Interaction between Cucumber mosaic virus 2b protein and plant catalase induces a specific necrosis in association with proteasome activity
- 2.6k Downloads
Cucumber mosaic virus (CMV) can induce a specific necrosis on Arabidopsis through the interaction between the CMV 2b protein and host catalase, in which the ubiquitin–proteasome pathway may be involved.
We previously reported that the CMV 2b protein, the viral RNA silencing suppressor, interacted with the H2O2 scavenger catalase (CAT3), leading to necrosis on CMV-inoculated Arabidopsis leaves. We here confirmed that CMV could more abundantly accumulate in the CAT3-knockout mutant (cat3), and that CAT3 makes host plants a little more tolerant to CMV. We also found that the necrosis severity is not simply explained by a high level of H2O2 given by the lack of CAT3, because the recombinant CMV, CMV-N, induced much milder necrosis in cat3 than in the wild type, suggesting some specific mechanism for the necrosis induction. To further characterize the 2b-inducing necrosis in relation to its binding to CAT3, we conducted the agroinfiltration experiments to overexpress CAT3 and 2b in N. benthamiana leaves. The accumulation levels of CAT3 were higher when co-expressed with the CMV-N 2b (N2b) than with CMV-Y 2b (Y2b). We infer that N2b made a more stable complex with CAT3 than Y2b did, and the longevity of the 2b–CAT3 complex seemed to be important to induce necrosis. By immunoprecipitation (IP) with an anti-ubiquitin antibody followed by the detection with anti-CAT3 antibodies, we detected a higher molecular-weight smear and several breakdown products of CAT3 among the IP-proteins. In addition, the proteasome inhibitor MG132 treatment could actually increase the accumulation levels of CAT3. This study suggests that the host proteasome pathway is, at least partially, responsible for the degradation of CAT3, which is manifested in CMV-infected tissues.
KeywordsCucumber mosaic virus Catalase Ubiquitin–proteasome 2b Protein Arabidopsis thaliana Necrosis
Cucumber mosaic virus
Programmed cell death
Reactive oxygen species
Reactive oxygen species (ROS), such as hydrogen peroxide (H2O2) and O2 −, are generated during numerous physiological processes, including photosynthesis, plant development, and resistance responses against pathogens. H2O2 serves as an important molecular messenger to induce a form of programmed cell death (PCD) and especially called as the hypersensitive reaction (HR) in plant–pathogen interactions. Catalase is one of the most important antioxidant enzymes that catalyze the decomposition of H2O2, thus playing a role in protecting cells from H2O2 toxicity. Arabidopsis has three catalase enzymes (Frugoli et al. 1996), and catalase 3 (CAT3) is the most abundantly expressed and controlled by a circadian rhythm. CAT3 expression is enhanced with plant age and is accompanied by H2O2 accumulation in vascular bundles (Zimmermann et al. 2006; Hu et al. 2010). CAT3 has been found to interact with several proteins, such as nucleoside diphosphate kinase, NDK1 (Fukamatsu et al. 2003), class 3 sucrose-nonfermenting 1-related kinase, SOS2 (Verslues et al. 2007), and LESION SIMULATING DISEASE1, LSD1 (Li et al. 2013), and even a viral protein, the 2b protein (2b) of Cucumber mosaic virus (CMV) (Inaba et al. 2011). The interactions between catalase and other proteins may cause the diverse effects on catalase’s function. For example, Zou et al. (2015) demonstrated that the interaction between CAT3 and calcium-dependent protein kinase 8 (CPK8) enhanced CAT3 activity to maintain H2O2 homeostasis in response to drought stress. On the other hand, some interactions cause functional disturbance of catalase resulting in the accumulation of H2O2 and subsequent cell death (i.e., necrosis); the interactions of CAT3 with LSD1 and with 2b have been reported to be involved in necrosis on Arabidopsis (Inaba et al. 2011; Li et al. 2013).
