SlNAC1, a stress-related transcription factor, is fine-tuned on both the transcriptional and the post-translational level



  • The plant-specific NAC (NAM, ATAF1,2, CUC2) transcription factors play significant roles in diverse physiological processes. In this study, we determined the regulation of a stress-related tomato (Solanum lycopersicum) NAC1 (SlNAC1) transcription factor at both the transcriptional and the post-translational level.
  • The SlNAC1 protein was found to be stable in the presence of proteasome-specific inhibitor MG132 or MG115 and ubiquitinated in plant cells, suggesting that the SlNAC1 is subject to the ubiquitin–proteasome system-mediated degradation. Deletion analysis identified a short segment of 10 amino acids (aa261–270) that was required for ubiquitin–proteasome system-mediated degradation, among which two leucine residues (L268 and L269) were critical for the protein instability of SlNAC1. Fusion of the degron (SlNAC1191–270) containing these 10 amino acids to green fluorescent protein was found to be sufficient to trigger the degradation of the fusion protein.
  • In addition, the SlNAC1 gene is strongly upregulated during Pseudomonas infection, while repression of the NAC1 ortholog in Nicotiana benthamiana resulted in enhanced susceptibility to Pseudomonas bacteria.
  • These results suggest that rapid upregulation of the NAC1 gene resulting in more protein production is likely one of the strategies plants use to defend themselves against pathogen infection.


Plants are continuously exposed to biotic and abiotic stresses. To survive, plants have evolved complex mechanisms to efficently adapt to these stresses. One of the major cellular events in plant defense response is the timely expression of stress-related genes whose products are involved in a variety of physiological and metabolic processes. In most cases increased tolerance to stress is a result of the action of a number of genes each contributing part of a multifaceted response (Fujita et al., 2006). Transcription factors play a pivotal role in induction of stress-related genes by directly or indirectly activating many target genes. However, the regulation of transcription factors involved in stress response is still largely unknown (Fujita et al., 2006; Yamaguchi-Shinozaki & Shinozaki, 2006).

In eukaryotic cells, a variety of proteins are degraded though the ubiquitin–proteasome system (UPS). These proteins are tagged with a ubiquitin chain, which is recognized by the 26S proteasome for proteolysis-based degradation. Ubiquitin–proteasome system-mediated protein degradation is determined by a specific degradation-relevant region, termed a degron, which serves as a signal for ubiquitination and degradation (Ravid & Hochstrasser, 2008; Schrader et al., 2009). In plants, UPS-mediated degradation not only cleans up misfolded or/and damaged proteins to avoid potential toxicity, it also adjusts the amount of regulatory proteins, including transcription factors, to control many physiological processes, including senescence, development and stress responses, in a temporal and spatial manner (Zeng et al., 2006; Stone & Callis, 2007; Santner & Estelle, 2010).

NAC (NAM, ATAF1,2, CUC2) transcription factors constitute one of the largest families of plant-specific transcription factors, and contain a conserved N-terminal NAC domain involved in DNA binding and a highly variable C-terminal domain responsible for transcriptional activation (Olsen et al., 2005). Although NAC transcription factors were originally identified because of their role in diverse developmental processes (Souer et al., 1996), a growing body of evidence suggests that they also play a significant role in both biotic and abiotic stress responses. These hypotheses have been made mainly based on observations of rapid induction of NAC genes in response to stress stimuli and enhanced stress tolerance in transgenic plants over-expressing these genes (Hu et al., 2006; Wu et al., 2009; Zheng et al., 2009; Jeong et al., 2010; Takasaki et al., 2010; Hao et al., 2011; Liu et al., 2011; Nakashima et al., 2011; Xue et al., 2011). In addition to rapid gene expression in response to stress stimuli, NAC transcription factors can be regulated at the mRNA or/and protein level. For example, the mRNA of CUC1 and CUC2, two genes controlling boundary establishment and maintenance of apical meristem in Arabidopsis, is targeted by miR164 for degradation (Laufs et al., 2004). Although it is not orthologous to SlNAC1 Arabidopsis AtNAC1 is ubiquitinated by a ring ubiquitin E3 ligase SINAT5 for proteasomal degradation, thereby attenuating auxin signaling (Xie et al., 2000, 2002). The rice NAC transcription factor RIM1, which is involved in jasmonic acid (JA) signaling, is also degraded via a 26S proteasome-dependent pathway in response to treatment with JA (Yoshii et al., 2010). Rice OsNAC4 transcription factor, which is involved in hypersensitive (HR) cell death during gene-for-gene resistance reaction (Fujiwara et al., 2004), is subjected to phosphorylation modification (Kaneda et al., 2009). The phosphorylation of OsNAC4 is required for its subcellular localization to the nucleus, where it may activate > 100 genes directly or indirectly (Kaneda et al., 2009).

