In eukaryotic cells, transduction of external stimuli into the nucleus to induce transcription and export of mRNAs for translation in the cytoplasm is mediated by nuclear pore complexes (NPCs) composed of nucleoporin proteins (Nups). We previously reported that Arabidopsis MOS3, encoding the homolog of vertebrate Nup96, is required for plant immunity and constitutive resistance mediated by the de-regulated Toll interleukin 1 receptor/nucleotide-binding/leucine-rich repeat (TNL)-type R gene snc1. In vertebrates, Nup96 is a component of the conserved Nup107-160 nuclear pore sub-complex, and implicated in immunity-related mRNA export. Here, we used a reverse genetics approach to examine the requirement for additional subunits of the predicted Arabidopsis Nup107-160 complex in plant immunity. We show that, among eight putative complex members, beside MOS3, only plants with defects in Nup160 or Seh1 are impaired in basal resistance. Constitutive resistance in the snc1 mutant and immunity mediated by TNL-type R genes also depend on functional Nup160 and have a partial requirement for Seh1. Conversely, resistance conferred by coiled coil-type immune receptors operates largely independently of both genes, demonstrating specific contributions to plant defense signaling. Our functional analysis further revealed that defects in nup160 and seh1 result in nuclear accumulation of poly(A) mRNA, and, in the case of nup160, considerable depletion of EDS1, a key positive regulator of basal and TNL-triggered resistance. These findings suggest that Nup160 is required for nuclear mRNA export and full expression of EDS1-conditioned resistance pathways in Arabidopsis.
In plants, a major barrier to pathogen infection is conferred by innate immune responses of individual cells. A first line of basal defense is mediated by cell-surface pattern recognition receptors that sense generic pathogen-associated molecular patterns (PAMPs); this is referred to as PAMP-triggered immunity (Jones and Dangl, 2006; Zipfel, 2009). Although PAMP-triggered immunity is normally sufficient to prevent non-adapted microbes from colonizing plant tissues, host-adapted pathogens have evolved effectors that are often secreted inside host cells during infection to suppress PAMP-triggered immunity, thereby allowing host invasion. Plant cells deploy a second, largely intracellular, layer of immunity to host-adapted pathogens that is generally conferred by nucleotide-binding domain/leucine-rich repeat-containing immune receptors known as resistance (R) proteins. R proteins recognize and respond to isolate-specific microbial effectors in a strong and robust defense response termed effector-triggered immunity. Effector-triggered immunity typically culminates in programmed cell death at attempted infection sites as part of the hypersensitive response (HR) to restrict pathogen growth (Greenberg and Yao, 2004; Jones and Dangl, 2006). Pattern recognition receptor and R protein activation triggers a major transcriptional reprogramming of affected host cells (Tao et al., 2003; Bartsch et al., 2006) that depends on coordinated relay of information by intracellular signaling systems into the nucleus and nuclear export of defense-related transcripts towards the cytoplasmic protein synthesis machinery. Thus, communication between the cytoplasm and the nucleus is required for both PAMP-triggered immunity and effector-triggered immunity.
The nucleo-cytoplasmic exchange of proteins and RNAs is mediated through nuclear pore complexes (NPCs), numerous perforations in the nuclear envelope (NE) that separates the cytoplasm and the nucleoplasm of eukaryotic cells. NPCs are supramolecular assemblies comprising multiple copies of approximately 30 constituent nucleoporin proteins (Nups). Nups are modularly assembled in distinct sub-complexes and arranged radially around a central channel that serves as the sole conduit for selective bi-directional exchange of molecular cargoes. Translocation of proteins typically depends on the recognition of nuclear localization/export signal sequence motifs on the cargo by nuclear transport receptors of the karyopherin (Kap) family that have the capacity to transiently interact with the NPC to facilitate nuclear import (importins) or export (exportins) (Tran and Wente, 2006; Terry et al., 2007). General export of mRNAs utilizes non-Kap export receptors and requires numerous additional RNA-binding proteins and export factors that assemble with mRNAs into export-competent messenger ribonucleoprotein particles (Cole and Scarcelli, 2006; Köhler and Hurt, 2007; Chinnusamy et al., 2008).
In plants, little is known about the composition of the NPC and nuclear import and export pathways that mediate spatial communication between the cytoplasm and nucleoplasm. The contribution of the nucleo-cytoplasmic trafficking machinery to plant immunity was first revealed in an Arabidopsis genetic screen that aimed to identify components of auto-immune responses and related growth inhibition caused by an constitutively active variant of the Toll interleukin 1 receptor/nucleotide-binding/leucine-rich repeat (TNL)-type R gene snc1 (suppressor of npr1-1, constitutive 1) (Zhang et al., 2003). Mutations in the nucleoporin genes MOS3 (MODIFIER OF SNC1, 3) and MOS7, which encode homologs of vertebrate/Drosophila Nup96 and Nup88, respectively, and in the importin α3 homolog MOS6 were identified as suppressors of snc1 (Palma et al., 2005; Zhang and Li, 2005; Cheng et al., 2009). The three MOS proteins participate in different nucleo-cytoplasmic trafficking pathways: MOS6 and MOS7 are involved in protein import and export, respectively, and MOS3/Nup96/SAR3 is implicated in nuclear mRNA export (Palma et al., 2005; Zhang and Li, 2005; Parry et al., 2006; Cheng et al., 2009). Recently, the predicted RNA-binding protein MOS11 was identified as another essential component of snc1 auto-immunity, probably functioning in the same mRNA export pathway as MOS3 (Nup96) at an early stage before the mRNAs exit the nucleus through nuclear pores (Germain et al., 2010).
