Wheat TaNPSN SNARE homologues are involved in vesicle-mediated resistance to stripe rust (Puccinia striiformis f. sp. tritici)

Subcellular localisation of SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptors) and their ability to form SNARE complexes are critical for determining the specificity of vesicle fusion. NPSN11, a Novel Plant SNARE (NPSN) gene, has been reported to be involved in the delivery of cell wall precursors to the newly formed cell plate during cytokinesis. However, functions of NPSN genes in plant–pathogen interactions are largely unknown. In this study, we cloned and characterized three NPSN genes (TaNPSN11, TaNPSN12, and TaNPSN13) and three plant defence-related SNARE homologues (TaSYP132, TaSNAP34, and TaMEMB12). TaSYP132 showed a highly specific interaction with TaNPSN11 in both yeast two-hybrid and bimolecular fluorescence complementation (BiFC) assays. We hypothesize that this interaction may indicate a partnership in vesicle trafficking. Expressions of the three TaNPSNs and TaSYP132 were differentially induced in wheat leaves when challenged by Puccinia striiformis f. sp. tritici (Pst). In virus-induced gene silencing (VIGS) assays, resistance of wheat (Triticum aestivum) cultivar Xingzi9104 to the Pst avirulent race CYR23 was reduced by knocking down TaNPSN11, TaNPSN13 and TaSYP132, but not TaNPSN12, implying diversified functions of these wheat SNARE homologues in prevention of Pst infection and hyphal elongation. Immuno-localization results showed that TaNPSN11 or its structural homologues were mainly distributed in vesicle structures near cell membrane toward Pst hypha. Taken together, our data suggests a role of TaNPSN11 in vesicle-mediated resistance to stripe rust.


Introduction
SNAREs (soluble N-ethylmaleimide-sensitive factor attachment protein receptors) are key components in vesicle trafficking in eukaryotic cells (Heese et al., 2001;Wick et al., 2003). Four different SNAREs form a SNARE complex to determine the specificity of intracellular fusion (Fukuda et al., 2000).
Several SNARE genes in plants have been shown to be involved in plant resistance against various pathogens (Inada and Ueda, 2014). For instance, HvSNAP34 in barley (Hordeum vulgare) was reported to participate in callose deposition during non-host resistance to powdery mildew (Collins et al., 2003). Further research on AtSNAP33, the homologue of HvSNAP34 in Arabidopsis thaliana, has revealed the functional SNARE complex PEN1-SNAP33-VAMP721/722 (Wick et al., 2003;Lipka et al., 2008). Recent study shows that SEC11 from Arabidopsis modulates PEN1dependent vesicle traffic by dynamically competing for PEN1 binding with VAMP721 and SNAP33 (Karnik et al., 2013). Tobacco (Nicotiana tabacum) NbSYP132 has been implicated in plant resistance against bacterial pathogen by mediating the secretion of pathogenesis-related protein 1 (Kalde et al., 2007). Another Golgi SNARE AtMEMB12 was targeted by miR393b* and involved in the accumulation of PR1 (Zhang et al., 2011). By interacting with potyviral 6K2 integral membrane protein, the Arabidopsis SNARE protein Syp71 is an essential host factor for successful Turnip mosaic potyvirus infection (Wei et al., 2013).
The Novel Plant SNARE (NPSN) genes are a family of SNARE genes that have no homologues in mammalian or yeast genomes (Sanderfoot et al., 2000). Three NPSN genes, NPSN11, NPSN12, and NPSN13, were identified in both Arabidopsis and rice (Oryza sativa) genomes and reported to locate on the plasma membrane (Zheng et al., 2002;Uemura et al., 2004;Bao et al., 2008a). Further research demonstrated that AtNPSN11 in Arabidopsis was immuno-fluorescently localized on the cell plate where it interacted with a t-SNARE protein KNOLLE during cytokinesis (Zheng et al., 2002). Interestingly, a recent study illustrated that AtNPSN11 might be involved in one of the two KNOLLE-containing tetrameric SNARE complexes, which jointly mediate membrane fusion during cytokinesis (El Kasmi et al., 2013).
So far, the physiological roles of NPSN genes in plantpathogen interactions have not been well characterised. In this study, three NPSNs (TaNPSN11, TaNPSN12, and TaNPSN13) and three plant-defence related SNARE homologues (TaSYP132, TaSNAP34, and TaMEMB12) were cloned from common wheat (Triticum aestivum) cultivar Xingzi9104 and their possible interactions were investigated by yeast two-hybrid and bimolecular fluorescence complementation (BiFC) assays. The transcriptional regulation of the three TaNPSNs and TaSYP132 in wheat response to Puccinia striiformis f. sp. tritici (Pst) was characterized using qRT-PCR and their functions were further tested by virus-induced gene silencing (VIGS) assay. Localization of TaNPSN11 on vesicle structures near cell membrane toward Pst hypha was clarified by immuno-cytochemical methods.

