Nucleoporin-Regulated MAP Kinase Signaling in Immunity to a Necrotrophic Fungal Pathogen

Arabidopsis nuclear pore complex protein Nup88/MOS7 interacts with Nup98a/b and promotes nuclear accumulation of MPK3 to mount a robust immune response against the necrotrophic fungus Botrytis cinerea.

Abstract

Pathogen-responsive mitogen-activated protein kinase (MAPK or MPK) cascades relay signals from activated immune receptors across the nuclear envelope to intranuclear targets. However, in plants, little is known about the spatial control of MAPK signaling. Here, we report that the Arabidopsis (Arabidopsis thaliana) nuclear pore complex protein Nup88/MOS7 is essential for immunity to the necrotrophic fungus Botrytis cinerea. The mos7-1 mutation, causing a four-amino acid deletion, compromises B. cinerea-induced activation of the key immunoregulatory MAPKs MPK3/MPK6 and reduces MPK3 protein levels posttranscriptionally. Furthermore, MOS7 contributes to retaining a sufficient MPK3 abundance in the nucleus, which is required for full immunity to B. cinerea. Finally, we present a structural model of MOS7 and show that the mos7-1 mutation compromises interactions with Nup98a/b, two phenylalanine-glycine repeat nucleoporins implicated in maintaining the selective nuclear pore complex permeability barrier. Together, our analysis uncovered MOS7 and Nup98 as novel components of plant immunity toward a necrotrophic pathogen and provides mechanistic insights into how these nucleoporins coordinate nucleocytoplasmic transport to mount a robust immune response.

The plant immune system has evolved the capability to recognize and respond to a broad range of pathogens with diverse life styles and infection strategies. Plants sense pathogens by two major types of immune receptors. At the cell surface, pattern-recognition receptors recognize microbe/pathogen-associated molecular patterns (MAMPs/PAMPs) such as fungal chitin and initiate PAMP-triggered immunity. Intracellularly, nucleotide-binding domain and leucine-rich repeat (NB-LRR)-containing Resistance (R) proteins perceive specific effector molecules of host-adapted pathogens and activate effector-triggered immunity (ETI; Jones and Dangl, 2006). ETI is an amplified PAMP-triggered immunity response and typically involves the accumulation of the defense hormone salicylic acid (SA) and a hypersensitive cell death response at infection sites that restricts the growth of biotrophic and hemibiotrophic pathogens. In contrast, defense against necrotrophic pathogens that kill their host is mediated by the hormones jasmonic acid (JA) and ethylene (ET; Glazebrook, 2005). The antagonistic interplay between SA- and JA-mediated defense signaling pathways appears to enable the plant to fine-tune its immune response to the type of pathogen attacking (Pieterse et al., 2012; Gimenez-Ibanez and Solano, 2013).

Pathogen-responsive mitogen-activated protein kinase (MAPK/MPK) cascades constitute important signaling modules that mediate the phosphorelay of information from activated immune receptors to downstream target proteins involved in immune response regulation. MAPK targets include enzymes and transcription factors (TFs) that regulate the synthesis of defense hormones and antimicrobial compounds or activate the expression of immune response genes (Ren et al., 2008; Han et al., 2010; Mao et al., 2011; Meng et al., 2013). The canonical MAPK cascade consists of a MAPK kinase kinase (MAP3K) that, upon activation by a stimulated receptor, phosphorylates a MAPK kinase (MAP2K), which in turn activates a MAPK via phosphorylation. The output of such a cascade is determined by the phosphorylation of specific MAPK substrates, whose activities, functions, and/or cellular localizations can change significantly (Lee et al., 2015). The Arabidopsis (Arabidopsis thaliana) genome encodes for 20 MAPKs (MAPK Group, 2002), of which the functions of the pathogen-responsive MPK3, MPK4, and MPK6 in disease resistance are characterized best. MPK3 and MPK6 are closely related and show a high level of functional redundancy; they are activated by the redundant MAP2Ks MKK4/MKK5 and positively regulate plant defense responses (Asai et al., 2002; Ren et al., 2008; Pitzschke et al., 2009; Meng and Zhang, 2013). MPK4 constitutes another signaling cascade with its MAP2Ks MKK1/MKK2 and the MAP3K MEKK1 (Gao et al., 2008; Qiu et al., 2008b). In unchallenged cells, MPK4 forms a nuclear complex with its substrate MKS1 and the TF WRKY33. Upon activation, MPK4 phosphorylates MKS1 to release WRKY33, which drives the expression of PHYTOALEXIN DEFICIENT3 (PAD3), encoding a biosynthetic enzyme of antimicrobial camalexin (Qiu et al., 2008a). Interestingly, WRKY33 also is a direct target of MPK3 and MPK6 in response to infection with the necrotrophic fungus Botrytis cinerea, and phosphorylation of WRKY33 by activated MPK3/6 is required for B. cinerea-induced camalexin production (Mao et al., 2011). Besides regulating camalexin accumulation in response to B. cinerea infection, MPK3 and MPK6 also are essential for B. cinerea-induced ET production by increasing both the protein stability of the ET biosynthetic enzymes ACS2 and ACS6 through direct phosphorylation and the expression levels of both genes via the phosphorylation of WRKY33, which binds to the promoters of ACS2/ACS6 (Han et al., 2010; Li et al., 2012). Multifunctionality of the MPK3/MPK6 signaling cascade can further be inferred by the ability to phosphorylate another TF, ETHYLENE RESPONSE FACTOR6 (ERF6), which leads to the stabilization of ERF6 and the induction of genes important for fungal disease resistance (Meng et al., 2013). Together, these reports show that an important route of pathogen-responsive MAPK cascades is the transmission of signals from activated receptors across the nuclear envelope to intranuclear targets, where the signal alters the expression of genes that contribute to the cell’s immune response.

The nuclear envelope acts as a selective barrier separating nuclear from cytoplasmic processes and provides eukaryotic cells with an important feature to control the specificity and timing of signaling events and gene expression. Selective macromolecular trafficking between the cytoplasm and the nucleus is mediated via nuclear pore complexes (NPCs) that are composed of nucleoporin proteins (Nups), which localize to distinct subcomplexes within the NPC (Tamura et al., 2010; Tamura and Hara-Nishimura, 2013). The translocation of proteins typically depends on the recognition of nuclear localization signals (NLSs) and nuclear export signals (NESs) on the cargo by nuclear transport receptors (NTRs). NTRs mediate nuclear import (importins) or export (exportins) by their ability to transiently interact with intrinsically disordered Phe-Gly (FG)-repeat Nups that create the selective permeability barrier of the NPC (Schmidt and Görlich, 2016).

Previously, we identified Arabidopsis Nup88/MOS7 in a forward genetic screen for components that contribute to the autoimmunity of the Toll/Interleukin1 Receptor (TIR)-type NB-LRR R gene mutant suppressor of npr1-1, constitutive1 (snc1; Cheng et al., 2009). In addition, our analyses revealed that MOS7 is indispensable for basal resistance, ETI conditioned by TIR- and coiled coil-type NB-LRR R proteins, as well as systemic acquired resistance against biotrophic and hemibiotrophic pathogens (Cheng et al., 2009). This suggests that MOS7 is a central regulatory component of plant immunity that integrates signals from diverse SA-dependent defense signaling pathways. Accordingly, we found that MOS7 promotes the nuclear retention of autoactive snc1 and the immune regulators ENHANCED DISEASE SUSCEPTIBILITY1 (EDS1) and NONEXPRESSOR OF PATHOGENESIS-RELATED GENES1 (NPR1), key components of basal and TIR-NB-LRR-mediated immunity and systemic acquired resistance, respectively (Cheng et al., 2009).

