Activation of defense against Phytophthora infestans in potato by down-regulation of syntaxin gene expression

Summary

• •

  The oomycete Phytophthora infestans is the causal agent of late blight, the most devastating disease of potato. The importance of vesicle fusion processes and callose deposition for defense of potato against Phytophthora infestans was analyzed.

• •

  Transgenic plants were generated, which express RNA interference constructs targeted against plasma membrane-localized SYNTAXIN-RELATED 1 (StSYR1) and SOLUBLE N-ETHYLMALEIMIDE-SENSITIVE FACTOR ADAPTOR PROTEIN 33 (StSNAP33), the potato homologs of Arabidopsis AtSYP121 and AtSNAP33, respectively.

• •

  Phenotypically, transgenic plants grew normally, but showed spontaneous necrosis and chlorosis formation at later stages. In response to infection with Phytophthora infestans, increased resistance of StSYR1-RNAi plants, but not StSNAP33-RNAi plants, was observed. This increased resistance correlated with the constitutive accumulation of salicylic acid and PR1 transcripts. Aberrant callose deposition in Phytophthora infestans-infected StSYR1-RNAi plants coincided with decreased papilla formation at penetration sites. Resistance against the necrotrophic fungus Botrytis cinerea was not significantly altered. Infiltration experiments with bacterial solutions of Agrobacterium tumefaciens and Escherichia coli revealed a hypersensitive phenotype of both types of RNAi lines.

• •

  The enhanced defense status and the reduced growth of Phytophthora infestans on StSYR1-RNAi plants suggest an involvement of syntaxins in secretory defense responses of potato and, in particular, in the formation of callose-containing papillae.

Introduction

The oomycete Phytophthora infestans is the causal agent of late blight, the most devastating disease of potato. The hemibiotrophic oomycete is able to invade, colonize and fulfil its life cycle on susceptible potato plants within several days to weeks. Resistance against P. infestans in commercially grown potato cultivars is conferred by resistance (R)-genes, which are, however, easily overcome by the pathogen (Fry, 2008). The rapid adaptability of the oomycete is considered to result from the presence of fast-evolving effector genes in repeat-rich regions in the P. infestans genome (Haas et al., 2009). Enhanced resistance against P. infestans can also be achieved by treatment of susceptible potato cultivars with chemicals such as DL-β-aminobutyric acid, which enhance the defense status of the plant (Si-Ammour et al., 2003; Altamiranda et al., 2008; Eschen-Lippold et al., 2010; Liljeroth et al., 2010; for general reviews see Jakab et al., 2001; Cohen, 2002), in a salicylic acid (SA)-dependent manner (Eschen-Lippold et al., 2010). A strong induction of local and systemic defense in potato is, moreover, triggered in response to application of the pathogen-associated molecular pattern (PAMP) Pep-13, a 13-amino-acid motif from an extracellular transglutaminase of Phytophthora (Brunner et al., 2002). PAMP-triggered defense responses in potato include the accumulation of the signaling molecules SA and jasmonic acid (JA), defense gene activation and hypersensitive cell death (Halim et al., 2004). In contrast to the antagonistic interaction of SA and JA during pathogen defense in Arabidopsis, Pep-13-induced responses in potato require both SA and JA (Halim et al., 2009).

The plant’s secretory system is an important factor for resistance at the cell periphery (Lipka et al., 2007). A coordinated up-regulation upon pathogen attack is observed for components of the secretory system involved in folding, modification and transport of defense-related proteins (Wang et al., 2005). The secretory machinery enables attacked cells to transport antimicrobial compounds and cell wall material to the site of attempted penetration (Schulze-Lefert, 2004; Hückelhoven, 2007; Kwon et al., 2008). This involves preceding focal reorganization of cellular actin microfilaments and cytoplasmic streaming towards the infection site as well as local organelle aggregation (Takemoto & Hardham, 2004). Transport via the plant secretory system relies on vesicles as cargo transporters, which fuse with the target membrane for cargo delivery. This machinery is involved not only in pathogen defense, but also in other fundamental developmental as well as stress acclimation processes (Carter et al., 2004; Lipka et al., 2007). Selective fusion is mediated by vesicle- and target membrane-resident SNARE-proteins (v-/t-SNAREs; SOLUBLE N-ETHYLMALEIMIDE SENSITIVE FACTOR ATTACHMENT PROTEIN RECEPTOR). These interact to form the SNARE complex, thereby providing specificity and the energy needed for membrane fusion (Weber et al., 1998; Jahn & Scheller, 2006; Diao et al., 2010). SNARE proteins share structural motifs, designated either Q_a, Q_b or Q_c with a conserved glutamine residue (t-SNAREs), or R with a conserved arginine residue (v-SNAREs; Fasshauer et al., 1998; Bock et al., 2001; Pratelli et al., 2004). Vesicle fusion processes are furthermore regulated by additional proteins, for example SEC1/MUNC18-family proteins, Rab-type small GTPases and different coat proteins (Hong, 2005; Jahn & Scheller, 2006; Südhof & Rothman, 2009).