Sequence analysis
The amino acid sequences of six TaSNARE homologues cloned in this study were analysed by Pfam (http://pfam.sanger.ac.uk/) for conserved domains or motifs. TargetP 1.1 (http://www.cbs.dtu.dk/ services/TargetP/) and TMHMM 2.0 (http://www.cbs.dtu.dk/services/TMHMM/) were used to predict localization and trans-membrane domains. Multiple sequence alignments and the Neighbour Joining tree were created using MUSCLE method by MEGA 5.1 (MEGA, Inc.). The same programme was used to perform 1000 bootstrap cycles to estimate the confidence of the different nodes of the tree.
Yeast two-hybrid assay Protein-protein interactions were assayed using the Matchmaker yeast two-hybrid system (Clontech, Mountain View, CA). Fragments of TaNPSN11, TaNPSN12, TaNPSN13, and TaMEMB12 without the trans-membrane region were inserted into pGBK7 vector as bait constructs. Fragments of TaNPSN12, TaNPSN13, TaSNAP33, TaSYP132, and TaMEMB12 without the trans-membrane region were inserted into pGADT7 vector as prey constructs (primers in Supplementary Table S1 available at JXB online). Bait and prey vectors were co-transformed in pairs into the yeast strain AH109. The Leu + and Trp + transformants were isolated and interactions were tested on SD-Trp-Leu-His and SD-Trp-Leu-His-Ade plates as described previously Yang et al., 2013). The control plasmids were provided by the manufacturer.

Bimolecular fluorescence complementation (BiFC) assay
Bimolecular fluorescence complementation (BiFC) assays were performed in tobacco protoplasts as described by Schütze et al. (2009). Based on the positive results from the yeast two-hybrid assays, the full-length TaNPSN11 and TaSYP132 cDNAs were recombined with the N-terminal and C-terminal part of YFP in the vectors pSY736 and pSY735, respectively (primers in Supplementary Table S1 available at JXB online). The fusion proteins were co-expressed in tobacco protoplasts using the polyethylene glycol method. Fluorescence was monitored between 24 and 48 hours after transformation using a Zeiss Axiovert 25 fluorescence microscope with the Zeiss YFP filter cube 46HE (excitation, BP 500/25; beam splitter, FT 515; emission, BP 535/30). Co-transformation of TaHSP90-pSY736 and TaRAR1-pSY735 vectors was used as the positive control and recombinant vectors with corresponding empty vectors were co-transformed as negative controls . For each treatment, numbers of fluorescent cells and observed cells were counted from ten fields of vision under objective 40×. Calculations for the mean, standard error and ratio of fluorescent cells were performed using SPSS 16.0 software (SPSS Inc.).

