The rubisco small subunit is involved in tobamovirus movement and Tm-2²-mediated extreme resistance.

The multifunctional movement protein (MP) of Tomato mosaic tobamovirus (ToMV) is involved in viral cell-to-cell movement, symptom development, and resistance gene recognition. However, it remains to be elucidated how ToMV MP plays such diverse roles in plants. Here, we show that ToMV MP interacts with the Rubisco small subunit (RbCS) of Nicotiana benthamiana in vitro and in vivo. In susceptible N. benthamiana plants, silencing of NbRbCS enabled ToMV to induce necrosis in inoculated leaves, thus enhancing virus local infectivity. However, the development of systemic viral symptoms was delayed. In transgenic N. benthamiana plants harboring Tobacco mosaic virus resistance-22 (Tm-22), which mediates extreme resistance to ToMV, silencing of NbRbCS compromised Tm-22-dependent resistance. ToMV was able to establish efficient local infection but was not able to move systemically. These findings suggest that NbRbCS plays a vital role in tobamovirus movement and plant antiviral defenses.

Rubisco catalyzes the first step of CO 2 assimilation in photosynthesis and photorespiration. The Rubisco holoenzyme is a heteropolymer consisting of eight large subunits (RbCLs) and eight small subunits (RbCSs). RbCL was reported to interact with the coat protein of Potato virus Y (Feki et al., 2005). Both RbCS and RbCL were reported to interact with the P3 proteins encoded by several potyviruses, including Shallot yellow stripe virus, Onion yellow dwarf virus, Soybean mosaic virus, and Turnip mosaic virus (Lin et al., 2011). Proteomic analysis of the plant-virus interactome revealed that RbCS participates in the formation of virus complexes of Rice yellow mottle virus (Brizard et al., 2006). However, the biological function of Rubisco in viral infection remains unknown.
In this study, we show that RbCS plays an essential role in virus movement, host susceptibility, and Tm-2 2 -mediated extreme resistance in the ToMV-host plant interaction.

NbRbCS Interacts with ToMV MP
To identify ToMV MP-interacting proteins, we used ToMV MP as bait and conducted yeast (Saccharomyces cerevisiae) two-hybrid (Y2H) screens of a tomato complementary DNA (cDNA) library previously constructed in the Y2H prey vector (Liu et al., 2002b). From the screens, we obtained a number of clones that encode proteins that may interact with ToMV MP. Three of these candidates were identified as tomato RbCS cDNAs. To test whether Nicotiana benthamiana RbCS (NbRbCS) also interacts with ToMV MP, we cloned the full-length NbRbCS cDNA (N. benthamiana tentative consensus annotator no. TC23245) in the Y2H activation domain (AD) vector to produce AD::NbRbCS. Indeed, BD::ToMV MP interacted specifically with AD:: NbRbCS, as indicated by the growth of yeast on Leuplates containing Gal (Fig. 1A). However, as the controls, AD::NbRbCS and bait LexA binding domain (BD; BD vector), AD vector and BD::ToMV MP, as well as AD vector and BD vector showed no interaction (Fig. 1A).
The ToMV MP-NbRbCS interaction was further confirmed by glutathione S-transferase (GST) pull-down assays. GST-ToMV MP and NbRbCS-33FLAG-63His fusion proteins were separately expressed in Escherichia coli. The GST-ToMV MP fusion protein was purified using glutathione-Sepharose beads and then mixed with total soluble proteins of E. coli expressing the NbRbCS-33FLAG-63His fusion protein or 33FLAG-63His control. Western-blot analysis using an anti-FLAG antibody showed that only GST-ToMV MP specifically pulled down NbRbCS-33FLAG-63His, demonstrating that ToMV MP was able to bind to NbRbCS (Fig. 1B).
In vivo pull-down assays were performed to test if ToMV MP directly interacts with NbRbCS in planta. ToMV MP tagged with a Myc epitope (ToMV MP-Myc) and NbRbCS tagged with a hemagglutinin (HA) peptide (HA-NbRbCS) were expressed in plants under the control of Cauliflower mosaic virus 35S promoter. Expression of ToMV MP-Myc was confirmed by westernblot analysis with anti-Myc antibody (Fig. 2, bottom). Total proteins extracted from leaves coinfiltrated with Agrobacterium tumefaciens carrying the ToMV MP-Myc expression cassette together with A. tumefaciens carrying either HA vector or HA-NbRbCS expression cassettes were immunoprecipitated using an anti-HA antibody. The resulting precipitates were analyzed by western blot using an anti-Myc antibody. ToMV MP-Myc was only detected in the immunoprecipitations with HA-NbRbCS but not in immunoprecipitations with the control HA (Fig. 2, top), although the anti-HA antibody could pull down both HA-NbRbCS and control HA proteins (Fig.  2, middle). These experiments clearly demonstrate that ToMV MP interacts with NbRbCS in plant cells.