Vertebrate Nup96 and its yeast homolog, Nup145C, are constituents of the Nup107-160 nuclear pore sub-complex (called the Nup84 complex in Saccharomyces cerevisiae), that is located symmetrically on both the nuclear and cytoplasmic sides of the nuclear pore and the largest subunit of the NPC (Alber et al., 2007). Of the 30 bona fide Nups so far identified in vertebrates, almost one third can be assigned to the Nup107-160 complex, which consists of Nup37, Nup43, Nup85, Nup96, Nup107, Nup133, Nup160, Sec13 and Seh1. The Nup107-160/Nup84 complex is multi-functional. It mediates mRNA export in interphase and plays roles in kinetochore functions and post-mitotic NPC formation in the NE (Fabre et al., 1994; Vasu et al., 2001; Harel et al., 2003; Walther et al., 2003; Zuccolo et al., 2007). Notably, one of its members in mice, Nup96, is required selectively for innate and adaptive immunity (Faria et al., 2006).
The involvement of vertebrate and Arabidopsis Nup96 in immune responses (Zhang and Li, 2005; Faria et al., 2006) and the requirement for three additional members of this sub-complex, Nup85, Nup133 and Seh1, for plant responses to symbiotic microbes (Kanamori et al., 2006; Saito et al., 2007; Groth et al., 2010) directed our attention towards identifying additional putative Arabidopsis Nup107-160 complex members and understanding their possible function in response to microbial pathogen challenge. To address this, we used a reverse genetics approach, and analyzed available T-DNA insertion mutants of eight putative Nup107-160 complex members predicted in the Arabidopsis genome for defects in disease resistance. We found that two members which have not previously been implicated in regulating plant immunity, Nup160 and the WD40-repeat nucleoporin Seh1, are both required for basal defense to virulent Pseudomonas syringae and contribute different activity to auto-immunity in snc1 and TNL-conditioned resistance in response to avirulent bacterial and oomycete pathogens. In contrast, coiled coil/nucleotide-binding/leucine-rich repeat (CNL)-mediated immunity operates largely independently of Nup160 or Seh1. We further establish that both the nup160 and seh1 mutants accumulate poly(A) mRNA inside the nucleus. In addition, protein levels of ENHANCED DISEASE SUSCEPTIBILITY1 (EDS1), a key positive regulator of basal and TNL protein-mediated resistance, are strongly depleted in the nup160 mutant. Thus, Nup160 is an important component of EDS1-dependent resistance in Arabidopsis, reinforcing an important role for the plant NPC in innate immunity.
Identification of putative Arabidopsis Nup107-160 complex members
We previously identified Arabidopsis MOS3/Nup96 from a mutant screen for genetic suppressors of snc1 (Zhang and Li, 2005). As the MOS3 homologs in vertebrates and yeast, Nup96 and Nup145C, respectively, are integral members of the vertebrate Nup107-160/yeast Nup84 complex (hereafter called the Nup107-160 complex), we wished to determine whether additional subunits of this largest nuclear pore sub-complex also exist in plants. We performed blast database searches to identify additional putative complex members in the Arabidopsis thaliana Col-0 reference genome. The vertebrate and yeast complexes comprise nine and seven members, respectively (Table S1), and our analyses revealed that eight of the nine vertebrate members have putative homologs in Arabidopsis. These include Nup160/SAR1 (At1g33410), Nup133 (At2g05120), Nup107 (At3g14120), Nup96/MOS3/SAR3 (At1g80680), Nup85 (At4g32910), Nup43 (At4g30840), Seh1 (At1g64350), and two genes similar to Sec13 (At3g01340 and At2g30050) (Figure 1 and Table S1). We conclude that the Nup107-160 complex, which appears to be evolutionary conserved between distantly related eukaryotes such as yeast and mammals (Belgareh et al., 2001; Walther et al., 2003; Bai et al., 2004), also exists in a similar form in plants, as suggested previously (Parry et al., 2006; Tamura et al., 2010).