Tomato (Solanum lycopersicum) has > 40 predicted NAC family members ( The SlNAC1 gene has been shown to be induced in leaves during Prf (Pseudomonas resistance and fenthion sensitivity )-mediated gene-for-gene resistance (Salmeron et al., 1996) to Pseudomonas syringae pv. tomato (Pst) and in roots upon high salinity treatment (Mysore et al., 2002; Ouyang et al., 2007). Prf-mediated resistance to Pst is determined by the indirect recognition of Pst-secreted effectors AvrPto and AvrPtoB by Prf resistance protein (Pedley & Martin, 2003). The Prf recognition partner Pto kinase physically interacts with AvrPto or AvrPtoB, which is presumably monitored by Prf and consequently triggers the resistance response (Pedley & Martin, 2003; Oh & Martin, 2011). In addition, the SlNAC1 protein interacts with Tomato leaf curl virus (TLCV) replication enhancer (REn), and the SlNAC1 gene is induced in a REn-dependent manner during TLCV infection (Selth et al., 2005). Given the likely role of SlNAC1 in the defense response, TLCV may have hijacked the plant defense system to facilitate viral replication during infection.

In this study, we show that the SlNAC1 gene plays an important role in disease resistance to Pseudomonas bacteria. Our results also suggest that the SlNAC1 protein is subject to UPS-mediated degradation in plant cells. We further identify a short segment of 10 amino acids (aa261–270) that is essential but not sufficient for UPS-mediated degradation, while aa191–270 can serve as a degron to trigger the degradation of green fluorescent protein (GFP) fusion proteins.

Materials and Methods

Plasmid construction

Assembly of C-terminal HA-tagged constructs expressed in S. lycopersicum) protoplasts involved PCR amplification of full-length and truncated SlNAC1 cDNAs and ligation into the KpnI and StuI sites of the pTEX vector. The fusion protein with a C-terminal HA tag was expressed under the control of the Cauliflower mosaic virus (CaMV) 35S promoter. Generation of L268D, L269D, L268A/L269A, L268D/L269D and L268R/L269R mutant proteins involved PCR amplification of cDNAs with primers containing corresponding mutations. Generation of GFP fusion constructs involved ligation of the GFP cDNA into the KpnI and StuI sites of pTEX, followed by subcloning of the SlNAC1 cDNA into the restored StuI site of pTEX. Constructs used for transient expression in N. benthamiana leaves were constructed by PCR amplification of SlNAC1 cDNA and ubiquitin cDNA followed by ligation into the KpnI–SalI and BamHI–SalI sites of pBTEX, respectively. The resulting plasmids are designed to express fusion proteins with a C-terminal FLAG or an N-terminal HA tag. All constructs were confirmed by DNA sequencing. Primers used in the cloning are listed in the Supporting Information, Table S1.

Tomato protoplast transient expression

Tomato protoplast preparation was carried out following a recently published protocol (Nguyen et al., 2010). Plasmid DNA (20 μg) of each construct was used for transformation of protoplasts in the presence of polyethylene glycol (PEG), as described by He et al. (2006). MG132 or cycloheximide (CHX; Sigma-Aldrich) were added at 1 h or 8 h after transformation, respectively. Protoplasts were collected by centrifugation at 12 000 g for 1 min and lysed with extraction buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA, 1% Triton X-100, 1 mM dithiothreitol (DTT), 10 μl ml−1 plant protease inhibitor cocktail from Sigma-Aldrich). Proteins were resolved by 10% sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE), and detected by Western blotting analysis using anti-HA antibody (Roche Applied Science).

Agrobacterium-mediated transient expression in N. benthamiana and in vivo ubiquitination assay

Agrobacterium tumefaciens GV2260 strains expressing FLAG-tagged SlNAC1 (SlNAC1-FLAG) or HA-tagged ubiquitin (HA-Ub) were infiltrated into N. benthamiana leaves at a concentration of 0.4 OD600 in the presence of 100 μM MG132. Leaf tissue was harvested 2 d after infiltration and proteins were extracted with extraction buffer (50 mM Tris, pH 7.5, 150 mM NaCl, 5 mM EDTA, 2 mM DTT, 10% glycerol, 1 mM PMSF, 1% PVPP and 10 μl ml−1 plant protease inhibitor cocktail). After centrifugation, supernatants were collected. One-tenth of each supernatant (v : v) were saved as the input material, and the rest of the supernatant was used for immunoprecipitation with anti-HA affinity matrix (Roche Applied Sciences) or anti-FLAG affinity matrix (Sigma-Aldrich) for 2 h at 4°C. Matrix beads containing the immunoprecipitated complex were pelleted and washed at least three times with extraction buffer. The SDS loading buffer was added to the matrix beads and boiled at 100°C for 5 min to release the immunoprecipitated complex. After a brief centrifugation at 500 g for 15 s, supernatants were loaded on 7.5% SDS-PAG for Western blotting using the alternate antibody (anti-FLAG or anti-HA) as the primary antibody.

Virus-induced gene silencing (VIGS)

Silencing of the NAC1 ortholog in N. benthamiana involved a unique 300-bp fragment of SlNAC1 cDNA (encoding aa201-301) cloned into the pTRV2 vector at the EcoRI and XbaI sites for TRV-based VIGS in N. benthamiana, as described previously (Fig. S1) (Liu et al., 2002). Two-week-old N. benthamiana seedlings were used for tobacco rattle virus infection (TRV), and NAC1-silenced N. benthamiana plants were used for the Pseudomonas inoculation 4 wk after TRV infection.