SOLUBLE N-ETHYLMALEIMIDE-SENSITIVE FACTOR ADAPTOR PROTEIN 33 (SNAP33) is important for host cell entry of the barley pathogen Blumeria graminis f.sp. hordei (Bgh; Pajonk et al., 2008). In Arabidopsis, the t-SNARE protein AtPEN1 (= AtSYP121; SYNTAXIN OF PLANTS 121) has been shown to be required for nonhost penetration resistance against Bgh; (Collins et al., 2003). The barley PEN1-homolog REQUIRED FOR MLO-SPECIFIED RESISTANCE (ROR2) is required for penetration resistance against Bgh (Freialdenhoven et al., 1996). Thus, the functional conservation of defense-associated exocytosis was demonstrated both in mono- and dicotyledonous plants, and in nonhost and basal resistance (Collins et al., 2003).

In addition to the two t-SNAREs, AtSYP121 and AtSNAP33, the formation of a functional SNARE complex in Arabidopsis requires the v-SNAREs VESICLE-ASSOCIATED MEMBRANE PROTEIN 721/722 (AtVAMP721/AtVAMP722; Kwon et al., 2008). Apart from this defense-related function, the AtSYP121 protein has recently been shown to be part of a tripartite SNARE-K⁺ channel complex, whose assembly induces channel gating and thus mediates K⁺ transport (Honsbein et al., 2009; Grefen et al., 2010a). In tobacco, the PEN1 homolog SYNTAXIN RELATED PROTEIN 1 (NtSYR1) functions in hormonal guard cell control (Leyman et al., 1999). Here, expression of a soluble inhibitory fragment of NtSYR1 led to elimination of ABA-induced stomatal closure by suppression of the transient elevation of cytosolic Ca²⁺ (Sokolovski et al., 2008).

To elucidate the importance of SNARE proteins involved in vesicle fusion for pathogen defense in potato, a functional characterization of StSYR1 and StSNAP33, the potato homologs of AtPEN1 and AtSNAP33, respectively, were performed. We generated transgenic plants expressing StSYR1- and StSNAP33-RNAi constructs and report on their phenotypic as well as biological and biochemical characterization with respect to defense-related processes. Particular attention was directed at effects on infection-induced local callose deposition and the formation of callose-containing papillae.

Materials and Methods

Cloning of plant transformation constructs and generation of transgenic plants

All RNAi fragments and full-length coding sequences were cloned from cDNA which was reverse-transcribed from RNA isolated from potato leaves (Solanum tuberosum L. cv Désirée) infiltrated with Pseudomonas syringae pv. maculicola M2 (Göbel et al., 2002) using the RevertAid™ H Minus First Strand cDNA Synthesis Kit (Fermentas; http://www.fermentas.de). PCR was performed with Pfu/Taq polymerases (Fermentas). In detail, the StSNAP33-RNAi fragment (353 bp) was amplified using the primers 5′-GTTATCACAAGAATGTCAGATG-3′ and 5′-CT-GGTTTACTTTCCAAGCAAGC-3′, designed from a potato expressed sequence tag (EST, TC121755; http://compbio.dfci.harvard.edu/cgi-bin/tgi/tc_ann.pl?gudb=potato). The StSYR1-RNAi fragment (369 bp) was cloned using the primers 5′-CT-ACAGGTCAAAGTGAGACATTC-3′ and 5′-CATCTCATT-TCTTCCATGGCTG-3′, designed from the potato EST (TC122378) with the highest homology to the StPEN1 partial sequence (AY616763) described in Pajerowska et al. (2005). The full-length coding sequences without stop codons were amplified using the primers StSYR1-fwd 5′-ATGAATGATCTTTT-CTCAGGATCG-3′, StSYR1-rev 5′-TTTCTTCCATGGCTG-AATAG-3′ (900 bp) and StSNAP33-fwd 5′-ATGCATGGT-CTTAAGAAGTCTC-3′, and StSNAP33-rev 5′-CTTTCCAAGCAAGCGCCGA-3′ (918 bp). All generated amplicons were subcloned into the pCR8/GW/TOPO^® vector (Invitrogen; http://www.invitrogen.com). In a subsequent step, the RNAi fragments were transferred to the pHELLSGATE8 vector (Wesley et al., 2001) by LR recombination (Invitrogen) to generate the RNAi transformation vectors. The full-length products were used to create C-terminal fluorescence protein fusions by LR recombination (Invitrogen) into pUBC-RFP-DEST (StSNAP33, StSYR1), pUBC-CFP-DEST (StSNAP33; Grefen et al., 2010b) or pK7FWG2.0 (StSYR1; Karimi et al., 2002). Potato plants (S.  tuberosum cv Désirée) were transformed with Agrobacterium tumefaciens AGL-0 (Lazo et al., 1991) carrying the RNAi transformation vectors. Transgenic plants were regenerated as previously described (Feltkamp et al., 1995).

Cultivation and treatment of plants

Potato plants (S. tuberosum cv Désirée) were grown from sterile explants in soil in a phytochamber with 16 h of light (140 μE) at 20°C and 60% humidity. Infiltration experiments were performed with 100 μM of the Phytophthora PAMP Pep-13 (Brunner et al., 2002) or with bacterial suspensions of either untransformed Agrobacterium tumefaciens AGL-0 and Escherichia coli DH5α (OD₆₀₀ = 0.3 in 10 mM MgCl₂) or Pseudomonas syringae pv. maculicola M2 (1 × 10⁸ cfu ml⁻¹ in 10 mM MgCl₂). Infection experiments with P. infestans (CRA208m2; Si-Ammour et al., 2003; 1 × 10⁵ spores ml⁻¹) were performed according to Eschen-Lippold et al. (2007) and with Botrytis cinerea B05.10 (2 × 10³ spores ml⁻¹) as described in Bethke et al. (2009).