Here, we demonstrate that MOS7/Nup88 is essential for resistance to the necrotrophic fungus B. cinerea and provide evidence for the molecular function of MOS7 in defense signaling against this pathogen. Whereas infection of wild-type plants with B. cinerea activates both MPK3 and MPK6 and results in significantly enhanced MPK3 protein accumulation, mos7-1 mutants are affected in the amplitude and timing of MPK3 and MPK6 activation. In addition, the mos7-1 mutation selectively reduces nuclear and cytoplasmic protein levels of MPK3 in unchallenged and B. cinerea-infected tissues without affecting MPK3 gene expression, suggesting that enhanced nuclear export rates of MPK3 in mos7-1 affect MPK3 protein stability. Consistent with MOS7 promoting the nuclear retention of MPK3, we show by transgenic complementation that full immunity to B. cinerea requires the maintenance of a sufficient MPK3 protein abundance in the nucleus. Finally, we present a structural model of the N-terminal β-propeller domain of MOS7 and show that the four-amino acid (₁₉₀PheAspLeuSer₁₉₃ [FDLS]) deletion of mos7-1 compromises interactions with the FG-repeat Nups Nup98a and Nup98b, thus providing a mechanistic link to enhanced protein export rates and related immunity defects of the partial loss-of-function mutant mos7-1.

RESULTS

mos7-1 Mutants Are Hypersusceptible to B. cinerea Infection

MOS7 is a central regulator of diverse SA-dependent resistance pathways in Arabidopsis (Cheng et al., 2009). We aimed to investigate whether MOS7 also is required for resistance to necrotrophic pathogens, such as the fungus B. cinerea, the causal agent of gray mold diseases on a large number of plant species (Amselem et al., 2011). To test this, we inoculated fully expanded leaves of 5-week-old soil-grown mos7-1 plants with 6-µL droplets of a B. cinerea strain B05.10 (Quidde et al., 1998) spore suspension (5 × 10⁴ spores mL⁻¹) and determined the lesion diameter 3 d post inoculation (dpi). Compared with the Columbia-0 (Col-0) wild-type control, mos7-1 mutants showed significantly enhanced lesion formation that was even more extreme than on the hypersusceptible controls mpk3, mpk6, and wrky33 (Fig. 1A). Staining of infected leaves with lactophenol Trypan Blue was used to monitor fungal growth and host cell death at 2 dpi. As can be seen in Figure 1B, mos7-1 mutants show enhanced formation of spreading lesions that precede the developing mycelium at this early phase of infection, enabling faster fungal propagation on the dead plant tissue (Shlezinger et al., 2011). Together, these results show that MOS7 is essential for plant resistance to the necrotrophic fungal pathogen B. cinerea.

Figure 1.

Arabidopsis MOS7/Nup88 is essential for resistance to the necrotrophic fungus B. cinerea. A, Five-week-old Arabidopsis plants of the indicated genotypes were inoculated with 6-µL droplets of B. cinerea spore suspension (5 × 10⁴ spores mL⁻¹), and the lesion diameter was determined at 3 dpi. Bars represent means of n > 25 measurements ± se. P < 0.001 using Student’s t test for pairwise comparison of wild-type (Col-0) and mutant plants. Experiments were repeated three times with similar results. Images below the graph show representative leaves that exhibit the average lesion size. Bar = 1 cm. B, Five-week-old leaves of the indicated genotypes were stained with lactophenol Trypan Blue at 2 dpi with 7.5-µL droplets of B. cinerea spore suspension (1 × 10⁵ spores mL⁻¹) to visualize fungal hyphae and dead plant tissue. LN, Local necrosis; SL, fungus-induced spreading lesions. The experiment was repeated with similar results. Bars = 1 mm.

mos7-1 Shows Reduced Levels of Activated MPK3 and MPK6 upon B. cinerea Infection

Arabidopsis MPK3 and MPK6 are activated through phosphorylation in response to treatment with different pathogens or PAMPs and act as key signal transducers in defense against B. cinerea (Asai et al., 2002; Ren et al., 2008). Unlike transient activation of MPK3 and MPK6 upon PAMP treatment, B. cinerea infection induces a prolonged activation of both kinases (Ren et al., 2008; Han et al., 2010). As MOS7 is essential for disease resistance against B. cinerea, we sought to investigate whether the activation of MPK3 and/or MPK6 is altered in mos7-1 upon challenge with B. cinerea. To test this, we used an antibody raised against the phosphorylated pTEpY motif of the MPK activation loop for immunoblot analyses of total protein extracts derived from unchallenged and B. cinerea-inoculated Col-0 and mos7-1 leaves. As shown in Figure 2A, phosphorylated MPK6 and, to a lesser extent, MPK3 were detectable 24 h after spray inoculation of Col-0 plants with B. cinerea spore suspension. In contrast, the phosphorylated MPKs were barely detectable in mos7-1 at this time point. Compared with the 24-h time point, levels of phosphorylated MPK3 in Col-0 further increased until 48 h post inoculation (hpi) but were markedly reduced in mos7-1 compared with the corresponding wild-type control. Interestingly, a stronger signal of phosphorylated MPK6 was detected in mpk3 mutant plants at 48 hpi that was not caused by the overaccumulation of MPK6 protein in mpk3 (Fig. 2A; Supplemental Fig. S1). This suggests that increased amounts of MPK6 become activated to compensate for the absence of MPK3 (Fig. 2A). The activation of MPK3 and MPK6 was specific to inoculation with B. cinerea, as it was not observed after mock inoculation of plants with Vogel buffer (Supplemental Fig. S2). We conclude that both the amplitude and timing of MPK3 and MPK6 activation in response to B. cinerea infection are affected in mos7-1.

Figure 2.

mos7-1 mutants show reduced levels of activated MPK3 and MPK6 upon B. cinerea infection and are selectively impaired in MPK3 protein accumulation. A, Immunoblot analysis of activated MPK3 and MPK6 in total protein extracts derived from 5-week-old unchallenged (−) Col-0 and mos7-1 leaf tissues and leaf tissues harvested 24 and 48 h after spray inoculation with B. cinerea spore suspension (2 × 10⁵ spores mL⁻¹). mpk3, mpk4 (in ecotype Landsberg erecta), and mpk6 were included as controls. Equal loading was monitored by probing the membrane with a PEPC-specific antibody and Ponceau S staining. The experiment was repeated twice with similar results. B, Immunoblot analysis as in A using antibodies raised against the indicated proteins. For MPK3, two exposure times are shown. Ponceau S staining of the membrane was used to show equal loading. Similar results were obtained in three independent experiments. C, RT-PCR analyses of total RNAs extracted from unchallenged Col-0 and mos7-1 leaf tissues and leaf tissues harvested 24 and 48 h after spray inoculation with B. cinerea spore suspension (2 × 10⁵ spores mL⁻¹). Transcripts of all indicated genes were amplified by 29 cycles of PCR using equal amounts of cDNA. PCR products were analyzed by agarose gel electrophoresis and ethidium bromide staining. ACTIN1 and TUBULIN4 expression were used as controls. Similar results were obtained in three independent experiments. ntc, No template control.

