function of the interaction between TaTLP1 and TaPR1 in wheat resistance to Pt, the BSMV-virus-induced gene silencing (VIGS) system was employed to knock down the transcription of TaTLP1 and TaPR1, respectively, and we analyzed the impact on wheat response to pathogen infection. Then, BSMV: TaTLP1-V2 and BSMV: TaPR1-V2 recombinant vectors were mixed together and rub-inoculated onto TcLr19 seedlings to detect the effect of TaTLP1-TaPR1 co-silencing on wheat response to Pt infection. All BSMV-inoculated plants displayed mild chlorotic mosaic symptoms depicting infection by the virus ( Fig  8A). Photo-bleached symptoms were observed on wheat leaves after inoculation with BSMV: TaPDS at 10 dpi when TaPDS was silenced (Fig 8A). After inoculation with fresh urediospores of the Pt race PHNT, a conspicuous hypersensitive response (HR) was elicited on leaves inoculated with BSMV: γ. Various numbers of Pt urediniospores were produced around the necrotic spots on the fourth and fifth leaves infected with BSMV: TaTLP1 and BSMV: TaPR1, as compared with BSMV: γ plants (Fig 8A). Silencing of both TaTLP1 and TaPR1 produced slightly larger pustules compared to the control, and more urediniospores were observed on the ). Each of the constructs was co-transformed with TaTLP1 into yeast. All transformants are able to grow on synthetic dropout medium without leucine and tryptophan (SD-WL) medium. Yeast colonies that were able to grow on selective medium (SD medium without leucine, tryptophan, histidine, and adenine supplemented with X-α-Gal and AureobasidinA [SD-HAWL]) and displayed blue coloration confirmed the protein-protein interaction. Positive control is MAPK9-AvrLm1. TaPR1 and pGADT7, TaTLP1 and pGBKT7, pGADT7 and pGBKT7 are used as the negative controls. BSMV: TaTLP1/TaPR1 plant leaves than those with single gene silencing at 14 dpi with Pt race (Fig 8A).
To determine the silencing efficiency of TaTLP1 and TaPR1 in plants that had been infected with recombinant vectors, we detected the relative expression levels of TaTLP1 and TaPR1 in infected leaves using qPCR. Our results showed that the relative expression of TaTLP1 was significantly reduced by 75%, 60.5%, and 69.3% after infection with avirulent Pt at 24, 48, and 120 hpi, respectively, whereas the corresponding TaPR1 transcript expression in infected BSMV: TaPR1 leaves was reduced by 62.4%, 61.7%, and 57.58%, respectively (Fig 8B). Approximately 77.4%, 63.96%, and 72.8% of the TaTLP1 transcripts and 72.5%, 67.6%, and 61.7% of the TaPR1 transcripts were suppressed for each respective time point in the TaTLP1/TaPR1 cosilenced plants compared with that in BSMV: γ inoculated plants, suggesting that both of TaTLP1 and TaPR1 were silenced successfully (Fig 8B). The silencing efficiency further confirmed that the enhanced susceptibility phenotypes observed on the leaves inoculated with the avirulent Pt race were due to the silencing of TaTLP1 and TaPR1.
The histological changes in both the susceptible control (Thatcher) and silenced plants infected with Pt were observed. The number of hyphal branches in Thatcher was higher than in all BSMV silenced plants at 24 hpi (Fig 9A and 9B), and the hyphal length was longer in Thatcher than all knock-down plants (Fig 9C). At 120 hpi, the hyphal branches, the hyphal length, and the infection areas increased in different plants, and TaTLP1/TaPR1-knockdown plants displayed more hyphal branching, larger infection areas, and longer hyphal length compared to the control group (BSMV: γ) at 120 hpi (Fig 9A-9D). These results indicate that cosilencing of TaTLP1/TaPR1 compromises the resistance response to Pt infection.
High concentration of ROS induces cell death in plants [35]. H 2 O 2 is accumulated in different defense responses during wounding, insect attacks, and pathogen infections [36]. H 2 O 2 accumulation was evaluated to further understand how TaTLP1/TaPR1 participates in wheat resistance. The areas of cell death and H 2 O 2 accumulation were significantly lower at 120 hpi in TaTLP1/TaPR1-knockdown plants than that observed in BSMV: γ, TaTLP1 alone, or TaPR1 alone inoculated leaves (Fig 10). To further confirm the decrease in ROS accumulation in silenced plants and to understand the mechanism behind the decreased oxidative stress tolerance, the expression of genes involved in ROS signaling, catalase (TaCAT) and superoxide dismutase (TaSOD), were assayed. As shown in S5