Pseudomonas inoculation into plants and real-time PCR analysis

Two P. syringae pv tomato strains, DC3000∆avrPto∆avrPtoB∆hopQ1-1 (Wei et al., 2007) and DC3000hrcC (Deng et al., 1998) used in this work were prepared as described previously (Anderson et al., 2006). Bacteria were collected and resuspended in 10 mM MgCl2 containing 0.003% Silwet-77 at a concentration of 2 × 107 colony forming units (CFU) ml−1 (for real-time PCR assay) or 2 × 10CFU ml−1 (for disease assay). Real-time PCR analysis of SlNAC1 expression used leaf tissue samples harvested at 0, 2, 4 and 8 h after infiltration. Total RNAs were isolated with TRIzol reagent (Invitrogen) and treated with DNase. Reverse transcription was conducted using SuperScriptII reverse transcriptase (Invitrogen) and real-time PCR analysis was performed on an ABI Prism 7700 sequence detection system using SYBR Green reagents (Life Technologies, Carlsbad, CA, USA). The tomato Actin gene was used as an internal reference. Relative expression ratios were determined using the rest software (Pfaffl et al., 2002). Primers used in real-time PCR are listed in Table S1.


SlNAC1 expression is induced by Pseudomonas infection

Previously, SlNAC1 has been shown to be induced during the Prf-dependent gene-for-gene resistance to P. syringae pv tomato (Pst) in tomato (Mysore et al., 2002). However, it is not known whether the induction of SlNAC1 is specific to the gene-for-gene resistance or is a general defense response to Pseudomonas infection. If the latter is the case, can Pseudomonas bacteria suppress the expression of SlNAC1 for pathogenesis during disease development? To address these questions, we examined the expression of SlNAC1 gene in three different interactions between tomato and Pst: (1) the incompatible gene-for-gene resistance interaction in which resistant RG-PtoR plants (expressing resistance gene Prf) were inoculated with PstDC3000; (2) the compatible disease interaction in which susceptible RG-prf3 plants (containing a 1 kb deletion in the Prf gene; Salmeron et al., 1996) were inoculated with PstDC3000; and (3) the basal defense interaction in which susceptible RG-prf3 plants were inoculated with the non-pathogenic PstDC3000 hrcC mutant strain (Deng et al., 1998). We monitored the transcript abundance of the SlNAC1 gene with real-time PCR using total RNA isolated at different time points after bacterial infiltration. The RG-prf3 plants infiltrated with 10 mM MgCl2 served as a mock control. As shown in Fig. 1, the expression of the SlNAC1 gene was rapidly induced in all three interactions. First, the SlNAC1 gene was induced by infiltration of the non-pathogenic PstDC3000 hrcC strain, suggesting that the SlNAC1 gene is involved in basal defense, which is triggered by recognition of pathogen-associated molecular patterns (PAMPs) from the pathogen. These PAMPs are highly conserved and functionally essential among pathogens, thus PAMP-triggered defense is a basal-level defense response against potential pathogens in plants (Boller & Felix, 2009). Second, SlNAC1 was also induced in RG-prf3 plants inoculated with the PstDC3000 strain, with a slightly delayed induction peak time. Third, in the case of gene-for-gene resistance (RG-PtoR plants inoculated with the PstDC3000 strain), the induction of SlNAC1 was elevated to a much higher level, suggesting that defense signaling is significantly amplified in gene-for-gene resistance. We also noted that SlNAC1 was transiently induced in mock inoculation at 2 h after infiltration, followed by a decline in expression to the basal level in 8 h. This transient induction in response to wound or mechanical stresses has also been observed in the expression of other defense-related genes (Cheong et al., 2002; Kang et al., 2003). Together, these data suggest that SlNAC1 is involved in the basal defense of tomato in response to Pseudomonas bacterium and defense signaling is significantly enhanced in the case of gene-for-gene resistance.

Figure 1.

Induction of the SlNAC1 gene during different plant–pathogen interactions. Resistant RG-PtoR (expressing the Prf resistance gene) or susceptible RG-prf3 (possessing a 1 kb deletion in the Prf gene) tomato (Solanum lycopersicum) plants were inoculated with appropriate PstDC3000 strains as indicated at an inoculum of 2 × 10colony forming units (CFU) ml−1 or mock solution (10 mM MgCl2). Total RNA was isolated at the indicated time-points after Pst infiltration and were used to generate first-strand cDNA. The relative expression level of the SlNAC1 gene was analy3ed by real-time PCR using SlNAC1 gene-specific primers. The expression of the tomato Actin gene served as an internal control for normalization. Values are means ± SE of three replicates. The experiment was repeated three times with similar results.