N-Terminal and Middle Domains of ToMV MP Are Responsible for Its Interaction with NbRbCS
TMV MP can be divided into N-terminal, middle, and C-terminal regions joined by two putative a-helical transmembrane domains (Brill et al., 2000). The N-terminal and C-terminal regions are thought to be cytoplasmic, and the middle region of MP is presumed to fold in the endoplasmic reticulum (ER) lumen in plant cells (Brill et al., 2000(Brill et al., , 2004. Moreover, the N-terminal and middle regions of MPs are conserved among tobamoviruses (Supplemental Fig. S1). According to this structural model, we generated a series of ToMV MPdeletion mutants in order to determine which amino acid domain(s) is responsible for ToMV MP binding to NbRbCS. The Y2H assays demonstrated that the N-terminal region of ToMV MP (amino acid residues 1-58) interacted with NbRbCS, as indicated by growth of yeast on Leuplates containing Gal and raffinose and blue yeast colonies on plates containing 5-bromo-4chloro-3-indolyl b-D-galactopyranoside. The middle region (amino acid residues 85-145) also interacted with NbRbCS (Fig. 3). By contrast, no interaction was observed between NbRbCS and the C-terminal fragment (amino acid residues 171-264) of ToMV MP. Surprisingly, there was also no interaction between NbRbCS and a longer N-terminal polypeptide (amino acid residues 1-171) or a longer middle amino acid section (amino acid residues 58-171). These two polypeptides contain the dual transmembrane domains, which may disturb the NbRbCS-binding activity of ToMV MP. As expected, the full-length ToMV MP protein did interact with NbRbCS (Fig. 3). These results suggest that the N-terminal and middle domains of ToMV MP are responsible for its interaction with RbCS.

Subcellular Localization of the NbRbCS-ToMV MP Interaction
The subcellular localization of the NbRbCS-ToMV MP interaction was studied by citrine yellow fluorescent protein (YFP)-based bimolecular fluorescence complementation (BiFC) assays, where YFP was split into N-terminal (nYFP) and C-terminal (cYFP) fragments for protein fusion . When cYFP-NbRbCS was transiently coexpressed with ToMV MP-nYFP in N. benthamiana, bright fluorescent signals were observed in mesophyll and epidermal cells. The YFP fluorescence was observed mainly in amorphous aggresomes in the cytoplasm (Fig. 4, A and E, arrows), while some fluorescence was also found in punctate structures that were likely to be PD in the cell walls of the epidermal cells (Fig. 4A, asterisks). Indeed, the punctate-localized YFP fluorescence from the ToMV MP-RbCS interaction coincided with the PD regions labeled by the cyan fluorescent protein (CFP)-tagged PD marker CFP-TMD (for transmembrane domain of Arabidopsis plasmodesmata-located protein1a [PDLP1a]; Thomas et al., 2008;Fig. 4, I-K). Interestingly, the subcellular localization of the NbRbCS-ToMV MP interaction was similar to that of ToMV MP itself. ToMV MP-YFP accumulated primarily in PD along cell walls and partially in cytosolic inclusion bodies (referred to as virus replication complexes [VRCs]; Supplemental Fig. S2), as described previously (Kahn et al., 1998;Ashby et al., 2006). RbCS is a nucleus-encoded chloroplast protein (Chua and Schmidt, 1978;Smith and Ellis, 1979); however, no signal for NbRbCS-ToMV MP interaction was detected in chloroplasts ( Fig. 4G). Moreover, no interaction signals were observed for cYFP-NbRbCS with nYFP-ToMV P50 (Fig. 4, B and F) or for various control combinations ( Fig. 4, C, D, G, and H). These results demonstrate that the interaction of ToMV MP with NbRbCS may occur at PD and VRCs.