RNA expression analyses

Northern analyses were performed as described (Halim et al., 2004). Hybridizations were carried out using radioactively labeled fragments of StSYR1, StSNAP33 and StPR1 (TC126356). cDNA synthesis was performed using RevertAid™ (Fermentas). For quantitative real-time PCR, the Maxima™ Probe qPCR Master Mix (Fermentas) was used. Samples were run on an Mx3005P qPCR System (Agilent, http://www.agilent.com). The following primers and real-time probes were used for StSNAP33-1: 5′-GGGGAGTCTTGGAGGCATA-3′, 5′-CCTGTAATTGGGCGACTTGT-3′and Roche Universal Probe Library Probe #139; for StSNAP33-2: 5′-GAATGCTAGGAATCACTCTGTGAA-3′, 5′-TGTCAGAATCAAAAGGATTTGAAC-3′and Roche Universal Probe Library Probe #131, for StSYR1-1: 5′-TGGATAGATCCAACGCTTCC-3′, 5′-ACAACAGACGTCCTCGTCCT-3′and Roche Universal Probe Library Probe #82 and for StSYR1-2: 5′-TCCCCAAATCAAGAATCACC-3′, 5′-TCGTCTTTAATGGTTTCTACATCTTC-3′ and Roche Universal Probe Library Probe #131.

Determination of SA concentrations

Salicylic acid and salicylic acid glycoside (SAG) were extracted according to the method of Verberne et al. (2002). Further details are given in Halim et al. (2004) and Eschen-Lippold et al. (2010).

Determination of P. infestans and B. cinerea biomass

P. infestans biomass was determined by real-time PCR of oomycete DNA in infected leaf tissue as previously described (Eschen-Lippold et al., 2007). B. cinerea biomass determinations were carried out as previously described (Bethke et al., 2009).

Transient expression and fluorescence microscopy

Particle bombardment and transient expression in onion epidermal cells (Allium cepa L.) were performed according to König et al. (2008). Fluorescence microscopy images were recorded using a Zeiss Axio Imager M1 (Zeiss; http://www.zeiss.de) with standard cyan, green and red fluorescent protein (CFP, GFP and RFP) filters.

Callose detection

Callose staining was performed with aniline blue according to Adam & Somerville (1996). Photographs were taken using a Nikon AZ100 stereo zoom microscope (http://www.nikoninstruments.com). The UV filter combination was 370/36 for the excitation filter, 400 (LP) for the dichromatic mirror and 405 (LP) for the barrier filter.

Electron microscopy

Leaf segments were fixed with 3% glutaraldehyde (Sigma-Aldrich) in sodium cacodylate buffer (SCB), pH 7.2, for 3 h at room temperature, washed with SCB, postfixed with 1% osmiumtetroxide (Carl Roth; http://www.carlroth.com) in SCB, dehydrated in a graded ethanol series, and embedded in epoxy resin (Spurr, 1969). The material was sectioned with an Ultramicrotome S (Leica; http://www.leica-microsystems.com). Ultrathin sections (80 nm) were transferred to formvar-coated grids and poststained with uranyl acetate and lead citrate.

For immunogold labeling, sections were incubated with a monoclonal antibody against β-1,3-glucan (Biosupplies; http://www.biosupplies.com.au) as first antibody and with a goat-anti-mouse 10 nm gold conjugate (Sigma-Aldrich) as secondary antibody. Subsequently, the sections were poststained with uranyl acetate and lead citrate.

The sections were observed with an EM 900 transmission electron microscope (Zeiss SMT; http://www.zeiss.com/smt) at an acceleration voltage of 80 kV. Electron micrographs were taken with a slow scan camera (Variospeed SSCCD camera SM-1k-120; TRS, Moorenweis, Germany).

Phylogenetic analyses

Phylogenetic analyses were conducted in MEGA4 (Tamura et al., 2007). The evolutionary history was inferred using the neighbor-joining method (Saitou & Nei, 1987). The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) is shown next to the branches (Felsenstein, 1985). The trees are drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the Poisson correction method (Zuckerkandl & Pauling, 1965) and are in the units of the number of amino acid substitutions per site. All positions containing gaps and missing data were eliminated from the dataset (complete deletion option). In the final datasets, there were totals of 244 positions (StSNAP33 tree) and 265 positions (StSYR1 tree).

Statistical analyses

Statistical analyses were performed using GraphPad Prism 5 (http://www.graphpad.com).

Results

StSYR1 and StSNAP33 belong to the class of plant SNARE-proteins

The potato homologs of Arabidopsis AtSYP121 (StSYR1) and AtSNAP33 (StSNAP33) were identified in microarray analyses as genes up-regulated upon treatment with Pseudomonas syringae pv. maculicola M2 (Psm; ftp://ftp.tigr.org/pub/data/s_tuberosum/SGED/073_Rosahl/). The full-length cDNAs of StSYR1 and StSNAP33 were amplified from RNA derived from Psm-treated potato leaves.