The SlNAC1 protein is degraded by the UPS pathway

To further assess the nature of SlNAC1 protein in vivo, we transiently expressed HA-tagged SlNAC1 in tomato protoplasts. SlNAC1-HA was driven by the strong CaMV 35S promoter and protein accumulation was monitored by Western blotting analysis using anti-HA antibody. In our first attempts, the SlNAC1 protein could not be detected by Western blotting (data not shown), suggesting the SlNAC1 protein was rapidly degraded in plant cells. To test if the SlNAC1 protein was indeed subjected to degradation and, if so, which proteolysis mechanism was responsible for its instability, we applied several proteolysis inhibitors, including MG132 and MG115 (proteasome-specific inhibitors), plant protease inhibitor cocktail (protease inhibitors mixture), BAF (bafilomycin, a specific inhibitor of vacuolar autophagy) and phenylmethylsulfonyl fluoride (PMSF – a serine protease inhibitor), 1 h after protoplast transformation. As shown in Fig. 2(a), the SlNAC1 protein was stabilized by the addition of proteasome-specific inhibitors MG132 or MG115, but not by other proteolysis inhibitors. It is notable that there is basal level accumulation of SlNAC1 in the absence of inhibitors, furthermore PMSF may cause a slight increase in the accumulation of SlNAC1. These results strongly indicate that SlNAC1 is specifically degraded via the proteasome pathway in plant cells.

Figure 2.

Ubiquitin–proteasome system (UPS)-mediated degradation of the SlNAC1 protein in plant cells. (a) Stabilization of SlNAC1 by proteasome-specific inhibitors. Tomato (Solanum lycopersicum) protoplasts were transformed with 35S::SlNAC1-HA in the presence of polyethylene glycol (PEG). One hour after transformation, protoplasts were treated for 6 h with 0.2% dimethyl sulfoxide (DMSO) (negative control), 50 μM MG132, 50 μM MG115, plant protease COCKTAIL (Sigma-Aldrich), 300 nM bafilomycin (BAF) or 1 mM phenylmethylsulfonyl fluoride (PMSF). Proteins were extracted and separated by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). The accumulation of SlNAC1 protein was determined by Western blotting using anti-HA antibody. The levels of the Rubisco band stained by Coomassie Blue served as a loading control. (b) Detection of the in vivo ubiquitination of SlNAC1 by co-immunoprecipitation. Agrobacterium tumefaciens GV2260 strains harboring the Cauliflower mosaic virus (CaMV) 35S promoter-driven constructs for epitope-tagged ubiquitin (HA-Ub) or SlNAC1 (SlNAC1-FLAG) were infiltrated using a syringe into Nicotiana benthamiana leaves at a concentration of 0.4 OD600. Agrobacterium tumefaciens GV2260 strain containing the empty vector (Ev) was used as a control. One hundred micromoles of MG132 was co-injected with the Agrobacterium suspension to prevent the degradation of SlNAC1 protein. Two days after agrobacterium infiltration, proteins were extracted for immunoprecipitation with anti-FLAG affinity matrix (left panel) or anti-HA affinity matrix (right panel), followed by Western blotting using the opposite antibodies (anti-HA or anti-FLAG) to determine the association of poly-ubiquitin chain with SlNAC1, which appears as smear bands on Western blot, as indicated. The input lanes contained one-tenth of the protein extract used for co-immunoprecipitation. The asterisk indicates a non-specific cross-reaction band detected by antibody.

SlNAC1 protein is ubiquitinated in plant cells

In general, proteasome-mediated degradation is based on recognition of the ubiquitin chain attached to the targeted protein. To establish the correlation between protein ubiquitination and degradation, it is necessary to determine the ubiquitination of SlNAC1 in vivo. To this end, we attempted to co-express SlNAC1 and ubiquitin in tomato protoplast cells to determine if they associate with each other in vivo. However, the efficiency of co-transformation of the two proteins in tomato protoplasts was very low so another approach was tried. We took advantage of a N. benthamiana leaf transient expression system that is more amenable to co-expression of two proteins (Dong et al., 2009). We co-expressed N-terminally HA-tagged ubiquitin (HA-Ub) and C-terminally FLAG-tagged SlNAC1 (SlNAC1-Flag) constructs in N. benthamiana leaves via Agrobacterium-mediated transient expression. The association of HA-Ub with the SlNAC1-Flag was examined by co-immunoprecipitation (co-IP) assay. After immunoprecipitation with anti-HA antibody matrix, the immunoprecipitated complex was verified by Western blotting using anti-FLAG antibody. As shown in Fig. 2b (left panel), a smear representing the polyubiquitinated SlNAC1 proteins was detected by anti-FLAG antibody in the anti-HA-immunoprecipitated complex from co-expression of HA-Ub and SlNAC1-Flag, but not in the immunoprecipitated complex from co-expression of HA-Ub and vector control (Fig. 2b, left panel), suggesting that SlNAC1 can be poly-ubiquitinated in vivo for proteasome-mediated degradation. This result was confirmed by a reverse co-IP experiment using the anti-HA antibody matrix to pull down HA-Ub followed by Western blotting using anti-Flag antibody (Fig. 2b, right panel).