Pst inoculation and qRT-PCR assays
Seedlings of wheat cultivar Xingzi9104 were maintained and inoculated with Pst avirulent race (CYR23) or virulent race (CYR32) as described by Kang and Li (1984). The wheat plants inoculated with sterile distilled water were used as mock-inoculated controls. The leaves were harvested at 0, 12, 18, 24, 48, 72 and 120 hours postinoculation (hpi) for RNA isolation. These time points were selected based on previous microscopic studies of the interactions between wheat and stripe rust fungi (Wang et al., 2007;Zhang et al., 2012). All samples were rapidly frozen in liquid nitrogen and stored at -80°C. Three independent biological replications were included for each time point.
The RNAs were isolated using Trizol Reagent (Invitrogen, Carlsbad, CA, USA). First-strand cDNA was synthesized using the GoScript Reverse Transcription System (Promega Corp., Madison, WI, USA). Primers for qRT-PCR were designed (Supplementary Table S1 available at JXB online). The wheat elongation factor TaEF-1a (GenBank accession number Q03033) was used as an internal reference for qRT-PCR analyses. Primer efficiencies were calculated using five 4-fold cDNA dilutions (1:1, 1:4, 1:16, 1:64, and 1:256) in duplicate as well as checking for amplification in a negative control without cDNA. Dissociation curves ranging from 60 to 94°C were generated for each reaction to ensure specific amplification. The threshold values (Ct) generated from the ABI PRISM 7500 Software Tool (Applied Bio-systems) were used to quantify relative gene expression using the Delta Ct method as described by Chen and Dubcovsky (2012). Transcript levels for all genes and treatments presented in this study are expressed as linearized fold-EF1a levels calculated by the formula 2 (ACTIN CT -TARGET CT) . The resulting number indicates the ratio between the initial number of molecules of the target gene and the number of molecules of EF1a and therefore the Y scales are comparable across genes and treatments. Calculations for the mean, standard error and two-sample t-tests for the statistics were performed using SPSS 16.0 software (SPSS Inc.).

Barley Stripe Mosaic Virus (BSMV)-mediated TaNPSNs and
TaSYP132 gene silencing The virus-induced gene silencing (VIGS) system is an effective reverse genetic tool in barley and wheat (Lee et al., 2012;Wang et al., 2012;Tang et al., 2013). In this study, the plasmids used for virus-induced gene silencing were constructed as described by Holzberg et al., 2002. A cDNA fragment (120 bp) of the wheat phytoene desaturase gene TaPDS was obtained using RT-PCR. This fragment, in antisense orientation, was used to replace the GFP coding sequence in BSMV-GFP (green fluorescent protein) to generate BSMV-TaPDS. Using a similar approach, BSMV-TaNPSN11, BSMV-TaNPSN12, BSMV-TaNPSN13, and BSMV-TaSYP132 were prepared (primers in Supplementary Table S1 available at JXB online). Possible RNAi off-target effects of these VIGS constructs were tested by si-Fi software against an established durum wheat transcriptome as previously described (Nowara et al., 2010;Krasileva et al., 2013).
Capped in vitro transcripts were prepared from linearized plasmids containing the tripartite BSMV genome using the mMES-SAGE mMACHINE® T7 Transcription Kit (Ambion, Austin, TX, USA). Three independent sets of plants were prepared for each of the six BSMV virus constructs (BSMV-00, BSMV-TaPDS, BSMV-TaNPSN11, BSMV-TaNPSN12, BSMV-TaNPSN13, and BSMV-TaSYP132) using a total of 144 seedlings. Another 24 seedlings were Mock inoculated with 1 × Fes buffer as a control. The second leaf of a two-leaf wheat seedling was inoculated with BSMV transcripts by gently rubbing the surface with a gloved finger, maintained in a growth chamber at 23 ± 2°C, and examined for symptoms at regular intervals (Holzberg et al., 2002;Hein et al., 2005;Scofield et al., 2005). Once virus phenotype was observed, the fourth leaf was inoculated with urediospores of Pst avirulent race (CYR23) or virulent race (CYR32). The stripe rust infection types were recorded at 14 days post-inoculation. The fourth leaves were also sampled at 24, 48 and 120 hours post-inoculation for histological observation and RNA isolation. A series of qRT-PCR assays were applied to test the silencing efficiency for each of the BSMV constructs. Using the comparative threshold (2 -ΔΔCT ) method (Livak and Schmittgen, 2001), relative gene expression of TaNPSN11, TaNPSN12, TaNSN13, and TaSYP132 in wheat plants inoculated with corresponding BSMV-TaNPSN11, BSMV-TaNPSN12, BSMV-TaNPSN13, and BSMV-TaSYP132 was compared with that in BSMV-00, respectively. Calculations for the mean, standard error and two-sample t-tests for the statistics were performed using SPSS 16.0 software (SPSS Inc.).