Comparison of the cDNA sequences to the potato genome sequence (Xu et al., 2011) revealed the presence of one highly homologus gene for StSNAP33 as well as one for StSYR1, with 99% sequence identity each at the nucleotide level of the coding regions (http://potatogenomics.plantbiology.msu.edu; Supporting Information, Fig. S1). In addition, for both StSNAP33 and StSYR1, a gene with 79% (SpSNAP33-2) and 80% (SpSYR1-2) sequence identity at the nucleotide levels is present in the Solanum phureja genome (Fig. S1). Southern blot analyses confirmed the presence of a limited number of StSNAP33 and StSYR1 genes in potato (data not shown). The SpSNAP33 genes consist of five exons, whereas SpSYR1 genes possess two.

Sequence homology analyses at the amino acid level revealed high levels of similarity between StSYR1-1 and tobacco NtSYR1 (AAD11808), as well as between StSNAP33-1 and the deduced amino acid sequence of tomato clone 133885F (BT014507; Table S1). Phylogenetic analyses with homologs of different plant species confirmed that StSYR1-1 and StSNAP33-1 cluster with NtSYR1 and the putative tomato SNAP33-homolog, respectively (Fig. 1). Multiple sequence alignments revealed that both the Q_a domain in StSYR1 and the Q_b- and Q_c-domains in StSNAP33 were highly conserved (Fig. S2a,b). The conserved Q (glutamine) of the Q_c-domain is exchanged for an H (histidine) in the StSNAP33-1, the SpSNAP33-1 protein and in the protein deduced from the tomato EST sequence. Instead, these proteins have a Q residue nine amino acids downstream in the sequence (Fig. S2a), the biological significance of which remains to be elucidated. The protein encoded by SpSNAP33-2, by contrast, contains the conserved Q residue.

Figure 1

Phylogenetic analyses of StSNAP33- (a) and StSYR1-deduced amino acid sequences (b). Evolutionary relationships are shown as optimal trees with the sum of branch lengths equal to 1.6234 (a) and 2.0188 (b). Details are given in the Materials and Methods section.

StSYR1 and StSNAP33 expression is pathogen- and wound-inducible

To verify the expression predicted from the microarray results, RNA was isolated from potato plants (cv Désirée) 8 h after infiltration with Psm. Northern hybridization with radioactively labeled cDNA probes revealed no detectable transcripts in untreated plants, but high accumulation of StSYR1 and StSNAP33 transcripts after bacterial infiltration (Fig. 2a). Moreover, expression of both StSYR1 and StSNAP33 was detectable in leaves 1 d after infection with P. infestans, peaking 3 d after inoculation (Fig. 2b). A transient expression of both StSNAP33 and StSYR1 was observed after treatment of potato plants with the Phytophthora PAMP Pep-13 (Halim et al., 2009), with significant expression 6, 12 and 24 h after infiltration (Fig. 2c). Wounding experiments revealed that StSNAP33 transcripts started to accumulate 30 min after wounding to an abundance that did not change for the duration of the experiment (Fig. 2d). By contrast, StSYR1 was already expressed 10 min after treatment (Fig. 2d). Transcript abundances decreased up to 1.5 h post-wounding and were similar to background values after 2 h. Thus, both StSYR1 and StSNAP33 genes are pathogen-, PAMP- and wound-responsive, being expressed at early time points after treatment.

Figure 2

Defense-related transcript accumulation of StSNAP33 and StSYR1. RNA of differently treated potato (Solanum tuberosum cv Désirée) wildtype plants was isolated and subjected to northern hybridization with radioactively labeled probes of StSYR1 and StSNAP33. rRNA was stained with ethidium bromide to visualize equal loading. Time points are given as: d post inoculation (dpi); or h post inoculation (hpi). (a) Leaves were infiltrated with Pseudomonas syringae pv. maculicola (1 × 10⁸ cfu ml⁻¹; +) or 10 mM MgCl₂ (−) as control treatment (harvest time point 8 hpi). (b) Leaves were drop-inoculated with a Phytophthora infestans zoospore solution (1 × 10⁵ spores ml⁻¹). Samples were collected at the time points indicated. (c) Leaves were infiltrated with 100 μM Pep-13 (harvest time points indicated). (d) Leaves were wounded with a forceps and samples were taken at the time points indicated. The experiments were performed three times with similar results.

StSYR1-1 and StSNAP33-1 are plasma membrane-localized

To determine the subcellular localization of StSNAP33-1 and StSYR1-1, fusion constructs with fluorescent proteins were created and used for transient expression in onion epidermal cells. Fluorescence microscopy revealed that StSYR1-1:RFP colocalized with a plasma membrane-localized marker protein (ATPase:EYFP; Fig. S3a), suggesting plasma membrane localization of StSYR1-1. Using mCherry as a cytoplasmic marker, no colocalization with StSYR1-1:GFP was observed (Fig. S3b). Colocalization of StSYR1-1 and StSNAP33-1 at the cell periphery was demonstrated using either StSYR1-1:RFP and StSNAP33-1:CFP (Fig. S3c) or StSYR1-1:GFP and StSNAP33-1:RFP (Fig. S3d). These results are consistent with previously reported plasma membrane localization of AtPEN1 (Collins et al., 2003) and AtSNAP33 (Heese et al., 2001) and reveal a colocalization of StSYR1-1 with StSNAP33-1 at the plasma membrane.