A short segment of 10 amino acids in the C-terminus of SlNAC1 protein is essential for its degradation

To identify the determinant region of SlNAC1 responsible for the UPS-mediated degradation, a series of HA-tagged SlNAC1 truncations were generated and transiently expressed in tomato protoplasts. These truncations expressed peptides of the N-terminal 290, 280, 270, 260, 250, 240, 220 and 200 amino acids and were referred to as SlNAC11–290, SlNAC11–280, SlNAC11–270, SlNAC11–260, SlNAC11–250, SlNAC11–240, SlNAC11–220 and SlNAC11–200, respectively. Protein stability of the truncations was assessed by Western blotting using anti-HA antibody. As shown in Fig. 3a, in the absense of the proteasome inhibitor MG132, truncations expressing amino acids within the N-terminal 260 amino acids (SlNAC11–200, SlNAC11–220, SlNAC11–240, SlNAC11–250 and SlNAC11–260) were stable, whereas other truncations expressing regions beyond the N-terminal 260 amino acids (SlNAC11–270, SlNAC11–280 and SlNAC11–290) were unstable to the same extent as the full-length SlNAC1 protein (aa1–301). Significantly, when proteasome inhibitor MG132 was added, all these unstable truncations could accumulate in plant cells (Fig. 3b). These results demonstrat that this short region (aa261–270) in the C-terminus of SlNAC1 is important for its UPS-mediated degradation.

Figure 3.

Identification of a short segment of 10 amino acids (aa261–270) essential for the ubiquitin–proteasome system (UPS)-mediated degradation of SlNAC1. A series of HA-tagged SlNAC1 truncations or green fluorescent protein (GFP) fused with SlNAC1 truncations were expressed from the Cauliflower mosaic virus (CaMV) 35S promoter in tomato (Solanum lycopersicum) protoplasts. Protein stability was assessed by Western blotting analysis using anti-HA antibody. (a) Stable SlNAC1 truncations not containing aa260–270. Protoplasts transformed with SlNAC1 truncations were harvested for Western blotting assay 8 h after transformation. (b) Stabilization of SlNAC1 truncations by proteasome inhibitor MG132. Fifty micromoles of MG132 was added 1 h after transformation and protoplasts were collected for Western blotting assay 6 h later. (c) Stability of HA-tagged GFP proteins fused with SlNAC11–260, SlNAC11–270 or SlNAC11–301 (the full length). Appropriate GFP fusion constructs, as indicated, were transformed in tomato protoplasts. Fifty micromoles of cycloheximide (CHX) was applied to prevent de novo protein synthesis and protoplasts were collected at different time-points to determine the stability of fusion proteins by Western blotting assay.

The significance of aa261–270 in UPS-mediated degradation was further verified by GFP fusion analysis. The GFP protein is stable when expressed in plant cells and has been used widely as a control for protein stablity assays (Sen et al., 2007). We expressed GFP fusion proteins with SlNAC11–260, SlNAC11–270 and full-length SlNAC11–301 in tomato protoplasts. Eight hours after protoplast transformation, 50 μM cyclohexmide (CHX) was added to prevent de novo protein synthesis. The stability of fusion proteins were assessed by Western blotting. As shown in Fig. 3(c), fusion of SlNAC11–270 or full-length SlNAC11–301 with GFP dramatically compromised GFP stability. By contrast, when GFP was fused with SlNAC11–260 the fusion protein was still stable. Again, these results suggest that aa261–270 of SlNAC1 is critical for UPS-mediated degradation.

Identification of a degron region in the SlNAC1 protein

We then tested whether the 10 amino acids of aa261–270 alone could function as a degron or not. We fused these 10 amino acids to GFP and tested the stability of the fusion protein when expressed in tomato protoplasts. As shown in Fig. 4a, the GFP-SlNAC1261–270 was as stable as GFP alone, indicating that aa261–270 was not sufficient to trigger protein degradation. It is likely that all or some of the N-terminal 260 amino acids are also required for recognition by UPS for degradation. To identify the minimum region that could act as a degron, we expressed GFP variants fused with SlNAC1241–270, SlNAC1211–270, or SlNAC1191–270. As shown in Fig. 4(b), the fusion of SlNAC1191–270 to GFP significantly triggered GFP degradation, whereas fusions with other short regions did not affect GFP stability. Thus, aa191–270 can serve as a functional degron that is sufficient for recognition and degradation by UPS, and aa260–270 are an essential part of this function. Fig. 5 shows the alignment of SlNAC1 with putative orthologs from Solanum tuberosum, N. benthamiana and Arabidopsis thaliana, highlighting the degron region and its terminal ten amino acids.

Figure 4.

Identification of a degron region in SlNAC1. Green fluorescent protein (GFP) was fused with different SlNAC1 fragments and expressed from the Cauliflower mosaic virus (CaMV) 35S promoter in tomato (Solanum lycopersicum) protoplasts. Fifty micromoles of cycloheximide (CHX) was applied to prevent de novo protein synthesis and protoplasts were collected at different time-points to determine the stability of fusion proteins by Western blotting using anti-HA antibody. (a) SlNAC1261–270 was unable to trigger the degradation of GFP. (b) A minimal region of aa191–270 is sufficient to trigger protein degradation. Similar to (a), GFP fusion proteins were expressed in protoplasts. Eight hours after transformation, 50 μM of CHX was added to prevent de novo protein synthesis and proteins were extracted at different time-points for Western blotting assay.