The mos7-1 Mutation Selectively Affects Protein Levels of MPK3

Since mos7-1 mutants are highly susceptible to B. cinerea infection (Fig. 1) and show lower levels of activated MPK3 and MPK6 (Fig. 2A), we analyzed whether decreased amounts of phosphorylated MPK3/MPK6 are attributed to overall reduced protein levels of both kinases. Our western-blot analyses of leaf total protein extracts revealed that levels of MPK3 are reduced significantly in unchallenged as well as B. cinerea-infected mos7-1 compared with Col-0 (Fig. 2B). In response to B. cinerea infection, MPK3 total amounts increased incrementally at 24 and 48 hpi in both the wild type and mos7-1 (Fig. 2B). However, B. cinerea-triggered MPK3 accumulation was substantially lower in mos7-1 at all time points. Enhanced MPK3 protein accumulation was specific to B. cinerea inoculation, as it was not observed 24 and 48 h after mock treatment of Col-0 and mos7-1 plants with Vogel buffer (Supplemental Fig. S2). Intriguingly, protein abundance of the closely related MPK6 was not altered in unchallenged and B. cinerea-infected mos7-1 and did not obviously change in response to attack by this necrotroph (Fig. 2B). This also holds true for MPK4, which promotes JA/ET-mediated immune responses (Brodersen et al., 2006). The accumulation of SCARECROW-LIKE14 (SCL14), a transcriptional coactivator of several B. cinerea-induced genes (Fode et al., 2008), increased moderately in response to B. cinerea inoculation but was not affected in mos7-1 (Fig. 2B). Protein amounts of the defense-unrelated control phosphoenolpyruvate carboxylase (PEPC) did not show alterations in mos7-1 or in response to B. cinerea infection. These results show that MPK3 accumulates in response to B. cinerea infection and that the mos7-1 mutation selectively reduces total amounts of MPK3 in unchallenged and B. cinerea-challenged tissues. We reasoned further that reduced levels of activated MPK3 (Fig. 2A) might be attributed to overall reduced amounts of this kinase being available for phosphorylation upon B. cinerea challenge. Since total levels of MPK6 in mos7-1 are indistinguishable from those in the wild type, both in unchallenged and B. cinerea-infected tissues (Fig. 2B), this scenario does not explain the diminished amounts of activated MPK6 (Fig. 2A) and suggests an additional role of MOS7 required for the full activation of MPK6.

We next examined whether the reduction of MPK3 accumulation in mos7-1 mutant plants is caused by reduced MPK3 gene expression. Compared with unchallenged tissues, we observed a moderate increase in MPK3 mRNA abundance 24 and 48 h after B. cinerea inoculation as determined by reverse transcription (RT)-PCR (Fig. 2C). Importantly, no difference in MPK3 transcript level was observed in unchallenged and B. cinerea-infected mos7-1 compared with Col-0 (Fig. 2C). Consistent with our immunoblot analyses of MPK6 and MPK4 (Fig. 2B), the expression of both genes was not altered in mos7-1 or upon challenge with B. cinerea (Fig. 2C). Expression of the B. cinerea-responsive marker genes PAD3, ORA59, PDF1.2a, and WRKY33 was markedly induced upon infection, whereas SCL14 transcripts were only weakly induced by B. cinerea. Expression of MKK4, encoding an upstream kinase of MPK3/MPK6, showed a similar level of B. cinerea-triggered induction to MPK3. However, the expression of all tested genes, including the housekeeping genes TUBULIN4 and ACTIN1, was not obviously altered in mos7-1 as compared with Col-0 (Fig. 2C). Together, our immunoblot and RT-PCR analyses show that reduced accumulation of MPK3 in unchallenged and B. cinerea-infected mos7-1 is not caused by impaired MPK3 transcript accumulation. Since accumulation of the closely related MPK6 as well as MPK4, SCL14, and PEPC is not affected in mos7-1, these results imply a selective, posttranscriptional role of MOS7 that is essential for full MPK3 protein accumulation.

Nuclear Accumulation of MPK3 Is Severely Impaired in mos7-1

As MPK3 and MPK6 localize to the cytoplasm and the nucleus (Ahlfors et al., 2004; Yoo et al., 2008; Lumbreras et al., 2010), we sought to investigate whether mos7-1 might affect the subcellular localization and/or abundance of MPK3 and MPK6 in steady-state and B. cinerea-infected tissues. Our immunoblot analyses of fractionated subcellular compartments revealed the localization of MPK3, MPK6, as well as MPK4 (Andreasson et al., 2005) to both the cytoplasm and the nucleus of unchallenged Col-0 wild-type plants (Fig. 3). In response to B. cinerea infection, MPK3 levels increased incrementally at 24 and 48 hpi in both compartments of Col-0. Compared with wild-type plants, levels of MPK3 were depleted in both the cytoplasm and the nucleus of unchallenged and B. cinerea-infected mos7-1 mutants to levels that were barely detectable in nuclear fractions (Fig. 3). In contrast, we did not detect obvious changes in the subcellular abundance of MPK6 and MPK4 in the response of Col-0 to B. cinerea infection, nor was the subcellular accumulation of both kinases altered in mos7-1. Our finding that selectively MPK3 levels are reduced in both the cytoplasm and the nucleus of mos7-1 is reminiscent of reduced nucleocytoplasmic levels of the immune regulator EDS1 (Cheng et al., 2009). One interpretation of these results is that increased nuclear export of MPK3 in mos7-1 affects overall MPK3 protein levels, which equilibrate between the nucleoplasm and the cytoplasm.

Figure 3.

Nuclear accumulation of MPK3 is severely impaired in mos7-1. Immunoblot analyses are shown for the indicated proteins in cytoplasmic (Cyt) and nuclear (Nuc) fractions of 5-week-old unchallenged (−) Col-0 and mos7-1 leaf tissues and leaf tissues harvested 24 and 48 h after spray inoculation with B. cinerea spore suspension (2 × 10⁵ spores mL⁻¹). Anti-PEPC and anti-histone H3 were used as cytoplasmic and nuclear markers, respectively, and to monitor equal loading. Similar results were obtained in three independent experiments.

Enhanced Nuclear Export of MPK3 Compromises Resistance to B. cinerea

To establish whether reduced nuclear accumulation of MPK3 causes the resistance defects of mos7-1 against B. cinerea, we mimicked enhanced MPK3 nuclear export rates in mos7-1 through the fusion of fluorescently labeled MPK3 to a NES. For this purpose, we transformed mpk3 mutants with genomic MPK3 and expressed it under the control of the endogenous promotor as a fusion to monomeric yellow fluorescent protein (mYFP) with attached functional NES (LALKLAGLDI; ProMPK3:gMPK3-mYFP-NES) or a mutated nonfunctional nes as a control (LALKAAGADA; ProMPK3:gMPK3-mYFP-nes; García et al., 2010). For each construct, three independent homozygous transgenic lines carrying a single insertion of the transgene were selected that accumulate the tagged MPK3 fusion protein to levels comparable to endogenous MPK3 in wild-type plants, as determined by anti-MPK3 immunoblot analysis of leaf total protein extracts (Supplemental Fig. S3). Protein gel blots also were probed with GFP antibody recognizing mYFP and revealed the expression of the MPK3-mYFP-NES/nes fusion proteins but the absence of free mYFP (Supplemental Fig. S3). Moreover, B. cinerea-induced activation of transgenically expressed MPK3-mYFP-NES/nes fusion proteins was similar to the level observed for wild-type MPK3 in Col-0 plants (Supplemental Fig. S3). Next, we investigated the intracellular localization of stably expressed MPK3-mYFP-NES/nes in leaves by confocal fluorescence microscopy. As shown in Figure 4A, MPK3-mYFP-NES fluorescence was detected in the cytoplasm, whereas the expression of MPK3-mYFP-nes resulted in a nucleocytoplasmic distribution of mYFP fluorescence, consistent with a nucleocytoplasmic localization of MPK3 in fractionation experiments (Fig. 3).