Discussion
PR proteins are encoded pathogenesis-induced genes and play a supporting role in plant defense and the accumulation of hormones associated with plant defense. However, the function of the PR family members and the coordination between them remains unclear. TLP is involved in plant defense against various biotic and abiotic stresses [5]. The evolution of the TLP gene family was investigated by performing a genome-wide identification and comparison in Arabidopsis, Oryza, Populus, Zea, Physcomitrella and Chlamydomonas, and functional network analysis demonstrated that some resistance genes work together with TLPs [7]. As we all know, TLPs and PR1 are often considered to be common marker genes for antipathogen responses and were also found to interact with pathogen receptors in recent years [8][9][10]. In our previous study, we reported that TaPR1 and TaTLP1 play important roles in wheat development and resistance to leaf rust pathogen attack [24,29]. Interestingly, now we identified TaPR1 as an interacting partner of TaTLP1. Although there are many reports of PR genes

PLOS GENETICS
TaTLP1-TaPR1 is required for wheat resistance to Pt involved in plant pathogen defense, as far as we know, this is the first report of the interaction of two PR molecules in wheat, which sheds some light on the mechanism of PR genes in plants and might provide a basis for further functional investigation of these gene families.
A PR1-derived defence signalling peptide from the C-terminus of TaPR1-1, known as CAPE1, enhanced the infection of wheat by Parastagonospora nodorum in an SnTox3-dependent manner, but played no role in ToxA-mediated disease [15]. The CAPE1 peptide with a PxGNxxxxxPY-motif was derived from the C-terminal end of tomato PR-1b by a novel peptidomics approach and shown to increase immunity against pathogen infection in tomato and Arabidopsis. It was speculated that the CNYx motif immediately upstream of CAPE1 is required for cleavage from PR-1b [37]. In this study, we tested to see if the interaction of TaP-R1-TaTLP1 required CAPE1. The CNYx motif and CAPE1 peptide consists of 11 amino acids (PxGNxxxxxPY-) encoded within the CTR (S2 Fig) of TaPR1 (amino acids 143-164) were mutated and showed that CAPE1 appears to not be required for interaction of TaPR1-TaTLP1 (Fig 6). These results are different in that a mutation of CNY sequence can abolish the cleavage of the AtCAPE1 peptide [38], and the peptide sequence shows that it appears conserved in most of the wheat PR1 proteins (S2 Fig). However, 52 amino acids located at C-terminal region of TaPR1 are required for its interaction with TaTLP1 (Fig 6). Precipitating from this work, we identified that a total of 15 amino acids from 113 to 127 located at C-terminal region of TaPR1 is required for the interaction. Most TLPs have 10 or 16 conserved cysteine residues, which might form eight disulfide linkages to stabilize proteins and resist unfavorable pH, proteases, and heat-induced denaturation [33,39]. Thus, we investigated the function of cysteine residues in the interaction between TaTLP1 and TaPR1. We found that cysteine is not completely indispensable for the interaction of TaPR1 and TaTLP1, though it can affect the interaction between TaTLP1 and TaPR1. We will find the particular region of TaTLP1 which might be required for the interaction with TaPR1 in the future.
Leaf rust forms an intimate connection with the host through the formation of haustoria, which emerge from extracellular pathogen infection structures and enter plant cells through the infection peg by invaginating the host plasma membrane [40], so the interaction between host and fungus happens in the space between the cell membrane and cell wall. Consistent with this scheme for pathogenesis, our co-localization results showed that TaTLP1 interacts with TaPR1 in the plant cell apoplast (Fig 4B). Based on the presented data, we investigated the stability and level of accumulation of TaTLP1 in the presence of TaPR1. The results suggest that TaPR1 potentially increases the stability of TaTLP1. Moreover, to our surprise, it showed that pure TaTLP1-TaPR1 complexes play a crucial role in suppressing the germination of urediniospores and hyphal growth which indicated that the interaction of TaTLP1 and TaPR1 contribute to wheat defense responses to Pt.
In vivo, BSMV-VIGS was conducted to co-silence TaTLP1 and TaPR1 individually or simultaneously. We have reported that BSMV-induced TaTLP1 gene silenced wheat plants exhibited obviously compromised resistance, which suggests that TaTLP1 is involved in Lr35mediated wheat defense in response to leaf rust attack [24]. However, characterization of PR-1 function is a challenging task because of the genomic redundancy of the PR-1 family. Lu reported that attempts to specifically knock down the PR-1-5 gene by VIGS were unsuccessful [14]. In this study, based on the identification of the functional regions and off-target prediction of TaPR1, two fragments of TaPR1 were knocked down successfully and showed be involved in wheat defense in response. In addition, we found that the silencing efficiency of TaPR1-V2 is higher than TaPR1-V1 based on phenotype, qPCR analysis, histological observations of fungal growth and H 2 O 2 accumulation (S6 Fig, S7 Fig). Thus, TaPR1-V2 was selected to detect the co-silencing of TaTLP1 and TaPR1 genes by the VIGS system and verified the function of their interaction. TaPR1 belongs to PR1-4 and shares 94.51% homology with PR1-5 (S2 Fig), we detected the copy number of TaPR1 by Southern blotting and found that there are 4 copies in the wheat genome which is consistent with BLAST analyses (S8 Fig, S9 Fig). Meanwhile, TaPR1 was located on 7D by chromosome localization using Chinese Spring and its null-tetrasomic lines (S10 Fig). A VIGS fragment was designed according to the specific region of TaPR1 so that it can represent the function of the gene (S6 Fig). Similar to Lr21 [41], in the case of TaTLP1, we have reason to believe that the level of suppression may actually be significantly greater than what was measured in this analysis. As for TaTLP1, it also has 3 copies located at 7A, 7B and 7D respectively, and their sequences are almost identical, so it is difficult to distinguish them and to identify a single polymorphic nucleotide that could be used to specifically measure independent gene expression (S11 Fig). Therefore, qPCR experiments likely measure the combined expression of TaTLP1 and its homologs.
Taken together, our findings indicate that the TaTLP1-TaPR1 complex positively contributes to wheat resistance to leaf rust and regulates ROS generation in the interaction between wheat and Pt (S12 Fig), which will broaden our understanding of the potentially versatile functions of PR proteins and provide valuable insight into the molecular mechanism of plant immunity. Because VIGS is a transient expression system, we found that the levels of expression for TaPR1 and TaTLP1 are higher at 48 hpi compare to 24 hpi and then decrease again at later time point after inoculation with Pt, which is similar to what was found during pathogenesis by stripe rust [42]. In addition, under diseased conditions where there are many proteins which are differentially expressed, the complex responses to pathogenesis can alter binding affinity between these two proteins. In the future, we plan to evaluate the resistance ability of transgenic plants combining both two PR proteins because it is likely that PR proteins may be effective only in combination with select other PR proteins [43].