Figure 5.

ClustalW2 alignment of four putative NAC1 orthologs. SlNAC1, StNAC1 and AtATAF1 correspond to accessions Solanum lycopersicum NP_001234482, S. tuberosum CAC42087 and Arabidopsis thaliana NP_171677, respectively. NbNAC1 is the predicted protein sequence based on the homologous coding region from Nicotiana benthamiana genome v0.4.2. Shaded residues correspond to SlNAC1 191–270. Three lysines residing in the degron region are indicated in bold type. Residues SlNAC1 261–270 are underlined. Consensus symbols asterisk, colon and period indicate fully conserved residue, residue with strongly similar properties and residue with weakly similar properties, respectively.

L268 and L269 in aa261–270 are critical for SlNAC1 degradation

We next attempted to identify amino acid residues in aa261–270 that are critical for UPS-mediated degradation. For this, we generated a single alanine substitution of the 10 amino acids (except Y264) in the SlNAC11–270 template. None of these single alanine subsitutions could stabilize SlNAC11–270 (Fig. S2). We then generated four additional truncations with only one amino acid difference, SlNAC11–266, SlNAC11–267, SlNAC11–268 and SlNAC11–269. Transient expression of these truncations in tomato protoplasts showed that SlNAC11–267 and SlNAC11–268 accumulate to a moderate amount (Fig. 6a), suggesting that L268 and L269 are important for the instability of SlNAC11–270. To verify this result, we substituted L268, or L269 or both, with aspartic acid (D), arginine (R) or alanine (A) in the SlNAC11–270 template and found that substitutions with aspartic acid but not arginine or alanine, affect SlNAC11–270 stability. As shown in Fig. 6(b), SlNAC11–270 (L268D) and SlNAC11–270 (L269D) could barely be detected without the addition of MG132, whereas the L268D/L269D double mutation dramatically stablized SlNAC11–270. All alanine or arginine substitution mutants were still not detectable. Together, our results show that both L268 and L269 are critical for mediating SlNAC11–270 degradation. As only a strong amino acid change at L268 and L269 (from a neutral and hydrophobic amino acid to an acidic residue) could eliminate protein instability it is possible that L268 and L269 provide a stuctural base for recognition by UPS.

Figure 6.

An additive role of L268 and L269 in the instability of SlNAC11–270. A series of HA-tagged SlNAC1 truncations or green fluorescent protein (GFP) variants fused with SlNAC1 truncations were expressed under the Cauliflower mosaic virus (CaMV) 35S promoter in tomato (Solanum lycopersicum) protoplasts. Protein stability was assessed by Western blotting using anti-HA antibody. Eight hours after transformation, protoplasts with (+) or without (−) the application of 50 μM MG132 were harvested and proteins were extracted for Western blotting using anti-HA antibody. (a) Stability of SlNAC1 truncations in protoplasts. (b) Stability of single mutation SlNAC11–270, SlNAC11–270(L268D) or SlNAC11–270(L269D), and double mutation SlNAC11–270(L268A L269A), SlNAC11–270(L268D L269D) or SlNAC11–270(L268R L269R) in protoplasts.

The NAC1 transcription factor plays a role in plant disease resistance

We next sought to determine the biological role of the NAC1 transcription factor in plant disease resistance. To this end, we first attempted to repress SlNAC1 expression in tomato plants using VIGS. However, as is common in tomatoes, VIGS did not silence SlNAC1 in tomato in three attempts. We thus took advantage of N. benthamiana, which is readily amendable to VIGS. The N. benthamiana NAC1 ortholog (NbNAC1) shares 81% identity with SlNAC1 (Fig. 5). Furthermore, previous research has shown that using tomato DNA can efficiently silence N. benthamiana orthologs (Ekengren et al., 2003). The presence of unique motifs in the C-terminus of subfamilies within the NAC transcription factor family (Zhu et al., 2012) suggests that targeting this region for silencing gives a high level of specificity. The unique 3′ region of SlNAC1 cDNA was used to target NbNAC1 for VIGS (Fig. S1) and silenced plants were assessed for the altered susceptibility to the virulent P. syringae pv tomato (Pst) DC3000∆avrPto∆avrPtoB∆hopQ1-1 strain (Wei et al., 2007). To differentiate the subtle effect on disease resistance caused by the silencing of NbNAC1, a low titer (2 × 104CFU ml−1) of bacterial inoculum was infiltrated into N. benthamiana plants and disease was assessed 4 d after bacterial infiltration. As shown in Fig. 7, there was greater bacterial growth and more severe disease symptoms in the NAC1-silenced N. benthamiana plants than in the control plants, indicating that NbNAC1-silenced N. benthamiana shows enhanced disease susceptibility to Pseudomonas bacteria. The RT-PCR analysis verified silencing of NbNAC1 in the N. benthaminana-silenced plants (Fig. 7c). Our results thus suggest that the NAC1 transcription factor plays an important role in plant disease resistance.