Figure 4.

Enhanced nuclear export of MPK3 compromises resistance to B. cinerea. A, Confocal images showing the fluorescence distributions of MPK3-mYFP-NES and MPK3-mYFP-nes in leaves of representative 5-week-old mpk3 stable transgenic plants expressing the fusion proteins under the control of the MPK3 promoter. Bars = 50 µm. B, Five-week-old mpk3 stable transgenic plants expressing MPK3-mYFP-NES or MPK3-mYFP-nes under the control of the MPK3 promoter were inoculated with 6-µL droplets of B. cinerea spore suspension (5 × 10⁴ spores mL⁻¹), and the lesion diameter was determined 3 dpi. Bars represent means of n > 33 measurements ± se. P < 0.001 using Student’s t test for pairwise comparison of the mpk3 mutant with the other genotypes; ns, not significant. The experiment was repeated twice with similar results.

B. cinerea drop-inoculation experiments with the transgenic ProMPK3:gMPK3-mYFP-NES/nes lines revealed that nuclei-excluded MPK3-mYFP-NES did not complement the enhanced disease susceptibility of the mpk3 mutant, whereas the expression of nucleocytoplasmic MPK3-mYFP-nes fully restored wild-type-like resistance (Fig. 4B). These data strongly suggest that attaining a sufficient abundance of MPK3 in the nucleus is required to mount a robust immune response against B. cinerea and imply that the reduced nuclear accumulation of MPK3 in mos7-1 compromises resistance against B. cinerea.

Predicted Structural Changes in mos7-1 That Underlie Increased Nuclear Export Rates

Our finding that the four-amino acid (FDLS) deletion of the mos7-1 protein selectively affects nuclear retention and accumulation of certain defense-regulatory proteins is intriguing and prompted us to investigate the molecular basis for this phenotype within the predicted protein structure of MOS7. We generated a homology model of MOS7 based on the crystal structure of the N-terminal domain (NTD; amino acids 7–452 of the yeast [Saccharomyces cerevisiae] homolog of Nup88 [ScNup82]; Yoshida et al., 2011) using Phyre2 (Kelley et al., 2015). Based on this model, 409 residues of MOS7 can be superimposed on the ScNup82^NTD structure with a backbone mean square deviation of 0.71 Å, suggesting that MOS7 amino acids 55 to 555, like the N-terminal domain of ScNUP82, fold into a disc-shaped β-propeller composed of seven circularly arranged blades (Fig. 5A; a schematic representation of the MOS7 protein domain organization is given in Fig. 5C). In yeast, the N-terminal β-propeller domain of ScNup82 forms a heterotrimeric complex with the C-terminal tail (T; amino acids 1,433–1,458) of ScNup159 and the C-terminal domain (CTD; amino acids 966–1,111) of ScNup116 (Yoshida et al., 2011). The FDLS deletion in mos7-1 maps to a region in the homology model that corresponds to the end of β-strand 3D in the ScNUP82 structure (Fig. 5B). In ScNUP82, this β-strand precedes an extended loop that contributes to the binding interface for ScNup116 (Yoshida et al., 2011; Fig. 5B). A similar loop is predicted in MOS7 (amino acids Asp-191 to Ala-222) and includes the DLS residues of the FDLS sequence. MOS7 Phe-190 corresponds to ScNUP82 Phe-154 that packs against the hydrophobic core of the protein. We reasoned that deletion of the FDLS motif in the mos7-1 protein might alter the length and/or orientation of the 3D4A loop and that this might affect interaction with an ScNup116 homolog in Arabidopsis. In order to identify a potential homolog of ScNup116 from Arabidopsis and to test its interaction with MOS7 and mos7-1, respectively, we used BackPhyre (Kelley et al., 2015) with ScNUP116^CTD as query structure. This search identified the C-terminal domains of At1g10390 (Nup98a) and At1g59660 (Nup98b), two FG-repeat Nups, as putative homologs of ScNup116 (see structural alignments and Nup98a/b protein domain organization in Supplemental Fig. S4). Indeed, our analysis was supported by a recent study that identified Nup98a as a MOS7/Nup88 interactor in a yeast two-hybrid screen (Park et al., 2014).

Figure 5.

Structural basis for enhanced nuclear export in mos7-1. A, Three-dimensional homology model (top view) of the N-terminal domain of MOS7 (NTD; amino acids 55–555; blue) based on the crystal structure of ScNUP82^NTD (amino acids 7–452 [gray]; Yoshida et al., 2011; Protein Data Bank identifier 3PBP). Individual blades are delimited by gray lines and numbered. The 3D4A loops are highlighted in green, and the MOS7 FDSL motif that is deleted in the mos7-1 protein is shown in a stick representation. B, Superposition of the MOS7^NTD homology model (blue) and the ternary complex comprising ScNUP82^NTD (gray), ScNUP116^CTD (magenta), and ScNup159^T (orange; Yoshida et al., 2011). The 3D4A loops are highlighted in green, and the MOS7 FDSL motif that is absent in the mos7-1 protein is shown in a stick representation. A magnified and slightly rotated view of the MOS7 FDLS motif is shown next to the MOS7^NTD homology model. C, Schematic representation of the predicted MOS7 protein domain organization. The positions of amino acids and the FDLS motif are indicated. CC, Coiled coil; NTD, N-terminal domain.

Interaction of Wild-Type MOS7 and Mutant mos7-1 with Nup98a and Nup98b

Next, we investigated whether the C-terminal domains of both Nup98a and Nup98b can interact with MOS7 in planta and whether mutant mos7-1 is impaired in this interaction. For this purpose, we used the well-established Agrobacterium tumefaciens-mediated transient expression system in Nicotiana benthamiana and coexpressed the GFP-tagged C-terminal domains of Nup98a (amino acids 873–1,041) and Nup98b (amino acids 833–977) with 3xHA-StrepII-tagged MOS7 and mos7-1, respectively. Three days after A. tumefaciens infiltration, we conducted coimmunoprecipitation assays using GFP-Trap magnetic particles (Chromotec) and immunoblot analyses. As shown in Figure 6A, MOS7-3xHA-StrepII coimmunoprecipitated with both GFP-Nup98a^CTD and GFP-Nup98b^CTD, suggesting that MOS7 interacts with both proteins. In comparison with MOS7, the interaction of 3xHA-StrepII-tagged mos7-1 with both GFP-tagged Nup98^CTD domains was significantly weaker, as seen by the reduced amounts of mos7-1 that copurified with Nup98a and Nup98b despite comparable amounts of the fusion proteins in total protein (input) fractions (Fig. 6A). Together, these results show that MOS7 is able to interact with both Nup98a and Nup98b in planta and provide experimental evidence for our structural model that the FDLS deletion in mos7-1 affects this interaction.

Figure 6.