Generation of potato plants with reduced expression of StSNAP33 and StSYR1

To assess the significance of StSNAP33 and StSYR1 for defense responses in potato, appropriate RNAi constructs targeted at the 3′-ends of both genes, comprising the coding region for the Q_a- and Q_c-domains of mature StSYR1 and StSNAP33 proteins (Fig. S2), respectively, were cloned under the control of the 35S promoter. The silencing constructs were transferred to potato plants via Agrobacterium tumefaciens-mediated transformation. Positive transformants among the regenerated plants were identified by Southern analyses (data not shown) and assayed for a reduction in StSNAP33 or StSYR1 gene expression after infiltration of Psm. For both StSNAP33 and StSYR1, several independent RNAi lines with highly reduced transcript abundances were obtained (Fig. 3a). Transgenic lines which did not show an RNAi effect were subsequently used as controls. Using gene-specific primers in quantitative RT-PCR experiments, a concomitant down-regulation of StSNAP33-1 and -2, as well as StSYR1-1 and -2 transcripts in the respective RNAi lines was observed (Fig. S4).

Figure 3

Phenotype of StSNAP33- and StSYR1-RNAi potato (Solanum tuberosum cv Désirée) plants. (a) StSNAP33-RNAi, StSYR1-RNAi and control lines were grown for 3 wk. Leaves were infiltrated with Pseudomonas syringae pv. maculicola (1 × 10⁸ cfu ml⁻¹) and sampled for RNA extraction 8 h later. Northern hybridization was carried out using radioactively labeled probes of StSNAP33 and StSYR1. rRNA was stained with ethidium bromide to visualize equal loading. WT, wildtype plant; EV, empty vector-transformed plants; F3, ineffective StSNAP33-RNAi line; A1–F4, effective StSNAP33-RNAi lines; D1, U2, ineffective StSYR1-RNAi lines; B4–W1, effective StSYR1-RNAi lines. Photographs show phenotypes of untreated leaves of the same developmental age of StSNAP33-RNAi (6-wk-old) (b) and StSYR1-RNAi (5-wk-old) (c) plus the respective control plants. The experiments were performed three times with similar results.

StSNAP33 and StSYR1-RNAi plants exhibit a chlorosis phenotype and contain enhanced SA concentrations

After c. 4 wk of growth on soil in a phytochamber under controlled standard conditions, the effective StSNAP33- and StSYR1-RNAi plants developed spontaneous necrosis and chlorosis. Starting on the lower leaves, these symptoms appeared all over the plants within another 2 wk, leading to a complete cessation of growth and occasional death of the plants, whereas control plants, including the ineffective RNAi lines, showed normal senescence on the lowest leaves at this time point. This phenotype generally developed faster and a little earlier on StSYR1- than on StSNAP33-RNAi plants. Fig. 3(b,c) shows a comparison of leaves of the same developmental stage of control plants and ineffective RNAi lines. Chlorosis was not restricted to leaves, but was also visible on stems of RNAi lines (data not shown). In all subsequent experiments, 3-wk-old plants were used before the onset of chlorosis and necrosis, unless stated otherwise.

The chlorosis phenotype is reminiscent of that of the Arabidopsis syp121syp122 double mutant, which exhibits enhanced SA concentrations as well as enhanced SA-induced gene expression (Zhang et al., 2007). We therefore determined the concentrations of SA and SAG in StSNAP33- and StSYR1-RNAi lines. In both cases, the amounts of free SA were significantly increased in untreated leaves of 3-wk-old RNAi plants compared with the controls. Moreover, in StSYR1-RNAi lines, SAG concentrations were elevated as well (Fig. 4a,c). Infiltration of the PAMP Pep-13 led to strong accumulation of SA and SAG in all plant lines, and no differences between RNAi plants and control plants were measured (Fig. 4a,c). SA concentrations of the individual lines are shown in Fig. S5. Thus, the stimulus-dependent maximum SA accumulation is not altered in the RNAi lines despite higher basal concentrations of SA.

Figure 4

Altered salicylic acid (SA) contents in StSNAP33- and StSYR1-RNAi plants. (a, c) Potato (Solanum tuberosum cv Désirée) plants were grown for 3 wk. Samples of untreated as well as Pep-13-infiltrated leaves 24 h post inoculation (hpi) were taken from control plants, StSNAP33-RNAi (a) and StSYR1-RNAi plants (c) and subjected to SA measurements. Absolute concentrations of free SA and salicylic acid glucoside (SAG) are shown. Diagrams show combined data of three independent experiments (n ≥ 12). Control, empty vector-transformed plants and ineffective StSNAP33-RNAi line (F3) or ineffective StSYR1-RNAi lines (D1, U2); StSNAP33-RNAi, lines A1–F4; StSYR1-RNAi, lines B4–W1. *P < 0.05; ***P < 0.001; Mann–Whitney test. Error bars represent SEM. (b, d) Potato plants were grown for 4 wk until the onset of spontaneous chlorosis and necrosis. RNA was extracted from leaf samples of untreated control plants, StSNAP33-RNAi plants (b) and StSYR1-RNAi plants (d). Northern hybridization was performed with a radioactively labeled probe of StPR1. rRNA was stained with ethidium bromide to visualize equal loading. The experiment was repeated twice with similar results. EV, empty vector-transformed plants.

StPR1 gene expression, which is inducible by SA (Eschen-Lippold et al., 2010), was analyzed in RNAi lines of different ages. No difference in StPR1 expression was observed in 2-wk-old StSNAP33-RNAi plants (Fig. S6a). Concomitant with the enhanced SA concentrations in 3-wk-old plants, StPR1 transcripts started to accumulate in untreated StSYR1-RNAi lines after c. 3 wk of cultivation in the phytochamber (Fig. S6b). With the onset of chlorosis formation in 4-wk-old plants, StPR1 gene expression was significantly elevated in untreated StSNAP33- and StSYR1-RNAi lines compared with untreated control plants or ineffective RNAi lines (Figs 4b,d, S6a). Thus, there was a correlation of reduced StSYR1 and StSNAP33 gene expression, higher SA concentrations and enhanced StPR1 gene expression.