Histological observations of fungal growth and host response
Wheat leaves infected BSMV were sampled at 24, 48 and 120 hours post-inoculation (hpi) with Pst. For further observation, the leaf samples were discoloured using ethanol and acetic acid. For samples collected at 24 hpi, DAB was used to specifically stain the H 2 O 2 generated at the infection site (Wang et al., 2007). The proportion of H 2 O 2 accumulation at each infection site was measured using Olympus DP70/DP30BW microscopy with Olympus software (DP-BSW Ver. 02. 03, Olympus). Only infection sites with substomatal vesicle formations were considered as successfully penetrated. For samples collected at 48 and 120 hpi, the proportion of phenolic autofluorogens accumulated at the infection site was measured using Olympus DP70/DP30BW microscopy (excitation filter, 485 nm; dichromic mirror, 510 nm; barrier filter, 520 nm). At least 50 infection sites from each of the five randomly selected leaf segments per treatment were examined.
Rust fungal structures were then specifically stained using Calcofluor White (Sigma Co., USA) as described by Kang et al. (1993). The leaf segments were fixed and dehydrated using chloral hydrate for 1 hour. Subsequently, samples were soaked twice in 50% alcohol for 15 min followed by three washes with distilled water. The leaf samples were further dehydrated twice with 0.5 M NaOH for 10 min and washed three times with distilled water. The rust fungal structures were stained using 0.1% Calcofluor White (Tris-HCl buffer, pH 8.5) for 10 min, and immediately washed with distilled water for 10 min. All samples were preserved in 25% glycerol for observation. The Pst hyphae at each infection site were observed using Olympus DP70/DP30BW microscopy (excitation filter, 485 nm; dichromic mirror, 510 nm; barrier filter, 520 nm), and their lengths were calculated by Olympus software (DP-BSW Ver. 02. 03, Olympus). Calculations for the mean, standard error and two-sample t-tests for the statistics were performed using SPSS 16.0 software (SPSS Inc.).

Purification of recombinant TaNPSN11 and western blot
A 564 bp TaNPSN11 cDNA fragment (1-188 aa, 21.19 kDa) was amplified and cloned into the pET-28a (+) vector (Novagen, Madison WI, USA). The construct was then transformed into E. coli strain BL21 (DE3). The expression of TaNPSN11 tagged with six histidine residues at the N-terminus was induced using 0.1 mM isopropyl β-thiogalactoside at 37°C for 4 hours. For recombinant protein purification, the bacterial cells were pelleted after induction, suspended in coupling buffer (20 mM Na 2 HPO 4 pH 7.4, 0.5 mM NaCl, 20 mM imidazole) containing 10 mg ml -1 of lysozyme and subjected to sonication on ice for 30 min using a UP100H Vibra Cell sonicator (Hielscher, Teltow, Germany). The resulting lysate was centrifuged at 4°C for 15 min at 12 000 g and the supernatant was put onto His-Trap HP resin for chromatography using an AKTA Purifier 10 (Amersham Pharmacia Biotech, Uppsala, Sweden). The recombinant protein was eluted with a 20-500 mM imidazole gradient in elution buffer (20 mM Na 2 HPO 4 pH 7.4, 0.5 mM NaCl, 500 mM imidazole). The purified protein was injected into a rabbit to raise anti-TaNPSN11 antibody as described by Harlow and Lane (1999). For the western blot analysis, protein was extracted from wheat leaves using protein extraction buffer: GTEN (10% (v/v) glycerol, 25 mM Tris pH 7.5, 1 mM EDTA, 150 mM NaCl), 10mM DTT, 2% (w/v) PVPP (polyvinylpolypyrrolidone), and 1 × protease inhibitor cocktail (Moffett, 2011). Wheat leaves were harvested and leaf materials cut off on either side of the middle vein with a razor blade. Next, 1 g of leaf tissue was weighed out and placed in a pre-chilled mortar. 2.5 ml of extraction buffer was added to each mortar and ground for 1-2 min to form a consistent slurry. The slurry was poured into a 2 ml Eppendorf tube and spun at full speed in a refrigerated micro-centrifuge for 2 min. Supernatant was transferred to a 1.5 ml Eppendorf tube and spun for an additional 10 min. Supernatant was then separated on a 15% SDS-polyacrylamide gel (SDS-PAGE). The protein was subsequently transferred onto a nitrocellulose membrane using a Semi-Phor Semi-Dry Transfer Unit (Amersham Pharmacia Biotech). The immuno-blot analysis was conducted using the polyclonal antibody raised against the recombinant TaNPSN11 as the primary antibody and a horseradish peroxidase (HRP)-conjugated goat-antirabbit IgG (Sigma-Aldrich, St. Louis, MO) as the secondary antibody. The immuno-reactivity was detected using an ECL Western Blotting Substrate kit (Amersham TM , UK) and photographed.