Interaction analysis of MOS7 and mos7-1 with Nup98a and Nup98b, infection phenotype of nup98a mutants, and Nup98a/b gene expression in wild-type and mos7-1 plants. A, The GFP-tagged C-terminal domains of Nup98a and Nup98b were transiently coexpressed with 3xHA-StrepII-tagged MOS7 or mos7-1 in N. benthamiana. At 3 d post infiltration, GFP-Nup98a^CTD and GFP-Nup98b^CTD fusion proteins were immunoprecipitated using GFP-Trap magnetic particles (IP: α-GFP), and coimmunoprecipitation of 3xHA-StrepII-tagged MOS7 or mos7-1 was detected by α-hemagglutinin (HA) immunoblots. The top two gels show total protein extracts (Input) probed with α-GFP and α-HA. Ponceau S staining of the membrane was used to show equal loading. Similar results were obtained in two independent experiments. B, Five-week-old plants of the indicated genotypes were inoculated with 6-µL droplets of B. cinerea spore suspension (5 × 10⁴ spores mL⁻¹), and the lesion diameter was determined at 3 dpi. Bars represent means of n > 38 measurements ± se. P < 0.001 using Student’s t test for pairwise comparison of wild-type (Col-0) and mutants. The experiment was repeated twice with similar results. C, RT-PCR analyses of total RNAs extracted from unchallenged (−) Col-0 and mos7-1 leaf tissues and leaves harvested 24 and 48 h after spray inoculation with B. cinerea spore suspension (2 × 10⁵ spores mL⁻¹). Transcripts of all indicated genes were amplified by 29 cycles of PCR using equal amounts of cDNA. PCR products were analyzed by agarose gel electrophoresis and ethidium bromide staining. TUBULIN4 expression was used as a control. Similar results were obtained in three independent experiments. ntc, No template control.

To investigate the requirement for Nup98 in resistance against B. cinerea, we made use of two independent nup98a mutant alleles carrying homozygous T-DNA insertions in the second (nup98a-1; SALK_103803) and third (nup98a-2; SALK_015016) exons of Nup98a (At1g10390). Three days after drop inoculation with B. cinerea spore suspension, both nup98a alleles showed enhanced lesion formation compared with the wild type (Fig. 6B), supporting the functional involvement of Nup98a in defense against B. cinerea.

To obtain additional insights into the involvement of Nup98a and Nup98b in MOS7-regulated resistance against B. cinerea, we analyzed Nup98a/b and MOS7/mos7-1 gene expression patterns by RT-PCR. Neither MOS7 nor mos7-1 transcript levels were altered in response to B. cinerea infection. Nup98a was expressed to equal amounts in Col-0 and mos7-1 rosette leaves, and its expression was not obviously altered 24 and 48 h after spray inoculation with B. cinerea spore suspension (Fig. 6C). Nup98b expression was barely detectable in unchallenged Col-0 and only weakly induced upon B. cinerea infection. Interestingly, in the mos7-1 background, Nup98b expression was higher than in wild-type plants before and after infection with B. cinerea (Fig. 6C), suggesting the existence of a feedback loop that compensates for the partial loss of MOS7 function in mos7-1 and/or the attenuated binding of mos7-1 to Nup98b upon infection with B. cinerea.

DISCUSSION

MAPK cascades are evolutionarily conserved signaling modules that allow eukaryotic organisms to translate environmental stimuli via a phosphorelay mechanism into appropriate cellular responses. In plant defense, pathogen-responsive MAPK cascades play pivotal roles by transducing signals from activated immune receptors to downstream substrates such as TFs that alter gene expression patterns (Pitzschke et al., 2009; Mao et al., 2011; Meng et al., 2013; Meng and Zhang, 2013). Arabidopsis MPK3 and MPK6 are closely related MAPKs that show a high level of functional redundancy in a multitude of biological processes and signaling pathways, including various defense responses to microbial pathogens (Liu and Zhang, 2004; Wang et al., 2007, 2008; Lampard et al., 2008; Ren et al., 2008; Han et al., 2010; Mao et al., 2011; Meng et al., 2013; Guan et al., 2014). The multifunctionality of the MPK3/MPK6 cascade in diverse response pathways implies tight regulation of the spatial and temporal kinase activities to avoid inadvertent cross-pathway activation and impose specific signaling outcomes. However, little is known in plants about the spatiotemporal control of MAPK intracellular defense signal transduction.

In this study, we identified the nucleoporin MOS7/Nup88 as an essential component of plant resistance to the gray mold fungus B. cinerea and we provide, to our knowledge, the first molecular evidence for the spatial regulation of MAPK signaling at the level of the NPC in plants. Initially, we observed that enhanced susceptibility of partial loss-of-function mos7-1 plants to B. cinerea (Fig. 1) coincides with a delayed timing and reduced magnitude of MPK3 and MPK6 activation/phosphorylation (Fig. 2A). Indeed, it has been reported that the physiological outcomes of MPK3/MPK6 activation depend on the kinetics and magnitude of their activation. For instance, induction of PAMP-triggered immunity results in a transient activation of MPK3/MPK6 that peaks within a few minutes, whereas induction of ETI or infection with B. cinerea leads to a strong and sustained activation of both kinases (Asai et al., 2002; Wan et al., 2004; Ren et al., 2008; Ranf et al., 2011; Tsuda et al., 2013; Fig. 2A). Therefore, is has been suggested that MPK3/MPK6 activation in response to B. cinerea infection needs to reach a certain threshold in amplitude and duration to mount a robust immune response (Ren et al., 2008; Meng and Zhang, 2013). This conclusion is consistent with our observation that amounts of activated MPK3 and MPK6 are reduced in the hypersusceptible mos7-1 mutant (Fig. 2A). Regarding MPK3 function, our analyses revealed that total amounts of MPK3 protein increased considerably upon challenge with B. cinerea. We observed a specific effect on MPK3 levels, which were reduced in both unchallenged and B. cinerea-infected mos7-1 compared with Col-0 (Fig. 2B). In contrast, the protein accumulation of MPK6 and MPK4 remained unaltered in response to B. cinerea infection and in the mos7-1 mutant (Fig. 2B). Since MPK3 and MPK6 are closely related proteins and show a high level of functional overlap, this molecular phenotype was unexpected, but it supports the previous notion that MPK3 and MPK6 play both redundant and distinct roles in resistance responses to B. cinerea (Ren et al., 2008; Galletti et al., 2011). Therefore, MOS7 may contribute to the differential regulation of MPK3 and MPK6 signaling functions through the modulation of their protein abundance. The depletion of MPK3 protein in mos7-1 (Fig. 2B) provides a likely explanation for the reduced levels of activated/phosphorylated MPK3 (Fig. 2A). Nevertheless, we cannot exclude the possibility that the accumulation and/or activity of proteins controlling MPK3/MPK6 phosphorylation/dephosphorylation are misregulated in mos7-1, since B. cinerea-induced activation of MPK6 is impaired in mos7-1 without its overall protein accumulation being affected (Fig. 2).

In the mpk3 mutant lacking functional MPK3 protein, we detected a stronger activation of MPK6 at 48 hpi compared with wild-type levels (Fig. 2A), whereas this phenotype was not observed in mos7-1, which has reduced MPK3 protein. This indicates that the loss of MPK3 influences the activity of MPK6 but that low levels of MPK3 in mos7-1 may be sufficient to avoid the hyperactivation of MPK6. Such compensatory activation of MPK6 in mpk3 could be one explanation for why mos7-1 is more susceptible to B. cinerea infection than mpk3 despite mos7-1 having residual levels of MPK3.