Plant materials and leaf rust isolate
Wheat seedlings of cultivar TcLr19 (Tc � 6/RL6040), which is a near-isogenic line containing the Lr19 gene, was requested from the Cereal Disease Lab of USDA located at University of Minnesota were preserved in the Laboratory of Leaf rust, Hebei Agricultural University. Pt pathotype 07-10-426-1 (PHNT), collected from China and preserved in our laboratory was used to inoculate wheat according to Roelfs' standard [44]. For plant extraction, second leaves inoculated with PHNT or sterile distilled water (control) were harvested at 0, 24, 48, 72, 96, 120, 144, 168 hours post inoculation (hpi). All samples were immediately frozen in liquid nitrogen after harvesting and stored at -80˚C. Each treatment included three independent biological replicates.

Binary vector construction, Agrobacterium-mediated transformation, and transient transformation assays
For transient expression, full length TaTLP1 was PCR-amplified from wheat infected with Pt using primers TaTLP1-pENTR-F and TaTLP1-pENTR-R. The truncated TaTLP1 lacking the signal peptide sequence was amplified using primers ΔspTaTLP1-pENTR-F/ΔspTaTLP1-pENTR-R. The PCR products were introduced into entry vector pENTR-D-TOPO (Invitrogen) to obtain TaTLP1-pENTR and ΔspTaTLP1-pENTR which were recombined into the binary vector pEarlyGate103 (with green fluorescence protein (GFP) tag) and resulted in constructs TaTLP1-pEarlyGate103 and ΔspTaTLP1-pEarlyGate103 respectively. Agrobacteriummediated transient expression in N. benthamiana was performed according to methods described previously [45]. The truncated TaPR1 lacking the signal peptide sequence was amplified using primers ΔspTaPR1-pENTR-F/ΔspTaPR1-pENTR-R and the PCR products were introduced into entry clone pENTR-D-TOPO (Invitrogen) to obtain ΔspTaPR1-pENTR. All pENTR clones were confirmed by sequencing.
For CoIP, ΔspTaTLP1-pENTR was recombined into the binary vector pGWB414 containing the C-terminal Hemagglutinin (HA) tag fused to TaTLP1. ΔspTaPR1-pENTR was recombined into the binary vector pEarlyGate103 containing the C-terminal GFP tag fused to TaPR1. The resulting plasmids ΔspTaPR1-pEarlyGate103 together with construct ΔspTaTLP1-pGWB414 were transformed into A. tumefaciens strain GV3101.

