Palmitoylation Is Indispensable for Remorin to Restrict Tobacco Mosaic Virus Cell-to-Cell Movement in Nicotiana benthamiana

Remorin (REM) is a plant-specific plasma membrane-associated protein regulating plasmodesmata plasticity and restricting viral cell-to-cell movement. Here, we show that palmitoylation is broadly present in group 1 remorin proteins in Nicotiana benthamiana and is crucial for plasma membrane localization and accumulation. By screening the four members of N. benthamiana group 1 remorin proteins, we found that only NbREM1.5 could significantly hamper tobacco mosaic virus (TMV) cell-to-cell movement. We further showed that NbREM1.5 interacts with the movement protein of TMV in vivo and interferes with its function of expanding the plasmodesmata size exclusion limit. We also demonstrated that palmitoylation is indispensable for NbREM1.5 to hamper plasmodesmata permeability and inhibit TMV cell-to-cell movement.


Introduction
Plasmodesmata (PD) are plant-specific plastic membranous nanopores connecting adjacent plant cells [1,2]. PD are communication channels for the intercellular transport of proteins and molecules, and they are utilized by plant viruses to establish intercellular trafficking [3,4]. The efficiency of cell-to-cell communication and transportation through PD is dependent on the PD size exclusion limit (SEL), which is strictly controlled by callose deposition in the PD neck region [5]. As obligate intracellular pathogens, plant viruses encode movement proteins (MPs) to expand PD permeability and promote viral cell-to-cell movement [6]. Several plant proteins are also found capable of controlling PD SEL, of which, the remorin family is one of the best-characterized plant plasma membrane-anchored proteins regulating plasmodesmata aperture and functionality [7][8][9]. As nanodomainorganized proteins located at the plasma membrane and PD, remorin proteins are found to be broadly involved in plant responses to biotic stress, including viruses, bacteria, fungi, oomycetes, etc. [10].
Although remorin proteins are specifically associated with the inner leaflet of the plasma membrane, they lack a transmembrane domain. The molecular mechanism for remorin proteins to specifically anchor at the membrane remains unclear [10]. The conserved remorin C-terminal anchor region (REM-CA) of potato (Solanum tuberosum) remorin group 1 isoform 3 (StREM1.3) is considered to be a strictly membrane-anchoring motif, by folding into an alpha helix and inserting itself into the hydrophobic core of the bilayer [11][12][13]. For most Arabidopsis remorin proteins, palmitoylation of cysteine residues in the C-terminal region has been shown to predominantly contribute to their membrane association but it is not a key determinant factor for their membrane nanodomain localization [14].
Palmitoylation, inter-changeably used with the term "S-acylation", is a type of lipid post-translational modification. Palmitoylation is the reversible addition of a long-chain

Plant Growth Conditions
The N. benthamiana plants were grown in the growth chamber at 25 • C under a 16-h light/8-h dark photoperiod with 60% relative humidity.

Agroinfiltration in N. benthamiana
For agroinfiltration, vectors were individually transformed into A. tumefaciens strain EHA105 cells, and then the transformed A. tumefaciens cells were grown individually in YEP medium containing 50 mg/L kanamycin and 50 mg/L rifampicin at 28 • C until an OD 600 of 0.6 was reached. The cultures were collected and re-suspended in the inoculation buffer (10 mM MgCl 2 , 100 mM MES (pH 5.7), 2 mM acetosyringone) at room temperature. The suspensions were adjusted to an OD 600 of 0.5 and then used for agroinfiltration in five/seven-leaf N. benthamiana plants.

Multiple Sequence Alignment and Phylogenetic Analysis
Multiple sequence alignment was analyzed through the Clustal W method ( Figure S1) [20]. Phylogenetic analysis was performed by MEGA X [21]. The evolutionary history was inferred by using the maximum likelihood method and an JTT matrix-based model [21]. The tree was tested via bootstrap analysis (1000 replicates) and was drawn to scale ( Figure S2).

Virus Inoculation and Detection
The TMV-GFP infectious clone (30B-GFPC3) [22] was infiltrated into N. benthamiana leaves by A. tumefaciens infiltration; infection fluorescence spots were observed and photographed under a portable UV lamp at 4 days post-infiltration (dpi) on the inoculated leaves. TMV-GFP infection spot areas were calculated by Image Pro Plus 6.0 software (Media Cybernetics, Silver Spring, MD, USA) and were analyzed by GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA, USA). Experiments were repeated at least three times. Statistical parameters SEM (standard error) are indicated in figure legends. Asterisks mark statistical significance (* 0.01 < p < 0.05; ** p < 0.01; *** p < 0.001; ns, p > 0.05, not significant) according to two-tailed Student's t-test.