Importantly, reduced MPK3 protein accumulation in unchallenged and B. cinerea-challenged mos7-1 was not attributed to reduced MPK3 gene expression (Fig. 2C). This shows that MOS7 contributes to cellular MPK3 protein accumulation through a posttranscriptional mechanism. Our subsequent analysis of the subcellular protein levels of MPK3 revealed that its nucleocytoplasmic distribution was not obviously affected in mos7-1. Instead, MPK3 levels were reduced in both the cytoplasm and the nucleus of unchallenged and B. cinerea-inoculated mos7-1, resulting in very low amounts of MPK3 being detected in the nuclear compartment (Fig. 3). This posttranscriptional depletion of MPK3 in both the cytoplasm and nucleus of mos7-1 is reminiscent of the reduced nucleocytoplasmic levels of the immune regulator EDS1 (Cheng et al., 2009). In contrast, mos7-1 reduced nuclear levels of the autoactive immune receptor snc1 without affecting its total amounts (Cheng et al., 2009). These data, together with our previous finding that mos7-1 enhances NES/Exportin1 (XPO1)-mediated nuclear protein export (Cheng et al., 2009) and reports that defects in animal Nup88 increase NES/CRM1-mediated nuclear export of activated nuclear factor-κB TFs (Roth et al., 2003; Xylourgidis et al., 2006; Takahashi et al., 2008), suggest that the enhanced nuclear export of MPK3 in mos7-1 affects MPK3 protein stability, for example by enhancing its cytoplasmic degradation. This would lower the cytoplasmic amounts of MPK3 available for nuclear import and result in the observed proportional depletion in the cytoplasm and the nucleus (Fig. 3). Unlike EDS1 (García et al., 2010) and snc1 (Cheng et al., 2009), we did not detect a predicted NES in the protein sequence of MPK3 that would mark it as an XPO1 cargo client (Haasen et al., 1999; la Cour et al., 2004), indicating that MPK3 may possess a novel or unrecognized NES. Accordingly, it has been shown that the nuclear export of other nucleocytoplasmic shuttling proteins without discernible NES (a good example being SCL14; Fode et al., 2008) is mediated by XPO1. Alternatively, MPK3 might be coexported from the nucleus with an NES-containing interaction partner, as reported for the animal nucleocytoplasmic shuttling MAPKs ERK1/2 (Adachi et al., 2000).

It remains to be determined why increased nuclear export in mos7-1 has an impact on MPK3 and EDS1 but not on snc1 protein accumulation. Examples of animal nucleocytoplasmic shuttling proteins show that their enhanced nuclear export decreases protein stability and allows the cell to regulate protein levels via proteasomal degradation in the cytoplasm (Knauer et al., 2005; Shin et al., 2008). Vice versa, delayed nuclear export can be associated with increased protein stability (Connor et al., 2003). The fact that we can express MPK3-mYFP-NES transgenically close to protein levels of MPK3 in wild-type plants (Supplemental Fig. S3) raises the possibility that the protein stability of MPK3 is regulated independently of its nuclear export rates and may involve additional cellular factors that are affected in mos7-1. Nevertheless, by increasing the rate of MPK3 export from nuclei through fusion to an NES, we demonstrated that the nuclear pool of MPK3 is essential for full resistance to B. cinerea (Fig. 4). This is consistent with MPK3 functioning inside the nucleus, where it phosphorylates TFs to regulate the expression of B. cinerea-responsive defense genes (Ren et al., 2008; Mao et al., 2011; Li et al., 2012; Meng et al., 2013), and corroborates the notion that mos7-1 plants are not able to attain a sufficient concentration of MPK3 inside the nucleus to mount a full defense response to B. cinerea. Because the susceptibility of mos7-1 to infection with B. cinerea was consistently more extreme than the susceptibility of mpk3 (Fig. 1A), we reasoned that MOS7 might be required for the nuclear accumulation of additional defense-regulatory proteins implicated in resistance against B. cinerea.

The selective role of MOS7 in modulating nuclear concentrations of certain defense-regulatory proteins is intriguing yet consistent with the immunoregulatory function of its animal homolog Nup88 that promotes nuclear factor-κB nuclear accumulation (Roth et al., 2003; Xylourgidis et al., 2006; Takahashi et al., 2008). Cheng et al. (2009) reported that a GFP fusion of the mutant mos7-1 protein shows the same nuclear envelope localization as the GFP-tagged wild-type protein, so until now the changes in mos7-1 that underlie increased nuclear export rates were unknown. Based on the crystal structure of its yeast homolog ScNup82, we now present a homology model of the N-terminal β-propeller domain of MOS7 and propose that the mos7-1 FDLS deletion maps to an extended loop of MOS7 that contributes to the binding interface for the FG-repeat nucleoporins Nup98a and Nup98b (Figs. 5 and 6A; Yoshida et al., 2011). Consistently, when coexpressed in N. benthamiana, considerably reduced amounts of mos7-1 coimmunoprecipitated with Nup98a^CTD and Nup98b^CTD compared with wild-type MOS7 (Fig. 6A). Deletion of the entire 3D4A loop from the ScNup82 β-propeller resulted in a 5-fold reduced binding affinity to ScNup116^CTD (the putative homolog of Nup98a/b), as determined in vitro by isothermal titration calorimetry (Dissociation konstant (K_d) of 50 ± 30 nm for ScNup82^NTD versus 250 ± 40 nm for ScNup82^NTD,Δ3D4A; Yoshida et al., 2011). In this regard, it was unexpected that deletion of the FDLS motif in mos7-1 causes substantial differences in complex formation with Nup98a/Nup98b (Fig. 6A). One possible explanation is that, compared with the yeast ScNup82^NTD/Nup116^CTD complex, the MOS7 3D4A loop might make a larger contribution to the overall binding interface for Nup98a/b. Alternatively, relatively small changes in the dissociation konstant (K_d) might be sufficient to affect MOS7-Nup98a/b complex formation in N. benthamiana, where endogenous NbMOS7 competes for binding to Nup98a/b, as reported recently for cargo/importin-α complexes (Wirthmueller et al., 2015).

FG-repeat Nups are essential for NPC function. They create the selective NPC permeability barrier and provide binding sites for NTRs that traverse the central channel during facilitated transport (Schmidt and Görlich, 2016). As the four-amino acid deletion of mos7-1 weakens interactions with the FG-repeat Nups, Nup98a and Nup98b (Fig. 6A), and decreases the nuclear accumulation of immunoregulatory proteins (Fig. 3; Cheng et al., 2009), we consider two possible scenarios for the defect in mos7-1 plants. In one model, reduced binding of Nup98a/b to mos7-1 relaxes the NPC permeability barrier, thus rendering NPCs in mos7-1 more permissive to passive diffusion. In the other model, reduced binding of Nup98a/b to mos7-1 decreases FG-repeat binding sites for distinct XPO1/cargo-protein export complexes. This would affect the retention time of export complexes in the NPC channel and allow XPO1/cargo complexes to vacate the central channel more easily. We favor the latter model, because the defect of mos7-1 mutants is surprisingly specific and most acutely influences the nuclear accumulation of particular nucleocytoplasmic defense regulators (Fig. 3; Cheng et al., 2009). Therefore, the most straightforward interpretation of our data is that MOS7 acts as a regulatory scaffold that tethers Nup98a/b to the NPC to attenuate the nuclear export of certain defense-regulatory proteins. Enhanced expression of Nup98b in mos7-1 plants (Fig. 6C), therefore, may represent a compensatory response of mos7-1 to replenish the level of its binding partner in an effort to delay the passage of respective export complexes through the NPC.