Nup160 and Seh1 are required for basal defense
As mutations in the Nup96 homolog MOS3 impair basal defense against the virulent bacterial pathogen Pseudomonas syringae (Zhang and Li, 2005), we tested the putative Nup107-160 complex members in a targeted reverse genetics approach for their involvement in basal resistance to this pathogen. T-DNA insertion mutants of the putative complex members were obtained from the Arabidopsis Biological Resource Center (Figure 1) (Alonso et al., 2003), and homozygous lines were isolated by PCR-based genotyping. For each gene except Nup107 (At3g14120) and Sec13A (At2g30050) (see below), at least one line with a T-DNA inserted within exonic sequence was obtained (Figure 1), and disruption of functional transcripts was confirmed by RT-PCR using cDNA-specific primers flanking the insertions (data not shown). For Sec13A (At2g30050) no T-DNA insertion mutant was available. Our RT-PCR analyses of the two intronic nup107 insertion lines revealed disruption of full-length transcripts in nup107-1, suggesting loss of Nup107 function in this mutant allele (data not shown). We named the nup160 mutant alleles nup160-3 (SAIL_877_B01) and nup160-4 (SALK_126801; sar1-4) because nup160-1 and nup160-2 have previously been assigned (Dong et al., 2006). An additional nup160 mutant allele, sar1-1 (Parry et al., 2006), was kindly provided by Mark Estelle (Section of Cell and Developmental Biology, University of California at San Diego, La Jolla, CA, USA). If possible, at least two alleles for each gene were included for the comprehensive analysis. Infection of the T-DNA mutant lines with virulent Pseudomonas syringae pv. tomato strain DC3000 (Pst DC3000) revealed that mutations in Nup160/At1g33410 and Seh1/At1g64350 impair basal defense responses towards this pathogen, similar to the defect observed in mos3-1 (Figure 2a). Compared to the complete breakdown of basal resistance in the control eds1 null mutant Col eds1-2 (Bartsch et al., 2006), the nup160 and seh1 lines showed a slightly lower level of susceptibility. Mutants in the other predicted Nup107-160 complex members did not show altered susceptibility upon inoculation with Pst DC3000 (Figure 2a). We therefore focused our functional analyses on Nup160 and Seh1.
The three independent nup160 lines (Figure 1) all compromised basal resistance to Pst DC3000 to a similar degree (Figure 2a), are slightly smaller than wild-type and showed an early flowering phenotype (Figure 5b and Figure S1). According to The Arabidopsis Information Resource (TAIR), SAIL_877_B01 (nup160-3) carries a T-DNA insertion in the 12th exon of At1g33410, whereas SALK_126801 (nup160-4/sar1-4) carries an insertion in the 17th exon (Parry et al., 2006). The point mutation in sar1-1 introduces a premature stop codon in the 11th exon of the gene (Parry et al., 2006) (Figure 1).
In contrast to nup160, seh1 mutant plants are indistinguishable from Col-0 wild-type in terms of size and morphology (Figure S2). As only one T-DNA insertion line for Seh1 was available (insertion in the 4th exon), we cloned genomic Seh1/At1g64350 with 2.2 kb upstream of its translational start codon and stably expressed it in the SALK_022717 (seh1-1) mutant background to test for complementation of the resistance defect. As shown in Figure 2(b), genomic Seh1 fully complemented the enhanced disease susceptibility phenotype of SALK_022717 in multiple independent stable transgenic homozygous plants, confirming that the T-DNA integration in Seh1/At1g64350 is responsible for compromising basal resistance to Pst DC3000.
These data show that, in addition to MOS3/Nup96, Nup160 and Seh1 are two additional subunits of the predicted Nup107-160 complex in Arabidopsis required for the plant’s response to infection by virulent Pseudomonas bacteria. As mutations in the other five putative Nup107-160 complex members did not compromise basal resistance (Figure 2a), we reasoned that these members are either not required for basal resistance to Pst DC3000, or may function redundantly with other Nups.
Nup160 and Seh1 are required for TNL-type R protein-mediated immunity
Nup160 and Seh1 are equally important for basal resistance to the invasive hemi-biotrophic pathogen Pst DC3000. This prompted us to assess the genetic requirement for Nup160 and Seh1 in race-specific resistance mediated by various R proteins. We found that resistance conferred by the TNL receptor, RPS4, to Pst DC3000 expressing the effector avrRps4 (Hinsch and Staskawicz, 1996) was compromised in nup160 and seh1 mutants as indicated by the approximately 20-fold (seh1-1) and 40–130-fold (nup160 alleles) increase in bacterial growth 3 days after inoculation compared to wild-type Col-0 (Figure 3a). Loss of RPS4 resistance in seh1-1 was consistently less severe in multiple independent experiments in comparison with susceptibility of all nup160 mutants and the control mos3-1 (Figure 3a). Conversely, resistance conferred by the CNL receptors RPM1 and RPS2 to Pst DC3000 expressing the effectors avrRpm1 and avrRpt2, respectively (Mackey et al., 2002, 2003; Axtell and Staskawicz, 2003), remained intact in nup160 and seh1 (Figure 3b,c). However, RPM1-mediated resistance triggered after inoculation with Pseudomonas syringae pv. maculicola (Psm) ES4326 expressing the effector avrB was less tight than RPM1 resistance triggered after avrRpm1 recognition, and slightly increased growth of Psm ES4326 (avrB) was observed in nup160 and mos3-1 (Figures 3b and S3). This phenotype of mos3-1 is consistent with previous data (Zhang and Li, 2005).
We next tested the genetic requirement for Nup160 and Seh1 in resistance to the avirulent oomycete pathogen Hyaloperonospora arabidopsidis (Ha) isolate Emwa1, which is recognized by the TNL protein RPP4 (van der Biezen et al., 2002). seh1 mutants consistently prevented pathogen sporulation on leaves, as in Col-0. Staining of leaves with lactophenol trypan blue 7 days after inoculation revealed that seh1-1 exhibited a combination of spreading HR lesions and occasional development of spatially restricted trailing necrosis that typically contained pathogen infection structures before spreading across the whole leaf (Figure 3d). By contrast, we observed a significant decrease in RPP4 function in the nup160 mutant as seen by trailing host cell necrosis across entire leaves and the emergence of occasional sporangiophores (Figure 3d). We conclude that Nup160 and, to a minor extent, Seh1 are essential for full resistance conferred by TNL proteins against bacterial and oomycete pathogens. Nup160 may also have a minor contribution to immunity conferred by the CNL protein RPM1 after recognition of avrB.