Immuno-cytochemical localization of TaNPSN11
Leaves of seedling wheat were infected with Pst avirulent race CYR23 and harvested at 48 hours post-inoculation (hpi). The leaf specimens (approximately 2 mm 2 ) were excised from the infected tissues, fixed with 3% glutaraldehyde (v/v) in 100 mM phosphate buffer (pH 6.8) for 3 hours at 4°C, and washed in the same buffer four times for 15 min. The sections were post-fixed with 1% (w/v) O S O 4 in 100 mM phosphate buffer (pH 6.8) and washed in the same buffer four times for 15 min. Following dehydration in a graded ethanol series, the sections were embedded in LR-white. Ultra-thin sections were cut using a vibratome (TPI, Series 1000) at a 45° angle and mounted on grids.
The grids were incubated three times in 5% (w/v) BSA for 30 min and subsequently incubated in a 1:40 000 dilution of the rabbit anti-TaNPSN11 antibody in BSA for 2 hours at room temperature. After washing three times and twice for 10 min in 1% BSA and 75 mM PBS (pH 7.2-7.4), the grids were incubated for 1 h at room temperature in a 1:20 dilution of the colloidal gold (15 nm)-conjugated goat-antirabbit IgG in PBS. After washing twice for 10 min in distilled water, the grids were dried, and the samples were post-stained with uranyl acetate and lead citrate. The sections were examined using a Zeiss-EM10 electron microscope (80 kV) (Kang and Buchenauer, 2002;Liu et al., 2010;Wang et al., 2010).

Results
Cloning of TaNPSNs and plant defence-related SNARE homologues SNARE genes from Arabidopsis, rice and barley were retrieved and blasted against the DFCI wheat EST database and IWGSC wheat genome database to identify the homologous wheat genes. Six wheat SNARE homologues with complete ORFs were cloned from cDNA of wheat cultivar Xingzi9104, and designated as TaNPSN11, TaNPSN12, TaNPSN13, TaSYP132, TaSNAP34, and TaMEMB12 (GenBank accession numbers JX104547, JX104548, JX104549, JX104550, JX104551, and JX104552). Subsequent RACE reactions resulted in cloning of cDNA fragments with 5' and 3' UTR for the three TaNPSNs.

TaNPSN11 interacts with TaSYP132 in yeast two-hybrid and bimolecular fluorescence complementation assays
Yeast two-hybrid assays were used to characterize all pairwise interactions among the three TaNPSNs and three plant defence-related SNARE homologues in wheat. The results showed that TaSYP132 specifically interacted with TaNPSN11 but not with TaNPSN12 or TaNPSN13. Other combinations showed no interactions (Fig. 2).
To further verify the interaction between TaNPSN11 and TaSYP132 in planta, bimolecular fluorescence complementation (BiFC) assays were performed in Nicotiana benthamiana protoplasts. YFP fluorescence was reconstituted and punctate localized to the cell membrane when YFP N -TaNPSN11 and YFP C -TaSYP132 were co-expressed. Stable YFP fluorescence at cytoplasm and nucleus was observed in positive controls using YFP N -TaHSP90 and YFP C -TaRAR1 (Fig. 3). Although the ratio of fluorescent cells in YFP N -TaNPSN11/ YFP C -TaSYP132 (6%) was lower than YFP N -TaHSP90/ YFP C -TaRAR1 (25%), recombinant vectors with corresponding non-fused YFP N and YFP C empty vectors generated no fluorescence (Table 1).