In metazoans, Nup98 is mobile and has multiple reported functions, including mRNA export (Powers et al., 1997), regulation of gene expression (Kalverda et al., 2010), maintenance of the NPC permeability barrier (Hülsmann et al., 2012), and as a facilitator of NTR-mediated transport, including distinct protein import pathways (Wu et al., 2001; Oka et al., 2010; Schmidt and Görlich, 2015). Our analysis now provides strong evidence that Nup98, when tethered to the NPC by MOS7, attenuates the nuclear export of certain defense regulators to promote their nuclear accumulation and mount an efficient immune response. Therefore, it is conceivable that MOS7- and Nup98-dependent nuclear retention depends on the conformation of a given XPO1/cargo complex that determines its binding sites within, and thus its export route through, the NPC. Support for this idea comes from studies showing that NTR conformation can be influenced by its specific cargo and that specific transport pathways through the NPC can be controlled by particular Nups (Wu et al., 2001; Yang et al., 2004; Kubitscheck et al., 2005; Wohlwend et al., 2007; Ma et al., 2012). As MOS7 is required for both SA-mediated defense responses to (hemi)biotrophic pathogens (Cheng et al., 2009) and JA/ET-mediated resistance to the necrotroph B. cinerea (Fig. 1), we speculate that MOS7 may impinge on SA-JA/ET signal integration and cross talk, thereby modulating plant defense responses to different types of pathogenic microbes at the level of the NPC.

MATERIALS AND METHODS

Plant Material and Growth Conditions

Arabidopsis (Arabidopsis thaliana) plants were grown in soil at 22°C and a 10-h photoperiod in environmentally controlled chambers. The mutants mos7-1 (Cheng et al., 2009), wrky33 (GABI_324B11; Birkenbihl et al., 2012), mpk3-DG (Li et al., 2002; Miles et al., 2005), and mpk4 (Petersen et al., 2000) have been described previously. The mpk6 (SALK_127507) and nup98a (SALK_103803 and SALK_015016) T-DNA insertion mutants were obtained from the Nottingham Arabidopsis Stock Centre. All mutants are in the Col-0 background except for mpk4 (Landsberg erecta). Stable transgenic plants expressing MPK3-mYFP-NES or MPK3-mYFP-nes under the control of a 1-kb endogenous MPK3 promoter sequence were generated by transforming mpk3-DG mutants with Agrobacterium tumefaciens strain GV3101 (pMP90RK) carrying the fusion gene constructs in the binary vector pXCG mYFP-NES/nes (García et al., 2010; for vector construction details, see below). Multiple independent homozygous lines with a single insertion of the transgene were selected and used for experiments.

Botrytis cinerea Inoculation

B. cinerea B05.10 was cultivated on potato dextrose agar plates at 22°C for 10 d. Spores were collected in 15% glycerol and filtered through one layer of Miracloth, and 1-mL aliquots were stored at −80°C at a concentration of 2 × 10⁶ spores mL⁻¹. For droplet inoculations, spores were diluted in one-quarter-concentrated potato dextrose broth (P6685; Sigma-Aldrich) to 5 × 10⁴ spores mL⁻¹ and kept at room temperature for 4 h to allow spore germination. Afterward, 6-µL droplets were placed onto fully expanded leaves of 5-week-old soil-grown plants, and the lesion size was measured 3 dpi using the digital caliper MarCal 16ER (Mahr). For spray inoculations, spores were diluted in Vogel buffer (15 g L⁻¹ Suc, 2.5 g L⁻¹ trisodium citrate·2H₂O, 5 g L⁻¹ K₂HPO₄, 0.2 g L⁻¹ MgSO₄·7H₂O, 0.1 g L⁻¹ CaCl₂·2H₂O, and 2 g L⁻¹ NH₄NO₃, pH 6) to 2 × 10⁵ spores mL⁻¹ and kept at room temperature for 4 h before inoculation. For mock treatment, plants were sprayed with Vogel buffer only. Inoculated plants were kept under a propagator lid to create a high-humidity atmosphere and incubated in a growth chamber at 22°C and a 10-h light period.

Transient Expression in Nicotiana benthamiana and Coimmunoprecipitation

A. tumefaciens GV3101 (pMP90RK) bacteria carrying the desired expression constructs and A. tumefaciens strain GV3101 expressing the silencing suppressor 19K were grown overnight in DYT medium (16 g L⁻¹ peptone, 10 g L⁻¹ yeast extract, and 10 g L⁻¹ NaCl, pH 7.2) containing the appropriate antibiotics. After centrifugation, cells were resuspended in infiltration medium (10 mm MgCl₂ and 150 μg mL⁻¹ acetosyringone) at an optical density at 600 nm = 0.9 and kept at room temperature for 2 to 3 h. For coexpression, strains carrying the expression constructs were mixed with A. tumefaciens GV3101^[19K] at a ratio of 1:1:1 and syringe infiltrated into leaves of 3- to 4-week-old N. benthamiana plants. Three days after infiltration, leaf material was harvested, frozen in N₂, and stored at −80°C. Protein extracts for coimmunoprecipitations were prepared by grinding 1 g of leaf material in liquid N₂ to a fine powder followed by resuspension in 1.5 mL of extraction buffer (50 mm Tris-HCl, pH 8, 150 mm NaCl, 1 mm EDTA, 5 mm dithiothreitol [DTT], 0.2% Nonidet P-40, and 1× protease inhibitor cocktail [Sigma]). Two milliliters of the extracts was transferred to 2-mL tubes and centrifuged at 17,000g and 4°C for 30 min. The supernatants were transferred to new tubes, and the centrifugation was repeated for 15 min. Afterward, 60 µL of each supernatant was boiled for 8 min with 20 µL of 4× SDS-PAGE sample buffer (0.25 m Tris, pH 6.8, 8% SDS, 40% glycerol, 0.04% Bromphenol Blue, and 0.4 m DTT) and kept as input fractions. For immunoprecipitation, 1.6 mL of each supernatant was combined with 10 µL of GFP-Trap_M magnetic beads (Chromotec) that were washed with 500 µL of extraction buffer and incubated on a rotating wheel at 4°C for 2 h. Beads were collected on a magnetic rack, washed four times with 1 mL of wash buffer (extraction buffer without protease inhibitor cocktail), and boiled in 50 µL of 4× SDS-PAGE sample buffer to elute proteins from the beads.