StSYR1-RNAi plants are more resistant to P. infestans

Salicylic acid is required for basal defence of potato against P. infestans (Halim et al., 2007). To assess whether the increase in endogenous SA concentrations and enhanced StPR1 gene expression had an impact on resistance, leaves of young RNAi plants were infected with the hemibiotrophic oomycete P. infestans. Three days after inoculation, all effective RNAi plants displayed enhanced necrosis formation not only at the infection sites, but also all over the leaf lamina as a type of runaway cell death (Fig. 5a). Quantitative real-time PCR was performed to measure pathogen biomass at the DNA level. StSYR1-RNAi lines allowed 35% less pathogen growth than the controls and thus displayed a higher resistance (Fig. 5b; for values determined in individual plants see Fig. S7a). This difference was not observed for the StSNAP33-RNAi lines, which displayed the same susceptibility as control plants (Fig. 5b). Infiltration of StSNAP33- and StSYR1-RNAi plants with the Phytophthora PAMP Pep-13 alone did not lead to alterations in the hypersensitive cell death compared with control plants (data not shown). To further characterize the defence status of StSNAP33- and StSYR1-RNAi lines, infection experiments with the necrotrophic fungus B. cinerea were carried out, but no significant differences in pathogen growth could be measured using quantitative real-time PCR (Fig. 5c; for values determined in individual plants see Fig. S7b).

Figure 5

Infection experiments with Phytophthora infestans and Botrytis cinerea. (a) StSNAP33-RNAi, StSYR1-RNAi and control plants were grown for 3 wk. Leaves were drop-inoculated with P. infestans (1 × 10⁵ spores ml⁻¹; 10 μl droplets). Photos and samples were taken 3 d later. Whole leaves and single inoculation sites are shown. The experiment was repeated twice with similar results. WT, wildtype plant; EV, empty vector-transformed plants; F3, ineffective StSNAP33-RNAi line; A1–F4, effective StSNAP33-RNAi lines; D1, U2, ineffective StSYR1-RNAi lines; B4–W1, effective StSYR1-RNAi lines. (b) Plants were treated as described in (a). Pathogen biomass was determined by quantitative real-time PCR. Diagrams show combined data of three independent experiments (n ≥ 55; control, empty vector-transformed plants and ineffective RNAi lines; for data on individual StSYR1-RNAi and StSNAP33-RNAi lines see Fig. S7a; different letters indicate classes of statistically significant differences; Kruskal–Wallis and Dunn’s multiple comparison tests; error bars represent SEM). (c) StSNAP33-RNAi, StSYR1-RNAi and control plants were grown for 3 wk. Leaves were drop-inoculated with B. cinerea (2 × 10³ spores ml⁻¹; 10 μl droplets) and sampled 3 d later. Pathogen biomass was determined by quantitative real-time PCR. Diagrams show combined data of three independent experiments (n ≥ 20; for data on individual StSNAP33-RNAi and StSYR1-RNAi lines, see Fig. S7b). (u), relative units.

In addition, infiltration experiments with bacteria revealed that both Escherichia coli (strain DH5α) and A. tumefaciens (strain AGL-0) induced strong cell death responses in the infiltrated areas of the RNAi plants (Fig. S8). In control plants, only E. coli weakly elicited cell death.

StSYR1-RNAi plants display aberrant callose deposition

As part of the defense response of potato against P. infestans, callose is deposited at the penetration sites (Vleeshouwers et al., 2000; Halim et al., 2007). Callose was visualized by aniline blue staining of P. infestans-infected StSNAP33-RNAi and StSYR1-RNAi lines as well as in wildtype and empty vector-transformed plants. In StSNAP33- and StSYR1-RNAi lines, callose was detected in dot-like structures, whereas larger areas surrounding the necrotic lesions were stained in wildtype and empty vector plants (Fig. 6a). There was a higher percentage of infection sites displaying this aberrant callose deposition in StSNAP33- and StSYR1-RNAi lines compared with the control plants (Fig. 6b).

Figure 6

Callose staining of Phytophthora infestans penetration sites. (a) StSNAP33-RNAi, StSYR1-RNAi and control plants were grown for 3 wk. Leaves, drop-inoculated with P. infestans (1 × 10⁵ spores ml⁻¹; 10 μl droplets), were harvested 3 d later. Inoculated leaf material was subjected to callose staining with aniline blue (fluorescing areas). Photos were taken from single infection sites using a stereo zoom microscope. WT, wildtype plants; EV, empty vector-transformed plants; StSNAP33- and StSYR1-RNAi lines are indicated; n = 40). Bars, 500 μm. (b) Inoculation sites derived from (a) were scored for the presence of aberrant, dot-like callose depositions. Results are expressed as percent values, StSNAP33-RNAi (light gray) and StSYR1-RNAi lines (dark gray) are indicated (n = 40). The experiment was repeated three times with similar results. Error bars represent SEM.