Figure 7.

Silencing of NAC1 in Nicotiana benthamiana affects resistance against bacterial pathogen Pseudomonas syringae pv tomato (Pst). Four-week-old NAC1-silenced N. benthamiana plants (TRV2::NAC1) or control plants (TRV2::00) were inoculated with PstDC3000∆avrPto∆avrPtoB∆hrpQ1-1 strain (Wei et al., 2007) at 2 × 104 colony forming units (CFU) ml−1. Susceptibility to Pseudomonas bacterium was assessed by disease symptoms and bacterial populations in leaves at 2 d and/or 4 d after bacterial infiltration. (a) Leaf disease symptoms caused by PstDC3000∆avrPto∆avrPtoB∆hrpQ1-1 4 d after Pseudomonas infiltration. Necrotic lesions were more severe in NAC1-silenced plant leaves than in control plant leaves. Similar phenotypes were observed in four independent experiments (three plants for each experiment). (b) Measurements of bacterial population in NAC1-silenced and control plants at 2 d and 4 d after bacterial infiltration. Graph data represent two independent experiments (n = 6). Statistical significance determined by Student's t-test and  0.0005 for all significant differences (indicated by letters above bars) between plants. Error bar, ± SD. Note that at both day 2 and day 4 the bacterial populations in TRV2::NAC1 plants were eightfold more than in the TRV2::00 control plants. (c) Verification of silencing of the NAC1 gene in NAC1-silenced N. benthamiana plants by semi-real-time reverse transcription PCR assay, with the EF1-α gene as the internal reference gene.


Ubiquitin–proteasome system-mediated degradation is a common mechanism by which levels of proteins are regulated. Transcription factors control the expression of a number of target genes, so it is important to strictly adjust the amount of these proteins to prevent unnecessary activation of target gene expression. In this study, we presented data showing that the stress-responsive SlNAC1 transcription factor is regulated by UPS. While SlNAC1 was barely detectable when expressed in tomato cells, proteasome-specific inhibitors MG132 and MG115, but not other proteolysis inhibitors, suppressd SlNAC1 degradation (Fig. 2a). Analysis of ubiquitination in planta showed that SlNAC1 was polyubiquitinated when transiently co-expressed with ubiquitin in N. benthamiana leaves (Fig. 2b). Many NAC transcription factors have been identified but with limited characterization of their post-translational modification. While not the ortholog to SlNAC1, AtNAC1, which is involved in auxin signaling, is the only NAC protein known, to date, to be ubiquitinated and degraded by UPS (Xie et al., 2002). Examination of other SlNAC1 orthologs shows that the potato ortholog StNAC1 is induced by oomycete pathogen Phytophthora and salt stress (Collinge & Boller, 2001), while the Arabidopsis ortholog of SlNAC1, ATAF1, is induced by drought, high-salinity, abscisic acid, methyl jasmonate, mechanical wounding and Botrytis cinerea infection (Wu et al., 2009). Moreover, it has been demonstrated that ATAF1 is a negative regulator of defense responses to both necrotrophic fungal and bacterial pathogens (Wang et al., 2009), suggesting that the same gene may play different roles in plant defense in different plant species. However, whether StNAC1 and ATAF1 are also subjected to the UPS-mediated degradation is still to be determined. Although no evidence of polyubiquitination has been reported, another A. thaliana NAC protein, ANAC, has been shown to interact with several RING domain ubiquitin E3 ligases (Greve et al., 2003). Recently, a rice NAC transcription factor, RIM1, has been shown to be stabilized by MG132 (Yoshii et al., 2010), implying that UPS is the cellular apparatus responsible for degradation. Thus, like other regulatory proteins, NAC transcription factors may undergo UPS-mediated degradation as a major regulatory process.

Protein degradation by UPS is mediated by recognition of degradation signals (degrons) within the target proteins. A degron is defined as the minimal element within a protein that is sufficient for recognition and degradation by the proteolytic apparatus (Ravid & Hochstrasser, 2008). In general, rather than serving as the ubiquitin binding site, the degron can direct binding of the target protein to ubiquitin E3 ligase via various means, such as providing a contacting interface with ubiquitin E3 or site(s) subject to post-translational modification(s) essential for ubiqutination (Ravid & Hochstrasser, 2008; Schrader et al., 2009). We identified a short segment of 10 amino acids (aa261–270) that is important for SlNAC1 degradation (Fig. 3). Among these 10 amino acids, L268 and L269 contribute additively to SlNAC1 degradation, as shown by the observation that the L268D L269D double mutation, but not the L268D or L269D single mutation, significantly stabilized the SlNAC11–270 protein (Fig. 6b). As there is no lysine residing in this region, it appears that aa261–270, in particular L268 and L269 residues, play an important role in directing ubiquitination for degradation through the proteasome. It is notable that only subtitution of leucine (L) with acidic amino acid aspartic acid (D) resulted in the elimination of protein instability. These two residues could provide a structural base, maintaining a conformation or involving a modification, both of which are primed for ubiquitination. Thus, we speculate that the aa261–270 region could serve as a contact interface for binding with a ubiquitin E3 ligase, or could mediate specific post-translational modification(s) that are critical for UPS recognition. The latter explanation is consistent with the GFP fusion analysis. When fused with GFP, aa261–270 was unable to trigger the degradation of GFP fusion protein (Fig. 4a). By contrast a longer region of aa191–270, including these 10 amino acids, could mediate GFP degradation (Fig. 4b), suggesting that the amino acids outside the aa261–270 region are also required for proteasome-specific degradation. Given the fact that GFP is not ubiquitinated (Sen et al., 2007) and a minimum region of aa191–270 can act as a functional degron for recognition and degradation by UPS, aa191–270 could confer sites for both ubiquitin E3 binding and ubiquitin chain attachment. It would be interesting to test whether the lysine residues (K191, K242 and K249) in the aa191–270 degron (Fig. 5) are required for ubiquitination and degradation.