Protein Analyses and Nucleocytoplasmic Fractionation

Total and nucleocytoplasmic protein extracts for immunoblot analysis were prepared from 5-week-old leaf material of unchallenged, mock-treated, or B. cinerea spray-inoculated plants. For total protein extracts, 20 leaf discs (diameter = 6 mm) were collected in 2-mL tubes, frozen in liquid N₂, and homogenized eight times for 45 s to a fine powder using a TissueLyser LT (Qiagen) and 3-mm steel beads. A total of 350 µL of 2× SDS-PAGE sample buffer (0.125 m Tris, pH 6.8, 4% SDS, 20% glycerol, 0.02% Bromphenol Blue, and 0.2 m DTT) was added, and samples were boiled for 8 min and centrifuged at 17,000g and 4°C for 30 min. Supernatants were transferred to clean tubes and stored at −20°C. Nucleocytoplasmic protein extracts were prepared by grinding 2 g of leaf material first in liquid N₂ followed by grinding in 2 mL of extraction buffer (10 mm Tris, pH 7.5, 10 mm NaCl, 10 mm MgCl₂, 10% [v/v] glycerol, 10 mm β-mercaptoethanol, and 1× protease inhibitor cocktail [Sigma]). The homogenate was spun for 3 min through a 95-µm nylon mesh at 1,500g and 4°C. The filtrate was transferred to 2-mL tubes and centrifuged at 3,000g and 4°C for 12 min to pellet the nuclei. The supernatant was transferred to new tubes and recentrifuged at 13,000g and 4°C for 15 min. An aliquot of this supernatant was mixed with 2× SDS-PAGE sample buffer, boiled for 8 min, and saved as the cytoplasmic fraction. Each nuclear pellet of the first centrifugation was resuspended in 1 mL of nuclei washing buffer (10 mm Tris, pH 7.5, 10 mm NaCl, 10 mm MgCl₂, 1 m hexylene glycol, 0.5% [v/v] Triton X-100, and 10 mm β-mercaptoethanol) and centrifuged at 1,500g for 10 min at 4°C. The supernatant was discarded, washing and centrifugation of nuclear pellets were repeated seven times, and finally each nuclear pellet was resuspended in 50 µL of extraction buffer. Both nuclear fractions of each genotype were pooled, boiled with 150 µL of 2× SDS-PAGE sample buffer, and centrifuged at 17,000g and 4°C for 2 min, and the supernatants were saved as nuclear fractions. Proteins were separated by SDS-PAGE on 10% gels and transferred onto nitrocellulose membranes (Amersham Protran; 0.45 µm). Antibodies used for immunoblot analyses were α-MPK3 (M8318 [Sigma]; predicted molecular mass, 43 kD), α-MPK4 (A6979 [Sigma]; predicted molecular mass, 43 kD), α-MPK6 (A7104 [Sigma]; predicted molecular mass, 45 kD), α-P-p44/42 (9101 [Cell Signaling Technology]), α-PEPC (100-4163 [Rockland]; predicted molecular mass, 110 + 117 kD), α-histone H3 (ab1791 [Abcam]; predicted molecular mass, 15 kD), α-SCL14 (Fode et al., 2008; predicted molecular mass, 86 kD), α-GFP (11814460001 [Roche]), and α-HA (H9658 [Sigma]). The secondary antibodies goat anti-rabbit horseradish peroxidase (HRP; #32460), goat anti-rabbit poly-HRP (#32260), goat anti-mouse HRP (#32430), and goat anti-mouse poly-HRP (#32230) were purchased from Thermo Scientific, and HRP activity was detected using the SuperSignal West Pico (#34080) and Femto (#34095) chemiluminescence substrates (Thermo Scientific).

Gene Expression Analyses

Total RNA was extracted from unchallenged and B. cinerea spray-inoculated leaves of 5-week-old soil-grown plants using QIAzol (Qiagen). A total of 1.5 µg of DNaseI-treated RNA was reverse transcribed using RevertAid H Minus reverse transcriptase (Fermentas) and 0.5 µg of oligo(dT)₁₈V primer at 42°C in a 20-µL reaction volume. Two-microliter aliquots of 1:5-diluted cDNAs were used for semiquantitative PCR. Primers used for RT-PCR analysis are listed in Supplemental Table S1.

Construction of Plasmids

Binary vector constructs for the expression of MPK3-mYFP-NES and MPK3-mYFP-nes under the control of the endogenous MPK3 promoter were generated by PCR amplification of a Col-0 genomic fragment containing the MPK3 coding region without the stop codon and 1 kb upstream of the start codon, using the primers PromMPK3.D-TOPO.F (5′-CACCGACTTAACGTCGCCATGCC-3′) and gMPK3.R.ΔStop (5′-ACCGTATGTTGGATTGAGTGC-3′). The fragment was cloned into pENTR/D-TOPO (Invitrogen) and sequenced, and an LR reaction was made with the binary destination vectors pXCG mYFP-NES and pXCG mYFP-nes (García et al., 2010) to generate the expression vector pXCG ProMPK3:gMPK3-mYFP-NES/nes. For the generation of 35S promoter-driven MOS7-3xHA-StrepII and mos7-1-3xHA-StrepII constructs, the full-length genomic sequences of MOS7 and mos7-1 lacking a stop codon were PCR amplified from Col-0 and mos7-1 genomic DNA, respectively, using the primers gMOS7.D-TOPO.F (5′-CACCATGAAATTTAACTTTAACGAGAC-3′) and gMOS7.R.ΔStop (5′-CATGAAACTGCTTTCTTGCG-3′). PCR products were cloned into pENTR/D-TOPO and sequenced, and LR reactions were made between the pENTR/D-TOPO clones and the destination vector pXCSG 3xHA-StrepII (Witte et al., 2004) to generate the expression vectors pXCSG gMOS7-3xHA-StrepII and pXCSG gmos7-1-3xHA-StrepII. The Nup98a sequence encoding amino acids 873 to 1,041 was amplified using primers Nup98aCfw (5′-CACCGGTGCAGATATCGAGGCTC-3′) and Nup98aCrv (5′-CTAAACTCCATCTTCTTCATCTTCG-3′) from Col-0 cDNA. The Nup98b cDNA sequence encoding amino acids 832 to 997 was amplified using primers Nup98bCfw (5′-CACCGGAGCAGATATTGAGTC-3′) and Nup98bCrv (5′-CTACACATCATATTCATCTCC-3′). Both PCR fragments were cloned into pENTR-D-TOPO and sequenced. To generate eGFP-fused Nup98a/b constructs, the Nup98a and Nup98b pENTR clones were ligated into pK7WGF2 (Karimi et al., 2002) using LR II Clonase (Invitrogen).

Confocal Fluorescence Microscopy

Leaf discs of stable transgenic Arabidopsis mpk3 plants expressing MPK3-mYFP-NES/nes under the control of the native MPK3 promoter were mounted in water, and confocal images were obtained at room temperature with a Leica SP5-DM6000 using an HCX PL APO 40.0x0.85 DRY objective (Leica) and an excitation wavelength of 514 nm (525–600 nm emission). Images were acquired with a Leica PMT2 detector and Leica LAS AF software (version 2.6.7266.0) and processed using Adobe Photoshop CS5.

Statistical Analysis

The two-tailed Student’s t test was used for all statistical analyses, and data are presented as means ± se. In the figures, asterisks indicate P < 0.001 and ns stands for not significant.

Accession Numbers

Sequence data from this article can be found in the Arabidopsis Genome Initiative or GenBank/EMBL databases under the accession numbers listed in Supplemental Table S1.

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Immunoblot analyses of MPK3 and MPK6 protein abundance in mpk6 and mpk3 mutant plants 48 h after spray inoculation with B. cinerea spore suspension.

• Supplemental Figure S2. Immunoblot analyses of MPK3 and activated MPK3/MPK6 in Vogel buffer (mock)-treated Col-0 and mos7-1.

• Supplemental Figure S3. Immunoblot analyses of stable transgenic mpk3 lines expressing genomic MPK3 under the control of the endogenous promotor as a fusion to mYFP with attached functional NES or a mutated nonfunctional nes.

• Supplemental Figure S4. Phyre2 alignments of the ScNUP116^CTD protein structure with the predicted protein structures of the C-terminal domains of Nup98a and Nup98b.

• Supplemental Table S1. List of gene accession numbers and oligonucleotides used for RT-PCR analysis.

Glossary

PAMP

pathogen-associated molecular pattern

ETI

effector-triggered immunity

NB-LRR

nucleotide-binding domain and leucine-rich repeat

SA

salicylic acid

JA

jasmonic acid

ET

ethylene

TF

transcription factor

NPC

nuclear pore complex

NES

nuclear export signal

NTR

nuclear transport receptor

FG

Phe-Gly

dpi

days post inoculation

Col-0

Columbia-0

hpi

hours post inoculation

RT

reverse transcription

DTT

dithiothreitol

HRP

horseradish peroxidase

HA

hemagglutinin