Electron microscopy studies showed infected cells in empty vector-transformed plants, which had reacted with the formation of large papillae at the penetration sites, leading to the encapsulation of Phytophthora structures (Fig. 7a,b; large arrow). Immunogold labeling with an antibody against β-1,3 glucan revealed the deposition of callose specifically in papillae (Fig. 7b; large arrow and enlarged section). The antibody also labeled Phytophthora cell walls, which contain glucans as a structural element (Bartnicki-Garcia & Wang, 1983; Fig. 7b; small arrow in enlarged section). In samples taken from StSNAP33-RNAi plants, no differences from the control were visible; the cells also reacted with the formation of callose-containing papillae (Fig. 7c,d). By contrast, in most P. infestans-infected StSYR1-RNAi samples, no, or only rudimentary, papillae and callose depositions could be observed (Fig. 7e,f). To quantify this effect, infection sites were scored regarding the presence of fully developed, rudimentary or no papillae. This analysis revealed a statistically significant reduction in callose-containing papillae in P. infestans-infected StSYR1-RNAi plants compared with StSNAP33-RNAi lines and control plants (Fig. 8). Thus, these data provide evidence for the involvement of StSYR1 in the proper deposition of callose-containing papillae beneath the sites of attempted penetration.

Figure 7

Electron microscopic investigation of Phytophthora infestans penetration sites. StSNAP33-RNAi, StSYR1-RNAi and control plants were grown for 3 wk. Leaves, drop-inoculated with P. infestans (1 × 10⁵ spores ml⁻¹; 10 μl droplets), were harvested 3 d later and subjected to electron microscopy. Photographs show electron micrographs of independent P. infestans penetration sites in empty vector-expressing control plants (EV) (a, b) as well as StSNAP33-RNAi (c, d) and StSYR1-RNAi (e, f) after labeling of callose with a β-1,3-glucan-specific antibody. Right panels represent magnifications of the boxed areas in the middle panel; large arrows indicate penetration sites (a, c, e) or papillae (b, d, f); small arrows indicate β-1,3-glucan-labeling in P. infestans cell walls. cy, plant cytoplasm; is, intercellular space; pa, papilla; Pi, P. infestans; Pw, P. infestans cell wall; StSNAP33- and StSYR1-RNAi lines are indicated. Electron micrographs were derived from six independent experiments with similar results.

Figure 8

Scoring of penetration sites with callose-containing papillae. Data shown represent the scoring of electron micrographs derived from the experiments shown in Fig. 7; 15–30 individual penetration sites of each StSNAP33-RNAi, StSYR1-RNAi and empty vector-expressing plants were scored for the presence of fully developed, rudimentary or no callose-containing papillae. Calculated scores are shown (3, fully developed papillae; 2, rudimentary papillae; 1,  no papillae). Different letters indicate classes of statistically significant differences; Kruskal–Wallis and Dunn’s multiple comparison tests; error bars represent SEM.

Discussion

Down-regulating the expression of the genes encoding the SNARE proteins StSNAP33 and StSYR1 in potato led to alterations in both development and pathogen defense responses. Both StSNAP33- and StSYR1-RNAi plants displayed an early senescence-like phenotype with spontaneous development of chlorosis and necrosis (Fig. 3). Defects in development were also observed in mutants of the Arabidopsis homologs AtSNAP33 and double mutants of AtSYP121 and AtSYP122, which displayed severe dwarfism and spontaneous necrotic lesions (Heese et al., 2001; Zhang et al., 2007). As described for the Arabidopsis double mutant syp121syp122 (Zhang et al., 2007), the transgenic potato StSYR1-RNAi plants grew normally for several weeks and developed spontaneous leaf necrosis at later stages of development. Single knockout mutants of AtSYP121 or AtSYP122 did not show these defects, suggesting redundancy of their function (Assaad et al., 2004; Zhang et al., 2007). Two genes each for StSNAP33 and StSYR1 are present in the potato genome (http://www.potatogenome.net; Xu et al., 2011) and expression of both was down-regulated by the respective RNAi constructs used here. Thus, a dominant negative effect on StSNAP33 and StSYR1 gene expression was achieved, allowing the functional analysis of SNARE genes of potato.

In correlation with the growth defects at later developmental stages, untreated StSNAP33- and StSYR1-RNAi lines accumulated elevated SA concentrations. This effect was more pronounced in the StSYR1-RNAi lines (Fig. 4a,c), but led to enhanced expression of the SA-dependent defense gene StPR1 in both types of RNAi plants (Fig. 4b,d). In the Arabidopsis atsyp121 single mutant, a similar molecular phenotype was described, which was even more pronounced in the syp121syp122 double mutant (Zhang et al., 2007). Introduction of the SA-signaling mutations eds5 or sid2 in this genetic background abolished constitutive PR1 expression and partially rescued the dwarf phenotype (Zhang et al., 2008). These stimulus-independent defense-like responses are reminiscent of those described for the Arabidopsis lesions simulating disease (lsd) and accelerated cell death (acd) mutants. Both exhibit spontaneous necrosis formation as well, correlating with highly elevated endogenous SA concentrations and PR1 expression (Dietrich et al., 1994; Greenberg et al., 1994; Weymann et al., 1995). Introduction of the bacterial salicylate hydroxylase gene NahG in lsd6/7 mutants abolished these symptoms (Weymann et al., 1995), thereby pointing to a possible interrelation of SA accumulation and developmental defects. To analyze the SA dependence of defense responses in StSYR1-RNAi lines, it is feasible to express the NahG gene in potato, as this has been shown to efficiently abolish SA accumulation in transgenic potato plants (Halim et al., 2004, 2007).