Silencing of NbRbCS Enables ToMV to Induce Local Necrosis But Reduces Virus Systemic Movement
To examine the biological relevance of the NbRbCS-ToMV MP interaction to ToMV infection of plants, we used the Tobacco rattle virus (TRV) virus-induced gene silencing (VIGS) vector (Liu et al., 2002a) to silence NbRbCS in N. benthamiana. Western-blot assays showed that silencing of NbRbCS dramatically reduced NbRbCS protein levels (Fig. 5A). We also noted that silencing of NbRbCS caused leaf chlorosis and upward leaf curling  (Supplemental Fig. S4A, top row) did not induce any necrosis in the NbRbCS-silenced plants, suggesting that ToMV specifically induced necrosis. Moreover, necrosis spread out along proximate veins and resulted in the collapse of large leaf patches or even the entire leaf lamina. However, no necrosis was observed in NbRbCSnonsilenced TRV control plants (Fig. 6, bottom row). In addition, NbRbCS silencing increased the number of TMV-GFP infection foci in the inoculated leaves and the percentage of total infected leaf area (Supplemental Fig.  S4), although it reduced the size of a single TMV-GFP focus (Supplemental Fig. S4). These results suggest that silencing of NbRbCS may enhance host susceptibility to tobamoviruses.
To follow up on the local infection results, we next examined the effect of NbRbCS silencing on systemic trafficking of ToMV. Viral systemic symptom development can be used as an indicator to monitor long-distance trafficking of ToMV and other plant viruses (Leisner et al., 1993;Li et al., 2005;Lewis and Lazarowitz, 2010). In N. benthamiana, ToMV systemic infection resulted in curved apical shoots, crinkled apical leaves, chlorotic mottling of newly emerged leaves, and backward winding of young leaves (Fig. 7A, right  column). Surprisingly, we found that NbRbCS silencing delayed the appearance of ToMV systemic symptoms. Systemic symptoms were first observed in approximately 5% of NbRbCS-silenced plants at 5.5 d post infection (dpi), but more than 40% of nonsilenced TRV control plants had systemic ToMV infection at 5.5 dpi (Fig. 7, C and D). Furthermore, ToMV RNA can be detected by reverse transcription (RT)-PCR at 5 dpi in the upper uninoculated leaves of TRV control plants with viral symptoms but not in the upper uninoculated leaves of NbRbCS-silenced plants without viral symptoms (Fig. 7B), supporting the idea that the viral symptoms can be used as an indicator of the presence of ToMV. Systemic symptoms appeared at 7 dpi in almost 80% of nonsilenced control plants compared with approximately 45% of NbRbCS-silenced plants. Furthermore, more than 20% of NbRbCS-silenced plants remained symptomless, but all NbRbCS nonsilencing plants became systemically infected with ToMV at 10 dpi (Fig. 7C). The general tendency was that systemic infection occurred 1 to 2 d later in NbRbCS-silenced plants than in nonsilenced plants (Fig. 7D).
Taken together, these findings suggest that silencing of NbRbCS enhances the local susceptibility of N. benthamiana to ToMV infection but reduces ToMV long-distance movement.
Silencing of NbRbCS Compromised Tm-2 2 -Mediated Extreme Resistance against ToMV Enhanced host susceptibility to viruses in NbRbCSsilenced plants suggests that NbRbCS may have a role in plant antiviral defense. To test this hypothesis, we used RT-PCR to examine the effect of silencing NbRbCS on the expression of the defense-related PR1a gene. As seen in Supplemental Figure S5, PR1a RNA levels were reduced in the NbRbCS-silenced plants compared with the nonsilenced TRV control plants agroinfiltrated with TRV empty VIGS vector. However, the level of 18S rRNA was not affected by NbRbCS gene expression. These results suggest that NbRbCS may participate in antiviral defenses in plants.
To further test this idea, we investigated whether the ToMV MP-NbRbCS interaction has a role in Tm-2 2 gene-mediated resistance against ToMV infection. This experiment was performed in transgenic Tm-2 2 N. benthamiana line T1-4, which is extremely resistant to both ToMV and TMV (J. Zhao and Y. Liu, unpublished data; Supplemental Materials and Methods S1). It should be noted that ToMV MP is the corresponding avirulence protein that is recognized by the Tm-2 2 gene product. As expected, the NbRbCS-nonsilenced T1-4 plants were extremely resistant to ToMV infection, and no hypersensitive response (HR) was observed in the inoculated leaves (Fig. 8A, bottom row). However, ToMV induced HR necrosis in the inoculated leaves of NbRbCS-silenced T1-4 plants at 4 to 5 dpi (Fig. 8A, top row, arrows). In addition, the noninfective plant sap control did not induce any necrosis in the NbRbCSsilenced T1-4 plants (Supplemental Fig. S3B). Furthermore, ToMV RNA was readily detectable by RT-PCR at 4 dpi in the inoculated leaves of NbRbCS-silenced plants but not in nonsilenced T1-4 plants (Fig. 8B). Moreover, we consistently failed to detect any ToMV RNA in the upper noninoculated leaves of both NbRbCS-silenced and control plants even at 6 weeks post ToMV infection (data not shown). These data suggest that NbRbCS is required for Tm-2 2 -mediated local extreme resistance.