CMV, the type member of the genus Cucumovirus, has a broad host range of more than 1000 plant species. It has a tripartite genome consisting of RNAs 1–3. RNAs 1 and 2 encode viral helicase and replicase, respectively, for viral replication, and RNA 3 encodes the viral movement protein 3a. RNA 4, a subgenomic RNA derived from the 3′ half of RNA 3, is the mRNA for the coat protein, while RNA 4A, a subgenomic RNA from RNA 2, encodes 2b (Ding et al. 1994). 2b is known as an RNA silencing suppressor (RSS) and also functions in viral cell-to-cell and long-distance movement (Ding et al. 1995; Ji and Ding 2001; Soards et al. 2002; Shi et al. 2003; Goto et al. 2007). In addition, 2b contains nuclear localization signals (NLSs) that are required for the manifestation of viral symptoms and for RSS activity (Lucy et al. 2000; Lewsey et al. 2009). We previously found a protein–protein interaction between 2b and Arabidopsis CAT3, which apparently causes H2O2 accumulation and subsequent necrosis in infected Arabidopsis leaves. The interaction between 2b and CAT3 also dramatically changes the localization of CAT3, which is normally localized in the cytoplasm; CAT3 was translocated to the nucleus in the presence of 2b (Inaba et al. 2011; Masuta et al. 2012).
Although catalase is well known to play an important role in regulating HR through the decomposition of H2O2 during plant–pathogen interactions, there are not many reports that describe the molecular details of the catalase-mediated pathways against viruses. For CMV in pepper plants, catalase activity was important for determining the degree of host susceptibility to CMV (Petrova et al. 2009). In addition, it was shown that CMV infection significantly induced catalase expression in squash plants (Havelda and Maule 2000). We also observed that 2b’s RSS activities were cancelled by a high level of CAT3 expression in the protoplast experiment (Inaba et al. 2011), and that CAT3-overexpressing transgenic Col-0 lines showed the suppression of CMV multiplication until 7 dpi, although the levels of CMV reached those of the nontransgenic control plants at 14 dpi. Therefore, CAT3 seems to play an antiviral role in CMV infection, but a role of CAT3 in CMV tolerance of Arabidopsis still remains unknown. For the necrosis induction, we reasoned that CAT3 and 2b were important, but we did not have any answer to explain the phenomenon that the necrosis severity greatly varied depending on the CMV strains and Arabidopsis ecotypes. Here, we further investigated the mechanism underlying the manifestation of necrosis symptoms observed in Arabidopsis infected with CMV. Our results of this study suggest that the stability of the CAT3-2b complex is important for the necrosis, and that the proteasome system is involved in degrading CAT and regulating the induction of necrosis in Arabidopsis.
Materials and methods
Plant materials and viruses
For Arabidopsis thaliana, ecotype Col-0 and the CAT3-knockout Col-0 mutant (cat3) were used in this study. Arabidopsis was grown in a growth chamber at 21 °C with 12 h photoperiod (150 μmol/m2/s). Nicotiana benthamiana, which were used for agroinfiltration, was grown at 23 °C with 16 h light/8 h dark. To create CAT3-complemented plants (CAT3/cat3), the homozygous cat3 plant (cat3/cat3, T-DNA insertion line) was transformed with the wild-type CAT3 cDNA. We first PCR-amplified the cDNA covering the ORF using a primer pair (the forward primer, 5′-GGACTAGTATGGATCCTTACAAGTATCGTCC-3′ and the reverse primer, 5′-GCGGAGCTCCTAGATGCTTGGCCTGACGTTCAG-3′) based on the sequence in GenBank (accession no. NM_001035996). The cDNA was then inserted in the plant expression vector, pIG121-Hm, to create pIG21-CAT3. cat3 plants were transformed with pIG21-CAT3 by the conventional floral dip method. T1 plants were selected for resistance to hygromycin and used for the subsequent inoculation experiments. CMV-Y infectious clones (pCY1, pCY2, and pCY3) (Suzuki et al. 1991) and the CMV-Y-based vectors of CMV-A1 (Otagaki et al. 2006) and CMV-H1 (Matsuo et al. 2007) were previously constructed. A1Ds and H1Ds were created by inserting the Ds-Red2 gene into CMV-A1 or CMV-H1, respectively (Takeshita et al. 2012). CMV-N has the CMV-Y backbone, but it contains a different C-terminal of the 2b protein; CMV-N was coincidentally created by inserting a 100-bp DNA fragment into the CMV-A1 vector.
Viral inoculation and fluorescence microscopy
The plasmids containing the full-length cDNAs of RNAs of CMV were transcribed in vitro. Leaves of 6-week-old plants of N. benthamiana were dusted with carborundum and rub-inoculated with the in vitro-transcribed RNAs. For Arabidopsis, 4-week-old plants were inoculated with the sap of infected tissues. Ds-Red2 fluorescence images were taken essentially according to Takeshita et al. (2012). In brief, red fluorescence of inoculated leaves of Arabidopsis Col-0 and cat3 was acquired using SMZ1500 (Nikon) with Ds-Red2 filter sets. For each inoculum, a set of four plants was used. Leaves of different plants were removed and used for imaging at 11 dpi.