Several NAC transcription factors have been found to be induced by virulent pathogens (Collinge & Boller, 2001; Nakashima et al., 2007; Kaneda et al., 2009; Nuruzzaman et al., 2010; Xia et al., 2010), but none of them has been examined with regard to their inducibility by a nonpathogenic strain. The induction of SlNAC1 by nonpathogenic PstDC3000 hrcC strain suggests that it is involved in plant basal defense, which is known to be triggered by detection of PAMPs. In fact, rice OsNAC4, involved in gene-for-gene resistance-mediated HR cell death, has been found to be induced by bacterial flagellin PAMP (Fujiwara et al., 2004). In the current model of plant–pathogen interactions, it is proposed that the basal defense is the first layer of mechanisms used by plants to defend against potential pathogens (Boller & He, 2009). Virulent pathogens, such as PstDC3000, have evolved mechanisms to suppress PAMP-triggered basal defense. Interestingly, unlike other basal defense-related genes, such as Nho1 (Kang et al., 2003; Li et al., 2005) and FRK1 (He et al., 2006), whose induction is repressed by PstDC3000, the induction of SlNAC1 was not repressed by PstDC3000 during compatible disease development, as shown in the case of susceptible RG-prf3 plants inoculated with virulent PstDC3000 strain (Fig. 1). This suggests that plant basal defense genes may be differentially manipulated by pathogens during infection. It is striking that the SlNAC1 was highly induced during gene-for-gene resistance response, as shown in the case of resistant RG-PtoR plants inoculated with the virulent PstDC3000 strain (Fig. 1). This suggests that gene-for-gene resistance may activate SlNAC1 expression through different signaling pathways, or simply amplify the PAMP-triggered defense signaling by an unknown mechanism. Regardless, SlNAC1 appears to play a role in both gene-for-gene resistance and basal defense, the latter can be expanded to include stress tolerance, as exemplified by the induction of SlNAC1 by high salinity (Ouyang et al., 2007).

Although we were unable to further examine the role of SlNAC1 in gene-for-gene resistance (also called incompatible interaction) because of inefficient VIGS of SlNAC1 gene in resistant tomato expressing the Prf resistance gene, our gene silencing analysis in N. benthamiana reveals that the NAC1 transcription factor is required for resistance to Pseudomonas infection (Fig. 7). Thus, our results suggest that the SlNAC1 transcription factor is fine-tuned at both transcriptional and post-translational levels and is important for plant disease resistance (Figs 1, 2, 7). It is reasonable to speculate that under normal condition, plants have to tightly regulate transcription factors such as SlNAC1 to prevent autoactivation of stress response signaling: at the post-translational level, a rapid turnover of SlNAC1 protein may help maintain the signaling balance, while at the transcriptional level, the SlNAC1 gene is expressed at a basal level to produce a limited amount of SlNAC1 protein. When exposed to biotic stress, such as pathogen infection or abiotic stress, such as high salinity, plants may use two strategies to counteract the degradation of SlNAC1 protein. The first strategy is rapid expression of the SlNAC1 gene in order to produce more protein, ideally producing more protein than can be degraded. Alternatively, plants could interfere with the ubiquitination of SlNAC1 to prevent its degradation, such as down-regulation of the gene encoding the ubiquitin E3 ligase that is responsible for SlNAC1 ubiquitination. Unfortunately, because of a lack of SlNAC1-specific antibody, we were unable to determine the ubiquitination and/or degradation of native SlNAC1 protein during biotic/abiotic responses. Generation of epitope-tagged SlNAC1-HA transgenic plants under the control of its native stress-inducible promoter would allow us address this question. In addition, it would be interesting to identify and characterize the ubiquitin E3 ligase that specifically targets SlNAC1 for ubiquitination and degradation.


We thank Drs Zonglie Hong and Allan Caplan and Anthony Trakas for critical reading of the manuscript. This work was supported in part by grants from the Agriculture and Food Research Initiative Competitive Grants Program from the USDA (No. 2010-6511-42056), the Idaho State Department of Agriculture and the Idaho Potato Commission (F.X.), the Idaho State Department of Agriculture and the Idaho Potato Commission (J.C.K.) and the National Science Fund Distinguished Young Scholars No. 30825030 (Y.L.)