RbCS May Interact with MPs from Many Tobamoviruses
Through a range of in vitro and in vivo proteinprotein interaction assays, we have identified NbRbCS as a new ToMV MP-interacting host protein. NbRbCS interacted with the N-terminal and middle domains of ToMV MP. These domains are conserved among tobamovirus MPs, suggesting that RbCS may commonly interact with MPs from tobamoviruses. Indeed, we found that NbRbCS associated and interacted with TMV MP in planta (Supplemental Figs. S6 and S7; Supplemental Results S1). RbCS is translated in the cytoplasm and then imported into chloroplasts (Keegstra et al., 1995;Schatz and Dobberstein, 1996). However, we found that NbRbCS interacts with ToMV and TMV MPs in the cytoplasm but not in the chloroplast. It is possible that both ToMV and TMV MPs may form a protein complex with NbRbCS and, consequently, hijack this protein prior to its entry into the chloroplast to facilitate virus infection of plants.

Role of RbCS in Host Susceptibility to Virus Infection and in Plant Defense
RbCS is expressed in mesophyll cells but not in nonphotosynthetic epidermal cells (Lu et al., 2002). Removal of the lower RbCS-free epidermis from plant leaves Figure 4. In vivo subcellular localization of the NbRbCS-ToMV MP interaction by BiFC assay. A to H, NbRbCS and control N-terminal luciferase (Nluc) were tagged at the N terminus with cYFP (cYFP-NbRbCS and cYFP-nLUC) and were coexpressed transiently with ToMV MP or control ToMV P50 tagged at the C terminus with nYFP (ToMV MP-nYFP and ToMV P50-nYFP) in N. benthamiana leaves, and any fluorescence was observed in epidermal (A-D) and mesophyll (E-H) cells by confocal microscopy at 2 dpi. The NbRbCS interaction with ToMV MP was observed in cytoplasmic aggregates (indicated with arrows) and in discrete punctate sites (indicated with asterisks) along the cell edges. I to K, BiFC constructs and the PD marker CFP-tagged TMD (CFP-TMD) were coexpressed in N. benthamiana leaf epidermal cells. YFP channel (I), CFP channel (J), and merged images (K) show that good colocalization existed between the subcellular localization of the ToMV MP-NbRbCS interaction (YFP) and CFP-TMD in discrete punctate sites along the cell border. Green fluorescence shows sites with the two YFP and CFP fluorescence merged. Arrows indicate the positions of overlain YFP and CFP fluorescence. Bars = 20 mm. (Wieringabrants, 1981), suggesting that RbSC may have a negative impact on these viruses. Consistent with this hypothesis, NbRbCS silencing enabled ToMV to establish severe local infections and produce necrotic lesions in inoculated leaves. Furthermore, silencing of NbRbCS caused a significant increase in the numbers of local TMV infection foci (Supplemental Fig. S4B), demonstrating that RbCS negatively affects host susceptibility to tobamoviruses. This conclusion is further supported by the finding that NbRbCS silencing reduced defenserelated gene expression and compromised Tm-2 2 genemediated local extreme resistance against ToMV (Fig. 8) and TMV-GFP (Supplemental Fig. S8; Supplemental Results S1). Strikingly, NbRbCS silencing only compromised Tm-2 2 gene-mediated local extreme resistance but not systemic resistance against ToMV.