The BiFC plasmid vectors for transient expression (Singh et al. 2009) were kindly supplied by Dr. S. Mano, National Institute for Basic Biology, Japan. The full-length cDNA of the CAT3 gene of Arabidopsis was cloned in either pGWnG or pnGGW, while the 2b gene was cloned in either pcCGGW or pGWcCG. All constructs were created by the Gateway Technology (Invitrogen). To amplify the designed fusion genes from the constructs containing the inserts, PCR was conducted using the forward primer (T7 promoter sequence + the 5′ end sequence of the ORF for the N-terminal protein) and the reverse primer (oligo-dT of 66 T residues + the sequence of the 3′ nontranslated region just before the terminator). Capped RNAs were then in vitro-transcribed from the PCR products and subsequently co-transfected into N. benthamiana protoplasts as essentially described before (Shimura et al. 2008a, b).
For the MG132 treatment, either healthy Col-0 leaves or CMV-Y-inoculated leaves at 2 dpi were detached from the basal part of the petiole, and the leaves were then transferred to glass tubes containing 50 μM MG132 (Sigma), which was originally dissolved in DMSO, and incubated at 21 °C with 12 h photoperiod for 3 days before protein extraction.
Total RNA was isolated using the Trizol reagent (Invitrogen) essentially as described before (Kim et al. 2008). Total RNA (100 ng) was used for the first-strand cDNA synthesis by AMV reverse transcriptase (Nippon Gene). Quantitative PCR was performed using Universal SYBR Select Master Mix (Applied Biosystems) in a StepOne Real-Time PCR System (Applied Biosystems). The Arabidopsis tubulin gene (AtTub) was used as an internal control. Primer sets for each gene amplification were as follows: 5′-GAGGGAGCCATTGACAACATCTT-3′ and 5′-GCGAACAGTTCACAGCTATGTTCA-3′ (for AtTub), 5′-GCGCGTCGACGTTGACGTCGAGCACCAAC-3′ and 5′-CCATCGATTGGTCTCCTTTTGGAGGCC-3′ (for CMV).
The plasmid construct of pBE2113:CAT3-FLAG with a FLAG tag sequence at the 3′ end has been already described (Inaba et al. 2011). In addition, the FLAG-CAT3 with a FLAG tag at the 5′ end and the 2b gene of CMV-N (N2b) were inserted in the Ti plasmid vector pBE2113 in this study. The agroinfiltration was conducted according to Goto et al. (2007). Agrobacterium (KYRT1) culture containing each construct (FLAG-CAT3, CAT3-FLAG, Y2b, and N2b) was prepared to an optimal density (OD) at 600 of 1.0 and infiltrated in N. benthamiana leaves using a 1-ml syringe. Total proteins were extracted 3 days postinfiltration (dpi) and then subjected to Western blot analysis.
Immunoprecipitation and Western blot analysis
Total protein was extracted from the inoculated leaves essentially as described before (Masuta et al. 1995). Immunoprecipitation was performed using Dynabeads protein G (Life Technologies) with anti-ubiquitin antibody (Abcam) according to the method essentially described by He and Kermode (2010). Western blots were probed using either anti-FLAG (Sigma) antibody or anti-CAT3 antibodies (Inaba et al. 2011). The anti-CAT3 antibodies can recognize the catalase(s) in N. benthamiana (at least NbCAT1) as well as Arabidopsis CAT3, because CAT3 and NbCAT1 share 94 % amino-acid sequence similarity.
Data were evaluated using Student’s t test. A P value of <0.05 was considered to be significant.