Transcriptional changes of TaNPSNs and TaSYP132 induced by Pst infection
Wheat cultivar Xingzi9104, possibly carrying YrSK and Yr18, was used as a host plant for Pst infection (Liu et al., 2006;Krattinger et al., 2009;Zhang et al., 2012). After inoculating seedling plants of wheat cultivar Xingzi9104 with Pst avirulent race CYR23 and virulent race CYR32, the infection type was scored at 14 days post-inoculation (dpi) using the 0-9 scale method (Supplementary Table S2 available at JXB online) (Line and Qayoum, 1992). Seedlings of Xingzi9104 exhibit an immune or few necrotic flecks phenotype (Scale 0-1) to Pst avirulent race CYR23. Severe sporulation phenotypes (Scale 8-9) could be observed on Xingzi9104 seedlings inoculated with Pst virulent race CYR32 at 14 dpi.
The transcriptional changes of the three TaNPSNs and TaSYP132 induced by Pst infections were measured by qRT-PCR (Fig. 4). In leaves inoculated with Pst avirulent race CYR23, expression level of TaNPSN11 showed a significant up-regulation at 120 hours post-inoculation (hpi); TaNPSN13 was up-regulated at 18 and 24 hpi but suppressed at 12 and 120 hpi; TaSYP132 was dramatically elevated at 12 hpi. In leaves inoculated with Pst virulent race CYR32, TaNPSN13 was significantly up-regulated at 12 and 18 hpi; TaSYP132 was strongly induced at 18 hpi. In contrast, TaNPSN12 did not show any significant inductions by either Pst avirulent or virulent races.
Knocking down TaNPSN11, TaNPSN13, and TaSYP132 expression reduced the resistance of wheat to Pst Based on the expression profiles of the three TaNPSNs and TaSYP132 during Pst infections, a virus-induced gene    silencing (VIGS) system was applied to characterize the role of these SNARE homologues during wheat-Pst interaction. Four pairs of primers were designed specifically to knock down TaNPSN11, TaNPSN12, TaNPSN13, and TaSYP132, respectively. None of the RNAi constructs, except TaNPSN12-VIGS with a weak recognition to Cysteine-rich receptor-like protein kinase 41, was predicted to possess effective off-targets or cross silencing other SNARE transcripts in an established durum wheat transcriptome as determined by the si-Fi software (Supplementary Table S3 available at JXB online). All BSMV-inoculated plants displayed mild chlorotic mosaic symptoms at 9 days post-inoculation (dpi), but no obvious defects were observed during further leaf growth. To confirm whether our VIGS system was functioning correctly, BSMV-TaPDS (BSMV vector carrying a segment from the wheat phytoene desaturase gene) was used as the positive control for the gene silencing system, which generates photo-bleaching in the fourth leaves of the inoculated plants (Fig. 5A).
To determine the efficiency of VIGS, qRT-PCR assays were performed on RNA samples extracted from the fourth leaves of wheat seedlings pre-infected with BSMV-00, BSMV-TaNPSN11, BSMV-TaNPSN12, BSMV-TaNPSN13, and BSMV-TaSYP132 at 24, 48 and 120 hours post-inoculation (hpi) with Pst avirulent race CYR23. Compared with BSMV-00 control, the abundance of the three TaNPSNs and TaSYP132 transcripts was significantly suppressed to different extents in corresponding knockdown plants (Fig. 5C).
After inoculating seedling plants of wheat cultivar Xingzi9104 with Pst avirulent race CYR23, an immune or few necrotic spots phenotype (Scale 0-1) was observed on wheat leaves pre-infected with Mock (buffer inoculated without BSMV), BSMV-00 (empty vector), and BSMV-TaNPSN12, whereas many necrotic spots (Scale 1-2) were observed on leaves preinfected with BSMV-TaNPSN11, BSMV-TaNPSN13, and BSMV-TaSYP132 at 14 dpi ( Fig. 5B; Supplementary Table S2 available at JXB online). Thus, the race-specific resistance to the stripe rust fungus was not blocked or eliminated through silencing the expression of the three TaNPSNs or TaSYP132. Nevertheless, knocking down the expression of TaNPSN11, TaNPSN13 or TaSYP132 resulted in a reduced resistance of wheat to stripe rust fungus. There were no significant phenotype differences between each treatment when inoculated with Pst virulent race CYR32, thus severe sporulation (Scale 8-9) was observed (Supplementary Figure S2 available at JXB online).
To observe the histological changes associated with the enhanced susceptibility to Pst in these knockdown plants, leaf segments from at least three plants inoculated with Pst avirulent race CYR23 were harvested from each treatment. Accumulation of H 2 O 2 at the infection site was observed at 24 hpi using microscopy after DAB staining (Fig. 6C). The H 2 O 2 accumulations in BSMV-TaNPSN13 and BSMV-TaSYP132 pre-infected wheat leaves were significantly (P < 0.05) lower than that in BSMV-00 infected leaves ( Fig. 6C; Table 2), suggesting the potential involvement of TaNPSN13 and TaSYP132 in ROS-mediated plant defence. The phenolic autofluorogen accumulations per infection site in BSMV-TaNPSN11, BSMV-TaNPSN13, and BSMV-TaSYP132 pre-infected wheat leaves were significantly (P < 0.01) lower than control at both 48 and 120 hpi ( Fig. 6A and B; Table 2), indicating that knocking down TaNPSN11, TaNPSN13, and TaSYP132 expression decreases plant defence reactions at infection sites. The leaf samples were further stained using Calcofluor White to enhance the rust hyphal detection (Fig. 6D). The Pst hyphal lengths in BSMV-TaNPSN11, BSMV-TaNPSN13, and BSMV-TaSYP132 pre-infected wheat leaves were significantly (P < 0.05) longer than those observed in BSMV-00-infected leaves at 24, 48 and 120 hpi (Table 2), indicating that knocking down TaNPSN11, TaNPSN13 and TaSYP132 expression enhances hyphal penetration of the rust fungus.