reduces infection by TMV or Tobacco necrosis virus
It is possible that silencing of NbRbCS completely disrupted Tm-2 2 gene-mediated systemic resistance against ToMV, but the virus cannot move systemically, because the systemic movement is prevented by the lack of NbRbCS. However, we cannot rule out the possibility that Tm-2 2 gene-mediated local and systemic resistance can be functionally separated.
In several cases, resistance gene-mediated HR and virus resistance have been found to require light (Chandra-Shekara et al., 2006), and the expression of Rubisco genes, including RbSC, is light dependent. It is well known that dark-treated plants are more susceptible to TMV infection and that light treatments decrease plant susceptibility to TMV infection (Helms and McIntyre, 1967). These findings suggest that RbCS may play a role in plant defense. In addition, NbRbCS silencing could have a general effect on plant metabolism, and this effect may reduce plant defenses and further cause an enhanced susceptibility to viral infections. It is possible that healthy plants have sufficient levels of RbCS and are thus less susceptible to infection by tobamoviruses. However, ToMV MP may interact with RbCS and disturb its function to optimize cellular environments that are suitable for the establishment of ToMV infection. On the other hand, virus infection could be effectively established in cells in which levels of RbCS are reduced as in NbRbCS-silenced plants.
Besides RbCS, the 33-kD subunit of the oxygenevolving complex (OEC) and Rab GDP dissociation inhibitor (GDI) are involved in TMV infection (Abbink et al., 2002;Kramer et al., 2011). Both the OEC 33-kD subunit and GDI interact with TMV 126-kD replicase and negatively regulate host susceptibility. The OEC 33-kD subunit was suggested to have a general role in defense, and GDI was suggested to participate in vesicle trafficking to enhance the establishment of an infection (Abbink et al., 2002). TMV MP was also suggested to function as an RNA silencing enhancer to influence host cell gene expression and thus to enhance  host susceptibility to TMV infection (Vogler et al., 2008;Amari et al., 2012). Whether RbCS works together with the OEC 33-kD subunit and/or GDI to affect virus infection and whether RbCS participates in RNA silencing need to be further investigated.