Protein extraction, pull-down experiments and protein identification
Proteins were extracted from N. benthamiana leaves 48 h after infiltration with a mixture of A. tumefaciens GV3101 containing corresponding constructs as previously described [46]. For pull-down assays, Chromotek GFP-Trap was used to bind GFP-tagged protein, Anti-GFP Affinity Gel beads were added into the protein mixture and the manufacturer's instructions were followed. The GFP-Trap products were separated by 15% SDS-PAGE and visualized by coomassie blue staining, then transferred onto membranes (NC) using wet blotting overnight. BSA (3%) was used as a blocking agent. A 1:1000 dilution of anti-GFP (Pierce) was used. The secondary antibody goat-anti-mouse (Pierce) was used at a 1:1000 dilution. The luminescent signal was visualized by Immobilon Western Chemiluminescent HRP Substrate using the Che-miDoc Touch Imaging System (Bio-Rad). The pull-down and western blot assays were repeated at least twice. Different bands in TaTLP1-pEarlyGate103 extraction were compared to control bands from pEarlyGate103 empty vector extraction, cut off from the gel, and then subjected to mass spectrometry entrusted to the Beijing Genomics Institute (BGI).

RNA isolation and quantitative PCR
Total RNA was extracted using the TaKaRa MiniBEST Universal RNA Extraction Kit (TaKaRa) according to the manufacturer's instructions. Contaminating DNA was digested with amplification-grade DNase I (TaKaRa) and cDNA was synthesized using the SuperScript II reverse transcriptase, oligo (dT)-18T, and RNase Inhibitor (TaKaRa). qPCR was performed with the LightCycler 96 (Roche). Expression of TaTLP1 and TaPR1 were investigated, and the wheat glyceraldehyde-3-phosphate dehydrogenase (GAPDH, GenBank accession No. AF251217) gene was used to calibrate the expression level of the query genes, as previously described [29] (S3 Table). In addition, the expression levels of superoxide dismutase (TaSOD, GenBank accession number CB307850) and catalase (TaCAT, GenBank accession number X94352) genes were analyzed by qPCR in order to confirm reduction of ROS accumulation in silenced plants. Quantification of the target gene was assessed by relative standard curves. The 2 −ΔΔCT method [47] was employed to quantify the relative gene expression levels. Statistical analysis Microsoft Excel software was used to calculate mean values and standard errors (SE). The statistical significance of differences was calculated using Student's t-test by the SPSS 21.0 (IBM SPSS Statistic) to obtain the P-value. Data were shown as mean ± SE of three biological replicates from one representative experiment. Significant differences between treatments and controls or between time-course points were represented by asterisk ( � p < 0.05).

Yeast two-hybrid assays
The Matchmaker Gold Yeast Two-Hybrid System (Clontech) was used to verify the interaction between TaTLP1 and TaPR1. TaPR1 (without signal peptide) was introduced into pGBKT7 as bait, while TaTLP1 (without signal peptide) was introduced into pGADT7 as prey. The bait and prey plasmids were co-transformed into yeast strain Y187 according to the manufacturer's instructions. Truncated TaPR1 mutants with different deletions in the N-or C-terminal region were generated by PCR amplification and were subcloned into pGADT7 (S3 Table). Cysteines were mutated to alanines in TaTLP1 by polypeptide synthesis. Transformed yeast cells were assayed for growth on synthetic dropout SD/-Trp-Leu plates and SD/-Trp-Leu-His-Ade plates containing X-α-Galactosidase (X-α-Gal) and Aureobasidin A (AbA).

BiFC and co-localization assays
A gateway vector system enabling large-scale combination and investigation of candidate proteins was used for BiFC analysis as described previously [31]. Protein expression was confirmed by western blotting using the appropriate antibody. Then, plasmids of ΔspTaTLP1-VYNE-GW and ΔspTaPR1-VYCE-GW were co-transformed into N. benthamiana. The interaction of AvrLm1 with BnMPK9 was used as the positive control [48]. Fluorescence in leaves of N. benthamiana was monitored at 2 d after agroinfiltration, and then imaged directly using a confocal laser scanning microscope (CLSM) (Olympus FluoView FV1000).