snc1 auto-immunity is suppressed to different levels by nup160 and seh1
Mutations in MOS3/Nup96 suppress the auto-immune phenotypes of the auto-activated TNL variant, snc1, suggesting that this nucleoporin contributes to snc1-mediated resistance (Zhang and Li, 2005). To determine whether Nup160 and Seh1 are also involved in snc1-mediated auto-immunity, we crossed snc1 with nup160 and seh1 to obtain snc1 nup160 and snc1 seh1 double mutants, respectively. snc1 plants are stunted and have curly leaves, but snc1 nup160 plants no longer exhibit snc1-like morphology and resemble the nup160 mutant, which is slightly smaller than wild-type (Figure 4a). Mutations in Nup160 further suppress constitutive expression of the pathogenesis-related (PR) genes PR-1 and PR-2 in the snc1 mutant background as determined by semi-quantitative RT-PCR (Figure 4c). The snc1 single mutant exhibits enhanced resistance to virulent Pseudomonas syringae (Li et al., 2001; Zhang et al., 2003). To test whether enhanced disease resistance is impaired in the snc1 nup160 double mutant, we infected plants with Pst DC3000. As shown in Figure 4(d), resistance to Pst DC3000 was completely suppressed when Nup160 function was disabled in the snc1 mutant background, with an approximately 10-fold higher titer of Pst DC3000 compared with Col-0. In addition to increased resistance, the snc1 mutant accumulates high levels of the defense hormone salicylic acid (SA). HPLC analyses of SA extracts revealed a marked reduction of endogenous total SA levels in the snc1 nup160 mutant to levels comparable with those of Col-0 (Figure 4e). Together, our results show that nup160 fully suppresses snc1-related auto-immune phenotypes, consistent with a function for Nup160 in TNL-mediated immunity (Figure 3a,d).
The rather complete suppression of snc1 by nup160 is in contrast to the partial suppression observed when combining snc1 with seh1-1. snc1 seh1-1 double mutants retain the typical stunted morphology and curly leaves of snc1, but are significantly bigger than snc1 plants (Figure 4a,b). The elevated PR1 and PR2 gene expression (Figure 4c) and accumulation of SA (Figure 4e) in snc1 seh1-1 resemble the snc1 single mutant. The enhanced disease resistance of snc1 to Pst DC3000 is only partially suppressed by seh1, with bacterial titers in the snc1 seh1-1 double mutant being intermediate between those in snc1 and Col-0 plants (Figure 4d). In summary, the partial suppression of snc1 auto-immune responses and related growth inhibition by seh1-1 are consistent with the data that the seh1-1 mutation partially impairs the function of the TNL receptors RPS4 and RPP4 (Figure 3a,d).
Defects in both Nup160 and Seh1 do not further impair basal defense responses
An important function of both Nup160 and Seh1 is to maintain the basal resistance layer to virulent Pst DC3000 (Figure 2a). To test the relationship between the two genes within this resistance pathway, we generated nup160 seh1 double mutants and infected them with Pst DC3000. Bacterial growth in nup160-4 seh1-1 was not further increased compared with the single mutants (Figure 5a). The susceptibility of other double mutant combinations between nup160 and nup43-2, nup107-1 or nup133-1 was also similar to that of the nup160 single mutant (Figure 5a). Whereas nup160 seh1, nup160 nup43 and nup160 nup107 double mutants are similar to the nup160 mutant in terms of morphology and the timing of floral transition, nup160-4 nup133-1 plants show compound developmental defects. Although nup133-1 mutants are indistinguishable from wild-type, the nup160 early-flowering phenotype is further exacerbated in the nup160-4 nup133-1 double mutant and accompanied by a further reduction in rosette size (Figure 5b). These plants produce siliques that are significantly decreased in size and contain few or no seeds (inset in Figure 5b). An even more severe phenotype was reported previously for nup160mos3 double mutants that exhibit seedling lethality (Parry et al., 2006). Our epistasis analyses suggest that Nup160 and Seh1 function within the same pathway that confers basal resistance to virulent Pseudomonas bacteria. Because nup160-4 nup133-1 double mutants show more severe developmental defects than nup160, our results indicate a partially redundant function of Nup133 and Nup160, and suggest that loss of both subunits results in a further decrease in NPC function that is important for several aspects of plant development.