Role of RbCS in Virus Movement
Both RbCS and RbCL were reported to interact with the P3 proteins encoded by different potyviruses (Lin et al., 2011). Furthermore, potyvirus P3 localizes to PD and acts in intercellular virus movement (Wei et al., 2010;Wen and Hajimorad, 2010). These reports are consistent with our finding that NbRbCS interacts with MPs of ToMV and TMV, which are involved in viral cell-to-cell movement. Indeed, silencing of NbRbCS reduced the size of TMV infection foci (Supplemental Fig. S4, A and D; Supplemental Results S1), suggesting that MP-binding NbRbCS participates in cell-to-cell movement of TMV. Viral infection induces the formation of irregularly shaped cytoplasmic inclusion bodies, termed X-bodies, viroplasms, or VRCs in plants Más and Beachy, 1999;Szecsi et al., 1999). TMV-induced VRCs are ER-derived aggregates that contain viral RNA, 126-and 183-kD replicase, MP, Coat Protein (CP), and host factors (Más and Beachy, 1999). MP-binding NbRbCS seems to be one of these host factors, since we found that the NbRbCS interaction with MPs of ToMV and TMV mainly localized to irregularly shaped cytoplasmic inclusion bodies ( Fig. 4; Supplemental Fig. S7) and NbRbCS could only form irregularly shaped cytoplasmic inclusion bodies when ToMV MP was coexpressed (Supplemental Fig. S9). Although the formation of inclusion bodies from ER in infected cells might be dispensable for viral replication, VRCs may exist to regulate TMV RNA movement efficiency (Boyko et al., 2000;Waigmann et al., 2007). NbRbCS may participate in viral cell-to-cell movement by regulating the function of MP within VRCs.
We found that silencing of NbRbCS reduces the systemic movement of ToMV and TMV ( Fig. 7; Supplemental Fig. S10; Supplemental Results S1). This reduction may be caused by reduced virus cell-to-cell movement in NbRbCS-silenced plants, because silencing of NbRbCS reduces the size of TMV infection foci (Supplemental Fig. S4, A and D). However, there is also a possibility that NbRbCS is involved in ToMV systemic movement through the change in flow of photoassimilates in NbRbCS-silenced plants. It was proposed that plant viruses, paralleling the flow of photoassimilates, spread systemically through the Whiskers represent 5% to 95% intervals of each grouped data set, and crosses represent maximal and minimal observations, respectively. vascular system (Wintermantel et al., 1997;Susi et al., 1999). Silencing of NbRbCS may cause a change in the flow of photoassimilates because of leaf chlorosis and reduced photosynthesis. Consistent with this hypothesis, TMV MP does affect carbohydrate metabolism and photoassimilate translocation during TMV infection. Transgenic overexpression of TMV MP resulted in palegreen leaves, elevated carbohydrate accumulation, and reduced rates of photoassimilate export in source leaves (Lucas et al., 1993;Olesinski et al., 1995Olesinski et al., , 1996Almon et al., 1997). Furthermore, TMV MP induced alterations in carbon partitioning in the mesophyll and is involved in regulating photosynthesis. However, this process is independent of change in the PD size-exclusion limit (Balachandran et al., 1995;Olesinski et al., 1995). TMV MP may also attenuate photosynthesis and carbon metabolism by affecting certain mesophyllic factors (Wolf and Millatiner, 2000). Along with a decrease of Rubisco during TMV infection (Itaya et al., 2002) and tobamovirus MP-RbCS physical interactions, it is possible that RbCS may act as a potential node linking MP with photoassimilate allocation and photosynthesis. Thus, reduction of functional RbCS levels may account for the inability of tobamovirus to traffic over long distances.

Y2H Screen and Interaction Assays
The full-length ToMV MP gene was PCR amplified and cloned into pYL302, in which the nuclear localization signal sequence of the LexA DNA BD is under the control of a GAL10 promoter (Y. Liu and S.P. Dinesh-Kumar, unpublished data), to generate the LexA DNA BD containing bait vector BD-ToMV MP. The full-length NbRbCS gene was PCR amplified and cloned into pJG4-5 (Kolonin et al., 2000) to produce the B42 DNA AD-containing prey vector AD-NbRbCS. The Y2H prey library containing tomato (Solanum lycopersicum) cDNA (Liu et al., 2002b) was used to screen ToMV MP-binding proteins. The Y2H screen and interaction assays were performed as described (Liu et al., 2002b).

Transient Expression by Agroinfiltration and VIGS
All plants were grown in pots at 22°C in a growth chamber under a 16-hlight/8-h-dark cycle with 120 mmol m -2 s -1 white light illumination. For Agrobacterium tumefaciens-mediated transient expression studies, GV2260 strains containing the relevant expression vector were grown overnight, pelleted, resuspended in infiltration buffer (10 mM MgCl 2 , 10 mM MES, and 200 mM acetosyringone, pH 5.6), and kept at room temperature for 4 h.
A. tumefaciens cultures (optical density at 600 nm = 1.0) were infiltrated into N. benthamiana leaves. The infiltrated leaves were detached at 60 to 72 h post infiltration for the corresponding assays. For coexpression, equal amounts of A. tumefaciens cultures were mixed and used for infiltration. VIGS was performed as described (Liu et al., 2002a).