TaNPSN11 localizes in vesicle structures near cell membrane towards the infection site of Pst
To investigate the distribution of TaNPSN11 during the interaction between wheat and Pst, a polyclonal antibody was raised in rabbit against a non-conserved region (amino acids 1-188) of TaNPSN11. A clear band with correct TaNPSN11 size (~21.2 kD) was detected by western blot using protein extracted from Pst-infected wheat leaves (Supplementary Figure S3 available at JXB online), suggesting that our polyclonal antibody binds to TaNPSN11 (although we cannot rule out the unlikely possibility that it also binds to some other proteins with very similar structure and size).
By immuno-cytochemical localization assays, we observed that TaNPSN11 (or very close structural homologues recognized by the antibody) was primarily (approximately 73%) distributed in vesicle structures towards the intercellular hyphae of Pst, especially vesicles near the cell membrane ( Fig. 7; Table 3). In addition, the vesicle transport seemed to be activated and redirected to the infection sites (Fig. 7).

Discussion
A total of 54 SNARE genes have been identified in Arabidopsis including three NPSN genes, a number similar to the ones identified in rice (Uemura et al., 2004;Bao et al., 2008b). Using a comparative strategy, we cloned three NPSN (TaNPSN11, TaNPSN12, and TaNPSN13) and three plant defencerelated SNARE homologues (TaSYP132, TaSNAP34, and TaMEMB12) from the wheat cultivar Xingzi9104. The phylogenetic tree shows that the NPSNs, MEMB12, SYP132, and SNAP34 form four well differentiated groups supported by 100% bootstrap values. The NPSN proteins seem to closer to the MEMB12 than to the other groups (68% bootstrap). Calculations for the mean, standard error and two-sample t-tests for the statistics were performed using the SPSS 16.0 software (b*, P < 0.01, b, P < 0.05).
Within the NPSN cluster, the NPSN forms a well-defined group including both Arabidopsis and grass species, suggesting that NPSN11 differentiated before the divergence between the monocots and dicots. The NPSN12 and NPSN13 clusters are well defined in the grasses but separate from the duplication that originated Arabidopsis NPSN12 and NPSN13, so these two genes have independent sub-functionalization stories in these two lineages.
Our yeast two-hybrid results show that TaSYP132 interacts with TaNPSN11. Since SNARE complex can be formed with low specificity in vitro (Scales et al., 2000), we further validated the interaction in tobacco protoplasts using bimolecular fluorescence complementation (BiFC) assay. The localization of the fluorescent signal confirmed that the interaction punctate occurred at the plasma membrane. Using DUAL membrane protein yeast two-hybrid and immuno-precipitation assays, the interaction between two Arabidopsis SNARE proteins Syp71 and Vap27-1 was verified (Wei et al., 2013). The interaction between another Arabidopsis SNARE protein SYP121 and SEC11 was characterized in detail by a ratiometric bimolecular fluorescence complementation (rBiFC) assay (Karnik et al., 2013). Since both TaNPSN11 and TaSYP132 were expressed with their transmembrane domain in our BiFC assay, the punctate YFP fluorescence at the plasma membrane might indicate that the interaction occurs around the trans-Golgi network (TGN) and plasma membrane (PM). The specificity of TaNPSN11/TaSYP132 interaction also suggests functional differences among NPSN members.In our qRT-PCR assay, expressions of TaNPSN11, TaNPSN13 and TaSYP132 were differentially induced by either Pst avirulent or virulent races, suggesting a possible involvement of these SNARE homologues in wheat response to stripe rust infection, which were further characterized by the virusinduced gene silencing (VIGS) assay. Specifically, phenolic auto-fluorescent compounds and H 2 O 2 accumulations can be used to estimate the strength of the plant defence response toward a pathogen (Holzberg et al., 