Seh1 is a nuclear and cytoplasmic protein
In vivo localization of translational Nup160 and Nup96/MOS3 fusions with green fluorescent protein (GFP) to the nuclear rim has previously been demonstrated in transient expression assays and roots of stable transgenic plants (Zhang and Li, 2005; Dong et al., 2006). Seh1 is a WD40-repeat protein that localizes to multiple subcellular locations in yeast, including the NPC, the nucleoplasm and the cytoplasm (Rout et al., 2000; Bai et al., 2004; Matsuyama et al., 2006; Alber et al., 2007), suggesting that part of its cellular pool is not permanently associated with the NPC. In transiently transfected Arabidopsis protoplasts, Seh1 has been reported to localize to the nucleus, the pre-vacuolar compartment and the Golgi complex, as determined by immunostaining of affinity-tagged Seh1 protein in fixed protoplasts (Lee et al., 2006). To investigate the subcellular localization of Seh1 in living Arabidopsis cells, we fused cyan fluorescent protein (CFP) to the C-terminus of Seh1 and expressed the fusion protein in the seh1-1 mutant background under the control of endogenous 2.2 kb sequence upstream of the Seh1 translation initiation codon or a CaMV double 35S promoter. Western blot analyses on leaf total protein extracts using an antibody recognizing CFP revealed expression of the Seh1–CFP fusion protein and the absence of detectable degradation products or free CFP (data not shown). To test whether the Seh1–CFP fusion protein is functional, multiple independent homozygous lines carrying a single insertion of the transgene were infected with virulent Pst DC3000. As shown in Figure 6(a), the Seh1–CFP fusion protein complemented the enhanced disease susceptibility phenotype of seh1-1 towards Pst DC3000 when expressed under the control of either the native promoter or the double 35S promoter, suggesting that the Seh1–CFP fusion protein is functional and correctly localized in Arabidopsis cells. The intracellular localization of stably expressed Seh1–CFP was examined in Arabidopsis leaves by confocal laser scanning microscopy (CLSM). We were unable to detect Seh1–CFP fluorescence in transgenic plants expressing the fusion protein under the control of the endogenous Seh1 promoter, probably due to low expression levels. In contrast, strong CFP fluorescence was observed inside the nuclei and cytoplasm of multiple independent transgenic plants expressing a translational Seh1 fusion to CFP under control of the double 35S promoter (Figure 6b). Localization of Seh1–CFP to the nucleus and cytoplasm is in contrast to the localization of Nup160 and MOS3/Nup96 to the NE (Zhang and Li, 2005; Dong et al., 2006), and suggests that Arabidopsis Seh1 shows dynamic association with the Nup107-160 complex. However, we cannot exclude the possibility that stable association of a proportion of Seh1–CFP with the NPC and thus concentration of CFP fluorescence at the nuclear rim of leaf cells is masked due to over-expression of the fusion protein by the double 35S promoter.
The nup160 and seh1 mutants are impaired in nuclear mRNA export and EDS1 protein accumulation
In vertebrate and yeast cells, the Nup107-160 complex has an important function in mRNA export during interphase, and defects in or depletion of individual members of this complex result in nuclear mRNA accumulation (Fabre et al., 1994; Vasu et al., 2001). In Arabidopsis, loss-of-function nup160/sar1 and mos3/sar3/nup96 mutant plants are also defective in nuclear mRNA export (Dong et al., 2006; Parry et al., 2006). As Seh1 is a predicted member of the Nup160/SAR1- and MOS3/SAR3/Nup96-containing Arabidopsis Nup107-160 nuclear pore sub-complex, and all three genes are required for disease resistance, we tested whether Seh1 is also essential for export of mRNAs from the nucleus to the cytoplasm. To localize poly(A) mRNAs in leaves of 7-day-old wild-type and seh1-1 seedlings, we performed whole mount in situ hybridization using a 5′-Alexa488-labeled oligo(dT)45 probe. CLSM revealed strong fluorescence inside nuclei of the seh1 mutant that was similar to the extent of fluorescence observed in nup160-3 and mos3-1 using identical microscope settings for image acquisition (Figure 7a) (Dong et al., 2006; Parry et al., 2006). In contrast, wild-type Col-0 showed only weak nuclear fluorescence (Figure 7a). These results indicate that seh1-1 mutant plants accumulate mRNA inside the nucleus, and that the functions of Nup160/SAR1, Nup96/MOS3/SAR3 and Seh1 are required for proper export of mRNAs from the nucleus to the cytoplasm.
In vertebrates, Nup96 plays an important role in immunity, where it is required for selective nuclear mRNA export of key interferon γ-regulated genes in response to viral infection (Faria et al., 2006). Depletion of Nup96 results in nuclear retention of these specific classes of mRNAs, and probably contributes to the lower expression level of interferon γ-regulated proteins. We therefore reasoned that the defects of the nup160 and seh1 mutants in basal and TNL-triggered immunity may, at least in part, be caused by reduced accumulation of regulators of these resistance pathways. To test this, we generated total protein extracts from leaves of both mutants and wild-type plants. Our immunoblot analyses revealed that total amounts of EDS1, an essential regulator of basal resistance and TNL receptor-mediated immunity (Parker et al., 1996; Aarts et al., 1998), are strongly reduced in nup160 mutant lines (to approximately 25% of the wild-type level) but less severely affected in the seh1-1 mutant (to approximately 70% of the wild-type level) (Figure 7b). EDS1 protein accumulation was unchanged in control nup85-1 plants that did not show enhanced susceptibility to Pst DC3000 (Figures 2a and S4). Significantly, the mutations did not obviously alter total levels of extractable PEPC, HSC70, TGA2, VDAC1 and histone H3 (Figure 7b), suggesting a preferential depletion of EDS1 in the nup160 mutant.