CMV symptom, accumulation, and movement in CAT3-knockout mutant (cat3)
Specific induction of necrosis through the interaction between 2b and CAT3
CAT degradation through the ubiquitin–proteasome pathway depending on a different combination of 2b and CAT3
To analyze the interaction between 2b and CAT3, we previously produced transgenic Col-0 plants that overexpress the CAT3 gene under the 35S promoter (Inaba et al. 2011). With difficulty, we finally obtained several CAT3-overexpressing transgenic lines, but they produced CAT3 at most ~twofold more than in the wild type (Inaba et al. 2011). We assumed that when the accumulation levels of CAT3 exceed a certain threshold, the CAT3 protein levels must be lowered by a well-regulated mechanism, because the CAT3 levels should be very critical to maintain the cellular redox balance. Based on the detection of the higher molecular weight, ubiquitinated smear band (above the 57 kDa) in the agroinfiltration experiment, we considered that the proteasome pathway regulated CAT3 degradation. CAT3 expression levels are drastically controlled by a circadian rhythm, in which the amplitude of the oscillations in CAT3 mRNA accumulation is ~fivefold; the lowest peak is in the early morning (Zhong and McClung 1996). If CAT3 functions according to the circadian rhythm and is regulated at the mRNA level, the synthesized protein should be quickly degraded. We thus consider that even in healthy Col-0 plants, CAT3 levels must be reset by protein degradation along with the circadian rhythm. In fact, a proteasome inhibitor, MG132, treatment increased 1.3-fold CAT3 accumulation compared to the untreated control, while the CAT3 levels were increased ca. twofold by MG132 when plants were infected with CMV-Y (Fig. 6). These results provide the evidence that the accumulation levels of CAT3 in the presence of CMV 2b are significantly affected by the ubiquitin–proteasome pathway.
As for the involvement of the proteasome in the regulation of cellular CAT3 levels and association of programmed cell death with the generation of ROS, this study is not the first one. He and Kermode (2010) have already demonstrated that white spruce plants actually used the proteasome to control the levels of CAT (a homolog of Arabidopsis CAT3) during the seedling development. One important observation that we share with them is the finding of several proteins with lower molecular weights as breakdown products of CAT after MG132 treatment. As shown in Fig. 5, we also detected several lower molecular weights bands (20–30 kDa) in our Western blots, suggesting that those breakdown products of CAT3 may have been generated through the same ubiquitin–proteasome pathway, as is the case for the white spruce CAT. When these results are considered together, it is likely that the proteasome-mediated Arabidopsis CAT3 degradation plays an important role in viral symptom expression and host defense in CMV-infected plants.
Although we here described that interaction between 2b and CAT3 promoted CAT3 degradation probably by the ubiquitin–proteasome pathway, the phenomena that viral proteins can promote degradation of host factors through the proteasome system are not rare events (reviewed by Verchot 2014). As for the viral RSS-mediated degradation of a host factor, Chiu et al. (2010) demonstrated that P25, the RSS of PVX, could induce the degradation of host AGO1 via proteasomes by agroinfiltration in N. benthamiana. The P0 proteins of Polerovirus and Enamovirus have been also shown to target AGO1 for degradation (Baumberger et al. 2007; Fusaro et al. 2012). On the other hand, it is noteworthy that many viral RSSs are ubiquitinated to be targeted for proteasome degradation (reviewed by Alcaide-Loridan and Jupin 2012). Conversely, the HC-Pro protein, the RSS of Papaya ringspot virus (PRSV), can inhibit host proteasomes, enhancing viral infection (Sahana et al. 2012). Therefore, viral RSSs appear to be deeply involved in the ubiquitin–proteasome pathway during plant–virus interactions whether viruses use it to degrade cellular proteins for their own benefit or inhibit it to prevent them from being a target.