Co-IP assay
To validate the protein interactions of TaTLP1 and TaPR1, Co-IP assays were performed as follows. Total protein from N. benthamiana leaves 48 h after infiltrating with mixture of A. tumefaciens GV3101 containing either ΔspTaPR1-pEarlyGate103 or ΔspTaTLP1-pGWB414 were isolated by homogenizing tissues with RIPA buffer 3 d after agroinfiltration, and were subjected to α-HA IP. Eluted proteins were subjected to immunoblot analysis with anti-GFP antibody. Approximately 25 μL of RIPA buffer containing the total proteins was loaded as an input control.

Antifungal activity assays
TaTLP1 and TaPR1 were cloned into the BamH I and EcoR I restriction sites of plasmid pGEX-6P-3 to generate the plasmid TaTLP1-GST, TaPR1-GST, respectively. The recombinant plasmids were transformed into E. coli Top10 for expression. Proteins were detected by SDS-PAGE, and were frozen at -80˚C in small aliquots until use. The Pt strains PHNT were tested for their response in the presence of pure TaTLP1-TaPR1 complexes, TaTLP1 alone and TaPR1 alone respectively. According to the experiment used to detect germination of urediniospores, petri dishes were filled with 15ml agar medium containing purified TaTLP1 or TaPR1 (1mg/ml), or TaTLP1 plus TaPR1 (1mg/ml). A final concentration of 116.67 ml -1 spore of PHNT was added into the agar medium. The plates were incubated at 25°C in the dark and the germination of spores, hyphal length, and branches were observed with the Nikon Ti2-LAPP Ti2 Laser Application System (Nikon Corporation). The same experiment was also carried out to evaluate growth inhibition with sterile water, buffer instead of the protein solution used as negative controls.

Virus-induced gene silencing (VIGS)
In order to confirm the specificity of the silenced fragments, fragments were selected that had no or lower similarity with other genes to prevent off-target, then one cDNA fragments of TaTLP1 and two cDNA fragment of TaPR1 for gene silencing with Not I and Pac I restriction sites (S3 Table) were obtained by reverse transcription PCR (RT-PCR) and inserted into the original barley stripe mosaic virus (BSMV) vector [49]. In vitro transcription was performed using the mMESSAGE mMACHINE Kit High Yield Capped RNA Transcription Kit (Ambion) with linearized plasmid as template according to the manufacturers protocol. Then 240 μl of the BSMV mixture was applied to fully expanded leaves of 2-leave stage TcLr19 wheat plants by rubbing according to the method described previously [50]. Leaf rust urediniospores were inoculated on the fourth and fifth leaves of TcLr19 and the susceptible wheat cultivar Thatcher. Sterile water was inoculated to wheat leaves as control. The more details of VIGS assay were performed according to a previous study [24]. The feasibility and silencing efficiency were tested using the wheat phytoene desaturase (TaPDS) as a positive control.

Histological observations of fungal growth and host responses
Auto-fluorescence of attacked mesophyll cells was observed in necrotic areas by epifluorescence microscopy (excitation filter, 488 nm; dichromic mirror, 510 nm; and barrier filter, 520 nm). 3, 3-Diaminobenzidine (DAB; Coolaber) staining was conducted to detect H 2 O 2 accumulation following the protocols described previously [51] and was then viewed by differential interference contrast optics. A minimum of 50 infection sites were examined on each of five randomly selected leaf segments for every treatment. The hyphal length, the branching of hyphae and H 2 O 2 accumulation were observed with the Nikon Ti2-LAPP Ti2 Laser Application System (Nikon Corporation). Two mutated TaTLP1 proteins with five cysteines to alanines at N-terminal region (ΔNcys-TaTLP1) and another five cysteines to alanines at C-terminal region (ΔCcys-TaTLP1), and four individual mutated cysteine at C-terminal (ΔCcys1-TaTLP1, ΔCcys2-TaTLP1, ΔCcys4-TaTLP1, ΔCcys5-TaTLP1) that were used for Y2H with TaPR1 respectively. All transformants are able to grow on synthetic dropout medium without leucine and tryptophan (SD-WL) medium. Yeast colonies that were able to grow on selective medium (SD medium without leucine, tryptophan, histidine, and adenine supplemented with X-α-Gal and AureobasidinA [SD-HAWL]) and displayed blue coloration confirmed the protein-protein interaction. Positive control is MAP-K9-AvrLm1. TaPR1 and pGADT7, TaTLP1 and pGBKT7, pGADT7 and pGBKT7 are used as the negative controls.