We next examined whether selectively impaired export of EDS1 transcripts from nuclei of the nup160 mutant may account for the strong depletion of EDS1 protein in nup160. Unexpectedly, our quantitative RT-PCR analysis of fractionated cytoplasmic and nuclear RNA did not reveal a preferential accumulation of EDS1 transcripts in nup160 nuclei. Instead, we observed that EDS1 mRNA levels were reduced in both the cytoplasmic and nuclear fraction of the nup160 mutant compared to wild-type levels (Figure 7c). We quantified EDS1 transcripts from total cellular RNA by quantitative RT-PCR to confirm a slight but significant reduction of EDS1 mRNA expression in nup160 (Figure S5). The reduction in overall EDS1 expression in nup160 (to approximately 75% of wild-type levels) was less severe than the decrease in EDS1 protein abundance (to approximately 25% of wild-type levels) (Figures 7b and S5). Our data suggest that both reduced nuclear mRNA export activity (Figure 7a) and overall reduced EDS1 expression (Figures 7c and S5) in nup160 affect EDS1-conditioned basal and TNL-type R protein-mediated resistance.
Transport of proteins and RNAs across the NE is fundamental for eukaryotic cellular communication and function, providing important means to control cell homeostasis, gene expression and signaling networks (Kaffman and O’Shea, 1999; Orphanides and Reinberg, 2002). Despite the pivotal role NPCs play in selectively regulating the bi-directional flow of information between the cytoplasm and the nucleoplasm, our knowledge of plant NPC composition and the function of its subunits is limited.
In this study, we used a reverse genetics approach to identify putative subunits of the Arabidopsis Nup107-160 nuclear pore sub-complex that are required for basal defense signaling. Of eight complex members predicted in the Col-0 reference genome (Figure 1 and Table S1), defense-related functions were demonstrated for two: Nup160 and Seh1. A third complex member, Nup96/MOS3, has previously been identified based on its requirement for constitutive resistance caused by the auto-activated TNL R gene, snc1 (Zhang and Li, 2005). Our pathology assays revealed that mutants of Nup160 and Seh1 show altered defense, not only in basal resistance (Figure 2) but also in TNL-type R protein-mediated immunity (Figure 3a,d). Although both genes are equally required for basal defense to virulent Pseudomonas bacteria, Nup160 appears to play a more prominent role in TNL receptor-triggered immunity to avirulent pathogens than Seh1 (Figure 3a,d). This is also reflected by the fact that the seh1-1 mutation only partially suppresses the constitutive resistance and growth inhibition conditioned by snc1 (Figure 4a,b,d). As SA accumulation and PR-1/PR-2 expression in snc1 seh1-1 were not significantly different from snc1 (Figure 4c,e), part of the nuclear and/or cytoplasmic activity of Seh1 may contribute to the SA-independent branch of plant defense responses activated in snc1 (Zhang et al., 2003). Significantly, growth of Pst DC3000 expressing the effectors avrRpm1 (recognized by the CNL-type R protein RPM1) or avrRpt2 (recognized by the CNL protein RPS2) was largely unaffected in nup160 and seh1 mutants (Figure 3b,c), demonstrating pathway specificity of these two components in plant defense signaling. Because resistance to Psm ES4326 expressing avrB was slightly relaxed in nup160 (Figure S3), we cannot exclude the possibility of a minor contribution of Nup160 to RPM1-triggered immunity after recognition of avrB, which may be less efficient than recognition of avrRpm1. The lack of obvious developmental defects in the seh1-1 mutant (Figures S1 and S2) and the rather mild pleiotropic defects in nup160 (Figures 5b and S1) further indicate a selective involvement of these two nuclear pore components in regulating responses to microbial pathogens in Arabidopsis.
Faria et al. (2006) previously reported that mice with reduced levels of the Nup107-160 complex member Nup96 are impaired in immunity due to specifically reduced export and thus translation of mRNAs encoding key immune regulators. The authors concluded that the reduced protein accumulation of immune regulators probably accounts for the immunity defects of Nup96-depleted mice (Faria et al., 2006). Because nup160 plants are selectively impaired in basal and TNL-type R protein-mediated resistance and accumulation of EDS1, a key positive regulator of both defense pathways (Figures 3 and 7b) (Feys et al., 2005; Wiermer et al., 2005), we assessed whether nup160 specifically affects EDS1 mRNA export from the nucleus. We did not observe a preferential retention of EDS1 mRNA in nuclei of the nup160 mutant, which is inconsistent with a primary role of Nup160 in mediating selectively the nuclear export of transcripts encoding EDS1. Our quantitative RT-PCR analysis of fractionated RNA does not exclude the possibility that EDS1 is among the transcripts retained inside nuclei of nup160. Instead, our data suggest a more general mRNA export defect in which the mRNAs of the housekeeping genes Ubiquitin5 (UBQ5) (Figure 7c) or Protein Phosphatase 2A (PP2A) (data not shown) used for data normalization show the same subcellular distribution pattern. The observed accumulation of poly(A) mRNA inside mutant nuclei would thus reflect a broad range of transcripts. However, the finding that the nup160 and seh1 mutants are selectively impaired in basal and TNL-triggered immunity (Figure 3a,d) and EDS1 protein accumulation (Figure 7b) is remarkable and inconsistent with a major inhibition of bulk mRNA export, as this would have a more significant impact on CNL-triggered immunity as well and probably result in catastrophic phenotypes and/or death. Our data therefore suggest that the mRNA export defect of nup160 and seh1 is rather mild, possibly because these mutants (like several other tested Nup107-160 complex mutants) may produce truncated proteins and thus may not be functionally null (Figure S6). This suggests that EDS1-dependent resistance pathways are particularly prone to disturbances in mRNA export. Loss of Seh1 function might have a minor effect on EDS1 protein abundance (Figure 7b) and TNL-triggered immunity (Figures 3a,d and 4) due to partial compensation by Nup160 or other Nups.