2002;Wang et al., 2007). The reduced accumulation of phenolic autofluorogens in TaNPSN11, TaNPSN13, and TaSYP132-knockdown plants, together with the significant decrease in hyphal lengths, provides a possible explanation for the more abundant and more visible necrotic areas observed on the surface of Pst-infected leaves of the TaNPSN11, TaNPSN13, and TaSYP132knockdown plants relative to the control. Interestingly, reduced accumulation of reactive oxygen species (ROS) was only observed in the TaNPSN13-and TaSYP132-knockdown plants, indicating that these two SNARE homologues might be involved in the delivery of ROS-related materials. SNARE proteins were normally involved in the plant-microbe interaction by mediating the secretory pathway of functional PRRs or pathogenesis-related (PR) proteins to PM (Beck et al., 2012;Watanabe et al., 2013). Considering the reduced resistance toward Pst in TaNPSN11-and TaSYP132-knockdown plants, as well as their verified protein interaction, we predicted that NPSN11 might complex with SYP132 in the host response to pathogens. In the immuno-localization assays, we observed that TaNPSN11 or its structural homologues was primarily localized on vesicle structures near the plasma membrane towards the Pst infection site, and the vesicle transport seemed to be activated and redirected to the infection sites. During cytokinesis in Arabidopsis, AtNPSN11 was shown to be localized on the newly formed cell plate using the immunofluorescent method (Zheng et al., 2002). Another study illustrated that all three AtNPSN proteins when expressed with GFP-tagging in Arabidopsis protoplasts were localized on the plasma membrane (Uemura et al., 2004). Compared with several electron microscopic observation assays on plant vesicles and Golgi apparatus (Coleman et al., 1988;Boevink et al., 1998;McCarthy et al., 2013), we speculate that the bubbly and layered structures observed in our immuno-localization assay are vesicles and Golgi apparatus, respectively. Consequently, our results might provide a snapshot of the position where the TaNPSN11 or its structural homologues mediate vesicle trafficking between the Golgi apparatus and plasma membrane. In conclusion, we predict a dual function for NPSN11 during cytokinesis and plant-pathogen interactions. In Arabidopsis, NPSN11 is involved in cytokinesis by forming a KNOLLE-NPSN11-SYP71-VAMP721/722 complex (Zheng et al., 2002;El Kasmi et al., 2013), whereas our results indicate that NPSN11 might also be involved in the transport of plant defence-related materials, possibly through interaction with SYP132. The idea that one SNARE might function in different biological pathways by forming different SNARE complexes is well supported by research concerning AtSNAP33/HvSNAP34. Thus, Arabidopsis AtSNAP33 interacts with KNOLLE in cytokinesis whereas it (or its barley homologue HvSNAP34) interacted with PEN1 (HvROR2 in barley) in non-host resistance to powdery mildew (Heese et al., 2001;Collins et al., 2003;Lipka et al., 2008). Our study provides initial evidence that these wheat SNARE homologues play important roles in vesicle-mediated plant immunity, but the detailed molecular mechanisms and their precise role in wheat defence to stripe rust fungi will require further investigation.

Supplementary material
Supplementary data can be found at JXB online.
Supplementary Table S1. Primers designed and used in this study.
Supplementary Table S2. Scales of stripe rust infection type (IT) in wheat plants.
Supplementary Figure S3. Western blot analysis of the polyclonal anti-TaNPSN11 antibody. Anti-TaNPSN11 and HRP-conjugated goat-anti-rabbit IgG antibodies were used as the primary and secondary antibodies in the western blot assay. A clear band with correct TaNPSN11 size (~21.2 kD) was detected by western blot using protein extracted from Pst-infected wheat leaves, suggesting our polyclonal antibody could bind to TaNPSN11 or its structural homologues.