RNA Extraction, RT, and RT-PCR
Total RNA was extracted using TRIzol reagent (Tiangen) according to the manufacturer's protocol. First-strand cDNA was synthesized from 2 mg of RNA using a 21-nucleotide [oligo(dT) plus two anchoring nucleotides] or gene-specific primer and Moloney murine leukemia virus reverse transcriptase (Tiangen). DNA fragments corresponding to 18S rRNA, Actin, NbRbCS, ToMV MP, CP, and GFP were PCR amplified using the respective primers. The sequences of the primers used in PCR are available upon request.

Monitoring Protein Expression Levels and in Vivo Pull-Down Assays
Total proteins from N. benthamiana leaves were extracted in buffer containing 150 mM Tris-HCl, 150 mM NaCl, 25 mM sodium fluoride, 0.1% Triton X-100, pH 7.5, 2 mM dithiothreitol, 0.5 mM phenylmethanesulfonyl fluoride (PMSF), and plant protease inhibitor cocktail (Sigma-Aldrich) as described (Serino et al., 1999). Protein extracts were separated by 10% SDS-PAGE and transferred to Immobilon-P polyvinylidene difluoride membranes (Millipore) for western-blot analysis using anti-RbCS, anti-HA, or anti-Myc (Santa Cruz) primary antibody. For in vivo pull-down assays, protein extracts were incubated with the anti-HA antibody (1:200 diluted) for 4 h at 4°C following overnight incubation with protein A/G plus agarose beads (AbMART) equilibrated with the extraction buffer. The beads were washed three times with the extraction buffer. Immunoprecipitated samples were used for westernblot analysis as described above using the anti-Myc primary antibody.

BiFC and Fluorescence Microscopy
Citrine YFP-based BiFC was performed as described . Live plant imaging was performed on a Zeiss LSM710 confocal microscope. Enhanced GFP-or citrine YFP-derived fluorescence was acquired using excitation 488-nm laser and emission 493-to 598-nm filters and excitation 514-nm laser and emission 519-to 587-nm filters, respectively. CFP fluorescence was excited with a 405-nm laser line, and the emissions were captured at 430 to 460 nm. For colocalization assays, two fluorescent channels were separately monitored in two tracks, and signals were pseudocolored in digital images as described in the figure legends. Twelve-bit confocal images were acquired with an EC Plan-Neofluar 103/0.30 M27 objective for 103 magnification and a Plan-Apochromat 403/0.95 Korr M27 objective for 403 magnification. Images were analyzed with ZEN 2009 Light Edition and Image-Pro Plus 6.0 software.

Infection and Statistical Analysis of ToMV
Plants were rub inoculated with ToMV. Viral spread and symptoms were monitored for at least 2 weeks. All experiments were performed at least three times, and at least three plants were used per construct each time. OriginPro 8.1 and SPSS 19.0 were used to conduct statistical analyses and to generate graphs.

Supplemental Data
The following materials are available in the online version of this article.
Supplemental Figure S1. Amino acid sequence alignment of tobamovirus MPs.
Supplemental Figure S3. Mock inoculation does not induce local necrosis in NbRbCS-silenced plants.
Supplemental Figure S4. Silencing of NbRbCS enhances local multiplication but reduces cell-to-cell movement of TMV-GFP.
Supplemental Figure S6. TMV MP interacts with NbRbCS in plants.
Supplemental Figure S7. BiFC visualization of subcellular localization of the TMV MP-NbRbCS interaction.
Supplemental Figure S9. Coexpression of ToMV MP resulted in the formation of NbRbCS in cytosolic aggresomes and punctae along the cell wall.
Supplemental Materials and Methods S1.

ACKNOWLEDGMENTS
We thank Jian-Min Zhou for providing the luciferase complementation imaging assay vectors.