Notably, our analyses also revealed a slight reduction of total EDS1 transcript in the nup160 mutant (Figure S5), suggesting that Nup160 may also contribute to transcriptional activation mechanisms. An intriguing possibility is that Nup160 itself promotes EDS1 expression. Studies in yeast demonstrated that members of the Nup84 complex (which is equivalent to the vertebrate Nup107-160 complex) are capable of activating transcription in vivo, and, importantly, do so in the normal context of the NPC by tethering target genes to the inner side of the nuclear pore, thereby coupling transcription with efficient mRNA export (Menon et al., 2005). Another recent study revealed a positive role for the Nup84 complex in RNA polymerase II-mediated transcription elongation that is functionally coupled to its role in mRNA export as part of an intact NPC (Tous et al., 2011). Transcriptional elongation probably contributes to the control of transcript accumulation and thus transcriptional efficiency (Akhtar and Gasser, 2007). Therefore, it is also possible that Nup160 may fine-tune EDS1 transcript accumulation at the level of elongation.
Alternatively, Nup160 might contribute to the nuclear accumulation of transcriptional regulators required for full expression of EDS1 and/or regulate nuclear translocation rates of immune regulatory proteins essential for basal and TNL-mediated resistance. Consistent with this idea, Parry et al. (2006) observed reduced nuclear accumulation of the transcriptional repressor IAA17 in the nup160/sar1 mutant, providing a possible explanation for its altered auxin sensitivity. In contrast, nuclear import of the cold response regulator ICE1 is unchanged in nup160 plants, which are also defective in tolerance to cold stress (Dong et al., 2006), suggesting a different extent to which reduced nuclear accumulation of protein regulators contributes to the observed defects in nup160. In this context, we tested whether mutations in Nup160 affect the proper coordination of EDS1 nuclear and cytoplasmic pools across the NE which appears to be essential for full immunity (Garcia et al., 2010). We observed that the level of EDS1 protein was reduced proportionally in both the cytoplasm and nucleus of nup160 plants (Figure S7). Although this may be due to impaired EDS1 expression and mRNA export in the nup160 mutant, we cannot exclude the existence of additional processes that influence EDS1 protein stability. A post-transcriptional effect was observed in Arabidopsis mos7/nup88 mutant plants, in which reduced nuclear retention of EDS1 apparently influences EDS1 protein stability, and reduced EDS1 amounts become equilibrated between cellular compartments over time (Cheng et al., 2009). Despite this reduction in EDS1 protein accumulation, mos7-1 plants show normal EDS1 transcript levels, and are, like their nup88/mbo counterpart in Drosophila, unlikely to accumulate mRNA inside the nucleus (Roth et al., 2003; Cheng et al., 2009). Also distinct from nup160 and seh1, mos7-1 mutants are strongly affected in CNL receptor-mediated resistance, implying selective contributions of individual Nups to the regulation of plant immune responses by nucleo-cytoplasmic transport.
Plant growth, mutant isolation and pathology assays
Plants were grown in soil at 22°C under a 10 h (for pathology assays) or 16 h light regime. The sar1-1 (Parry et al., 2006) and mos3-1 (Zhang and Li, 2005) mutants have been described previously. T-DNA insertion mutants were obtained from the Arabidopsis Biological Resource Center, and genotyped by PCR using insertion-flanking primers. Plant cell death and H. arabidopsidis infection structures were visualized at 7 days post-inoculation under a light microscope after staining leaves with lactophenol trypan blue (Aarts et al., 1998). Virulent and avirulent Pst DC3000 strains were as described previously (Aarts et al., 1998). For bacterial growth assays, suspensions of 1 × 105 colony-forming units ml−1 5 mm MgCl2 with 0.001% v/v Silwet L-77, http://www.lehleseeds.com/cgi-bin/hazel.cgi?action=DETAIL&item=85 were vacuum-infiltrated into leaves of 5-week-old plants, bacterial titers were determined after 1 h (d0) and 3 days (d3), and colony numbers were compared between lines using a two-tailed Student’s t-test.