In this study, we demonstrated that CAT3 would work as a modulator of the 2b-inducing necrosis, and that the catalase binding to 2b would be degraded through the proteasome pathway. Considering the catalase activity to erase H2O2 triggering HR, it is conceivable that the observed necrosis may be a simple result of catalase consumption by the interaction between 2b and CAT3. However, there are also some other explanations. For example, Li et al. (2013) previously demonstrated that the interaction between LSD (an important negative regulator of PCD) and Arabidopsis catalases (CATs) played an important role in pathogen-induced PCD, which requires the accumulation of salicylic acid. This LSD1-mediated PCD may be involved in the CMV-induced necrosis. Alternatively, based on the observation by He and Kermode (2010) that CAT was ubiquitinated and degraded by the proteasome, in white spruce just before extensive PCD occurred, we may raise another possibility that the 2b’s binding to CAT3 can induce necrosis by a certain specific mechanism, which is not necessarily explained only by the catalase activity. This idea is consistent with our finding that the necrosis severity and development depend on the 2b’s sequences; we here assume the involvement of resistance gene (R gene)-mediated HR. Because the necrosis induced by CMV infection accompanied H2O2 generation and expression of the PR genes (Inaba et al. 2011), this necrosis could be regarded as an HR-like defense response. In addition, N2b could induce very severe necrosis (even lethal necrosis) not only in Arabidopsis but also in N. benthamiana, suggesting the involvement of systemic HR-like necrosis driven by an R gene, depending on the nature of the interaction between 2b and catalase. Several reports, indeed, suggest that the ubiquitin–proteasome pathway is deeply involved in HR-like cell death in host defense responses (Liu et al. 2002; Kim et al. 2003; Yang et al. 2006; Sadanandom et al. 2012). For example, gene silencing of the COP9 signalosome, a multiprotein complex involved in protein degradation via the ubiquitin–proteasome pathway compromised the R gene (N gene)-mediated HR on tobacco infected with Tobacco mosaic virus (TMV) (Liu et al. 2002). Similarly, silencing of an ACRE276, E3-ubiquitin ligase in the proteasome pathway resulted in loss of two HRs in tobacco: N-mediated HR to the Avr protein p50 of TMV and also Cf9-mediated HR to the Avr protein Avr 9(4) of Cladosporium fulvum (Yang et al. 2006). Although any R gene that can recognize CMV 2b has not been identified, we rather prefer the idea that a CAT3-2b complex may induce HR-like cell death in a putative R gene-mediated resistance, which is associated with the ubiquitin–proteasome pathway.
Author contribution statement
KM, HS, and CM designed the experiments and wrote the paper. MT conducted the inoculation experiments using the CMV vector expressing Ds-Red.
We thank Dr. David Baulcombe for the PVX vector. We are also grateful to Dr. S. Mano for providing the BiFC plasmids and Mr. T. Taenaka for his technical assistance. This work was partly supported by JSPS KAKENHI, Grants-in-Aid for Scientific Research (B) 24380025.
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- Hu YQ, Liu S, Yuan HM, Li J, Yan DW, Zhang JF, Lu YT (2010) Functional comparison of catalase genes in the elimination of photorespiratory H2O2 using promoter- and 3′-untranslated region exchange experiments in the Arabidopsis cat2 photorespiratory mutant. Plant Cell Environ 33:1656–1670CrossRefPubMedGoogle Scholar
- Masuta C, Tanaka H, Uehara K, Kuwata S, Koiwai A, Noma M (1995) Broad resistance to plant viruses in transgenic plants conferred by antisense inhibition of a host gene essential in S-adenosylmethionine-dependent transmethylation reactions. Proc Natl Acad Sci USA 92:6117–6121CrossRefPubMedPubMedCentralGoogle Scholar
- Takeshita M, Koizumi E, Noguchi M, Sueda K, Shimura H, Ishikawa N, Matsuura H, Ohshima K, Natsuaki T, Kuwata S, Furuya N, Tsuchiya K, Masuta C (2012) Infection dynamics in viral spread and interference under the synergism between Cucumber mosaic virus and Turnip mosaic virus. Mol Plant-Microbe Interact 25:18–27CrossRefPubMedGoogle Scholar
- Verslues PE, Batelli G, Grillo S, Agius F, Kim YS, Zhu J, Agarwal M, Katiyar-Agarwal S, Zhu JK (2007) Interaction of SOS2 with nucleoside diphosphate kinase 2 and catalases reveals a point of connection between salt stress and H2O2 signaling in Arabidopsis thaliana. Mol Cell Biol 27:7771–7780CrossRefPubMedPubMedCentralGoogle Scholar
- Yang CW, Gonzalez-Lamothe R, Ewan RA, Rowland O, Yoshioka H, Shenton M, Ye H, O’Donnell E, Jones JDG, Sadanandom A (2006) The E3 ubiquitin ligase activity of Arabidopsis PLANT U-BOX17 and its functional tobacco homolog ACRE276 are required for cell death and defense. Plant Cell 18:1084–1098CrossRefPubMedPubMedCentralGoogle Scholar
- Zou JJ, Li XD, Ratnasekera D, Wang C, Liu WX, Song LF, Zhang WZ, Wu WH (2015) Arabidopsis CALCIUM-DEPENDENT PROTEIN KINASE8 and CATALASE3 function in abscisic acid-mediated signaling and H2O2 homeostasis in stomatal guard cells under drought stress. Plant Cell 27:1445–1460CrossRefPubMedPubMedCentralGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.