In addition to developmental defects, syntaxin-deficient plants displayed alterations in their response to pathogens. Loss of AtSYP121 or barley HvROR2 resulted in the loss of penetration resistance against the obligate biotrophic fungus Bgh in a nonhost or host interaction, respectively (Collins et al., 2003). In our study, down-regulation of StSYR1, the closest potato homolog to AtSYP121 and HvROR2, did not lead to enhanced susceptibility but rather to a 35% enhanced resistance against P. infestans, which, possibly, is a result of the enhanced cell death response at the site of infection (Fig. 5a). The observation that 3-wk-old StSNAP33- and StSYR1-RNAi plants developed rapid cell death after infiltration of A. tumefaciens, and even of nonpathogenic E. coli (Fig. S8), suggests a hypersensitivity towards otherwise tolerated stresses. Thus, rapid necrosis formation in response to infection with P. infestans (Fig. 5) could be inhibitory to the infection process. The hemibiotrophic oomycete requires a biotrophic phase in which nutrients from living host cells are acquired via intracellular haustoria-like structures. Rapid host cell death might hinder the biotrophic phase of the pathogen. This is in accordance with the enhanced resistance of the Arabidopsis syp121syp122 mutant to the biotrophic pathogen Golovinomyces cichoracearum, which was attributed to the powdery mildew-induced hypersensitive cell death (Zhang et al., 2007). By contrast, growth of the necrotrophic pathogen B. cinerea was not altered in StSNAP33- or StSYR1-RNAi plants (Fig. 5c). It is interesting to note that StSNAP33-RNAi plants also showed the hypersensitivity to stresses, but are not more resistant to P. infestans.

As an alternative mechanism, enhanced resistance might be the result of the higher SA concentrations in StSYR1-RNAi plants. SA is required for basal defense of susceptible potato cultivars against P. infestans, since NahG plants, which were deficient in SA as a result of the expression of a bacterial salicylate hydroxylase, allowed enhanced growth of the oomycete compared with control plants (Halim et al., 2007). Therefore, the higher SA concentrations present in leaves before the onset of spontaneous necrosis might lead to increased resistance in potato. The enhanced susceptibility of NahG plants correlated with compromised callose depositions. In fact, mesophyll cells of NahG plants, penetrated by P. infestans, showed less papillae formation or callose depositions than cells from control plants, suggesting that the lack of SA results in the inability to form callose (Halim et al., 2007). Strikingly, in the present study, StSYR1-RNAi plants had higher amounts of SA, but were unable to form wildtype levels of callose-containing papillae upon infection by P. infestans (Fig. 7). This observation suggests a role for StSYR1, and thus functional vesicle trafficking, for the directional delivery of cell wall material to the site of penetration. The decreased papillae formation and the aberrant callose deposition in StSYR1-RNAi plants has not been observed in the Arabidopsis syp121 mutant, where callose lining of attacked cells is a prominent response to infection by Bgh (Assaad et al., 2004). In addition, the ability to form papillae has not been found to be impaired in syp121 mutants, and only the timing of papilla assembly is delayed by 2 h (Assaad et al., 2004). In a different study, callose accumulation in papillae was also shown to be independent of syp121 (Meyer et al., 2009).

Callose-containing papillae were thought to be important for pathogen defense as they form a physical barrier against the invading pathogen. However, the Arabidopsis mutants powdery mildew resistant 4 (pmr4) or glucan synthase-like 5 (gsl5), which are defective in a gene encoding a callose synthase responsible for the callose depositions at wound and penetration sites, were more resistant against powdery mildew infection (Jacobs et al., 2003; Nishimura et al., 2003). Interestingly, papillae size and shape upon fungal infection were unaffected, despite the loss of callose depositions. The enhanced resistance against otherwise virulent pathogens suggests that callose is not required for resistance and that, in fact, the loss of callose synthase is responsible for resistance. In line with this, pmr4 and gsl5 displayed constitutive SA signaling, leading to the up-regulation of known SA-responsive genes (Nishimura et al., 2003). Infection with G. cichoracearum caused necrosis formation not observed in wildtype plants (Nishimura et al., 2003). Thus, with respect to impaired callose depositions and activation of the SA pathway, loss of callose synthase in Arabidopsis results in similar phenotypes to the down-regulation of StSYR1 in potato. Both syntaxins and callose synthase might be considered to act as negative regulators of SA signaling, leading to spontaneous necrosis and enhanced resistance to several pathogens (Nishimura et al., 2003).

The functional analysis of the SNARE proteins StSYR1 and StSNAP33 in potato thus revealed a role for vesicle trafficking for defense against P. infestans. In particular, reduced StSYR1 expression resulted in a reduced ability of potato plants to form callose-containing papillae in response to infection with P. infestans, suggesting a previously unreported role for StSYR1 in the delivery of papillae components to the site of penetration.

Acknowledgements

Simone Fraas is gratefully acknowledged for excellent technical assistance. We thank TIGR for performing the microarray analyses. Peter Waterhouse (CSIRO) is acknowledged for providing pHELLSGATE8 and Felix Mauch (Université de Fribourg) for P. infestans 208m2. This work was supported by the “Deutsche Forschungsgemeinschaft” SFB 648 (TP A4 to S.R., TP B10 to I.H. and TP Z1 to G.H.).