Nuclear/cytoplasmic fractionation, RNA extraction and gene expression analyses
Nuclear and cytosolic fractions for RNA extraction were obtained using a protocol modified from that described by Park et al. (2005). Leaf material (2 g) from 4-week-old soil-grown plants was ground in liquid nitrogen and mixed with 2 ml cell-wall disruption buffer [10 mm Tris pH 7.5, 10 mm NaCl, 10 mm MgCl2, 10 mmβ-mercaptoethanol, 10% v/v glycerol, 100 U ml−1 Ribolock (Fermentas, http://www.fermentas.com)]. The homogenate was spun through a 95 μm nylon mesh, and the flow-through was centrifuged at 2500 g and 4°C for 10 min to pellet the nuclei. The supernatant was re-centrifuged at 13 000 g and 4°C for 15 min, and the supernatant of this second centrifugation was saved as the cytoplasmic fraction. The nuclear pellet from the first centrifugation was carefully resuspended in nuclei washing buffer (10 mm Tris pH 7.5, 10 mm NaCl, 10 mm MgCl2, 10 mmβ-mercaptoethanol, 1 m hexylene glycol, 0.5% Triton X-100), and centrifuged at 1500 g for 10 min at 4°C. The supernatant was discarded, and washing and centrifugation of nuclear pellets was repeated five or six times. Each nuclear pellet was resuspended in 75 μl cell-wall disruption buffer, and nuclear fractions of each genotype were pooled for RNA extraction. As quality controls for the fractionation, immunoblot analyses were performed using the nuclear marker protein histone H3 and the cytoplasmic marker protein PEPC. Total RNA was isolated from cytoplasmic and nuclear fractions or directly from 4-week-old soil-grown plants using TRIzol (Invitrogen, http://www.invitrogen.com/). A 1.5 μg aliquot of RNA was reverse-transcribed using RevertAid H Minus M-MuLV reverse transcriptase (Fermentas) and 0.5 μg oligo(dT)18V primer at 42°C in a 20 μl reaction volume. Aliquots of reverse transcription reaction products were used for semi-quantitative PCR (0.5 μl) or quantitative PCR (1 μl of a 1:7.5 dilution). Quantitative RT-PCRs were performed using a CFX96 real-time PCR detection system (Bio-Rad, http://www.bio-rad.com/) using SsoFast EvaGreen Supermix (Bio-Rad). Ubiquitin5 (UBQ5, At3g62250) and Protein Phosphatase 2A (PP2A, At1g13320) transcript levels were used for data normalization. Relative transcript levels were calculated using cfx Manager software (version 2.1) and the non-linear regression method. All primers are listed in Table S2.
DNA constructs, transgenic plants and protein localization studies
The construct used to complement the seh1-1 mutation was generated by PCR amplification of a Col-0 genomic fragment containing the Seh1/At1g64350 coding region and 2.2 kb upstream of the ATG start codon as well as 2.2 kb downstream of the stop codon. Primers Seh1.F.BamHI (5′-CGCGGATCCGACTTGTATATACCGTCTCG-3′) and Seh1.R.EcoRI (5′-CCGGAATTCGTCGCAGACTGAGATTCTTG-3′) were used for PCR, and the BamHI/EcoRI-digested fragment was cloned into pGreenII0229 (Hellens et al., 2000) for complementation analysis. The construct was introduced into Agrobacterium tumefaciens strain GV3101 (pMP90) harboring the helper plasmid pSOUP (Hellens et al., 2000), and transformed into SALK_022717 using the floral-dip method (Clough and Bent, 1998). To generate plants expressing Seh1–CFP under the control of the 35S promoter or the native Seh1 promoter, a genomic fragment spanning the full-length Col-0 Seh1 gene without the stop codon was PCR-amplified either with 2.2 kb upstream of the translation initiation codon (promSeh1-gSeh1) or without the promoter sequence (gSeh1), using the primers gAt1g64350.D-TOPO.F (5′-CACCATGGCGAAATCAATGGCGACG-3′) and Seh1R.GW (5′-GGAGGGAACTGGTTCAAGCG-3′) for gSeh1, and Seh1.F.GW (5′-CACCGACTTGTATATACCGTCTCG-3′) and Seh1.R.GW (5′-GGAGGGAACTGGTTCAAGCG-3′) for promSeh1-gSeh1. Both fragments were cloned into pENTR/D-TOPO (Invitrogen) and sequenced. Gateway LR reactions were performed between the pENTR/D-TOPO clones and the binary destination vectors pXCG-CFP and pXCSG-CFP (Feys et al., 2005) to generate the expression constructs pXCG-promSeh1::gSeh1::CFP and pXCSG-prom35SS::gSeh1::CFP, respectively. Constructs were transferred to A. tumefaciens GV3101 (pMP90RK) and transformed into SALK_022717 as described above. CLSM was performed on a Leica SP5-DM6000 (http://www.leica.com/) using an excitation wavelength of 458 nm.
Endogenous SA levels were determined as described previously (Li et al., 1999).
Whole-mount in situ localization of mRNA
Poly(A) RNA in situ hybridizations were performed on 7-day-old seedlings grown on half-strength MS medium as previously described (Gong et al., 2005; Germain et al., 2010) using a 5′-Alexa488-labeled oligo(dT)45 probe. Leaves were observed using a Nikon PCM-2000 confocal laser scanning microscope (http://www.nikon.com/) equipped with a 488 nm argon excitation laser. All images were taken using a 40× objective at equivalent laser intensity.
We thank the Arabidopsis Biological Resource Center for providing T-DNA insertion mutants, Mark Estelle (Section of Cell and Developmental Biology, University of California at San Diego, La Jolla, CA, USA) for sar1-1 seeds, Jane Parker (Department of Plant-Microbe Interactions, Max Planck Institute for Plant Breeding Research, Cologne, Germany) for Col eds1-2 seeds and purified EDS1 antisera, and Rene Fuchs for help with CLSM. We are grateful for financial support to M.W. by a research fellowship from the Alexander von Humboldt Foundation and the Deutsche Forschungsgemeinschaft. We acknowledge a Natural Sciences and Engineering Research Council of Canada postgraduate fellowship to Y.T.C., and thank the Natural Sciences and Engineering Research Council of Canada (Discovery Grant program) for funding to X.L.