Acyl-Resin Assisted Capture (Acyl-RAC)
The experiment was performed according to the method previously reported [23]. Briefly, N. benthamiana leaves that have expressed proteins of interest by agroinfiltration were harvested at 48 h post infiltration (hpi). Total proteins were extracted with a lysis buffer (25 mM HEPES, 25 mM NaCl, 1 mM EDTA, PH 7.5) containing a protease inhibitor cocktail (Roche, Rotkreuz, Switzerland). Total protein extraction was diluted with one volume of blocking buffer (100 mM HEPES, 2.5% SDS, 1 mM EDTA, 0.5% MMTS, PH 7.5), followed by incubating at 40 • C for 30 min to block free thiols. Then, three volumes of cold acetone were added, and samples were allowed to be precipitated at −20 • C for 20 min. The precipitated proteins were collected by centrifugation at 5000× g for 10 min. The pellet was then washed with 70% cold acetone, and finally re-suspended with a binding buffer (100 mM HEPES, 1% SDS, 1 mM EDTA, PH 7.5). The samples were split into two equal parts followed by adding 60 µL of pre-washed thiopropyl Sepharose 6b resins (Sigma, St. Louis, MO, USA). Then, 40 µL of 2 M neutral hydroxylamine (NH 2 OH) (pH 7.5) or 2 M NaCl (used as a negative control) were added, respectively, into two equal parts of the mix and processed in parallel. A total of 100 µL of each mix was saved as the total input (input). The mixtures were rotated at room temperature for 2 h then centrifuged at 5000× g for 1 min; the pellets containing thiopropyl Sepharose 6b resins were washed 5 times using the binding buffer, and finally, the resins were suspended in 100 µL binding buffer containing 50 mM dithiothreitol (DTT). A total of 2 × SDS-PAGE loading buffer was added to the samples and the samples were heated to 100 • C for 10 min, followed by immunoblotting detection.

Bimolecular Fluorescence Complementation (BIFC) and Subcellular Localization Assays
In BiFC and subcellular localization assays, all the plasmids were transformed into the A. tumefaciens strain EHA105. The N. benthamiana leaf epidermal cells were assayed for fluorescence using a Zeiss LSM 880 or 980 Airyscan TM upright laser scanning confocal microscope (Carl Zeiss Meditec AG, Jena, Germany) at 48 hpi. Subcellular localization assays were performed in 5-7-leaf-old RFP-histone 2B (RFP-H2B) transgenic N. benthamiana leaves [25]. The GFP fluorophore was excited with 488 nm laser lines and the emission was detected at 490-540 nm. The RFP fluorophore was excited with 561 nm laser lines and emission was detected at 588-648 nm. For subcellular assay in protoplast, the N. benthamiana leaves that have expressed proteins of interest by agroinfiltration at 48 hpi were cut into small pieces (2 mm × 2 mm) followed by incubation at 28 • C in enzyme solution [1% cellulose (Yakult, Tokyo, Japan), 0.5% BSA (Sigma, St. Louis, MO, USA), 1% Driselase (Sigma, St. Louis, MO, USA) and 0.55 M mannitol (pH 5.9)] for 2 h without shaking. The released protoplasts were collected and examined by confocal microscopy. The Z-stacks of optical sections were constructed to view the localization in the protoplast using ZEN Black software (Carl Zeiss Meditec AG, Jena, Germany). In the BiFC assay, NbREM1.1 or NbREM1.5 was fused into a pCAMBIA1300-35S::MCS::nYFP vector carrying the N-terminal 159 amino acids of the YFP by T4 DNA ligase (ThermoFisher, Waltham, MA, USA) generating NbREM1.1-nYFP or NbREM1.5-nYFP. NSvc4 or TMV-MP was fused into a pCAMBIA1300-35S::MCS::cYFP vector carrying the C-terminal 80 amino acids of YFP by T4 DNA ligase (ThermoFisher, Waltham, MA, USA), generating NSvc4-cYFP or TMV-MP-cYFP. Co-infiltration of Agrobacterium strains containing the BiFC constructs that were used to test the interaction was carried out at OD 600 of 0.5:0.5. The NbREM1.1-nYFP and NSvc4-cYFP interaction was used as a positive control; the NbREM1.1-nYFP and TMV-MP-cYFP interaction was used as a negative control.

Callose Deposition Assay
Callose was stained using aniline blue fluorochrome (Biosupplies, Bundoora, Victoria, Australia) according to the user manual, with minor modification. Briefly, the N. benthamiana leaves that were used for aniline blue staining were cut into small pieces (4 mm × 4 mm) followed by being vacuumed for 10 min in 0.1% aniline blue fluorochrome. Leaf pieces were then observed by using LSM 880 or 980 (Carl Zeiss) and excited at 405 nm. The callose intensity was measured by using Image Pro Plus 6.0 software (Media Cybernetics, Silver Spring, MD, USA).

Quantification and Statistical Analysis
All experiments involving measurements, quantifications, and imaging were repeated at least three times. GraphPad Prism version 8.0 (GraphPad Software, San Diego, CA, USA) was used for data plotting and statistical tests. Statistical parameters such as mean ± SD (standard deviation), SEM (standard error), or 95% confidence intervals are indicated in figure legends. In graphs, asterisks mark statistical significance (* 0.01 < p < 0.05; ** p < 0.01; *** p < 0.001; ns, p > 0.05, not significant) according to two-tailed Student's t-test.
NbREM1.1 has been proven to be palmitoylated [8], so we used it as a positive control for palmitoylation. In acyl-RAC assays, after the blocking of free thiols in total protein extraction with methyl methanethiosulfonate (NMT) and the cleaving of thioester linkages with hydroxylamine (NH 2 OH), palmitoylated NbREM1.1, NbREM1.3, NbREM1.5, or NbREM1.8 was captured by thiopropyl Sepharose in hydroxylamine-treated samples and detected by immunoblotting (Figure 1a-d,j). In contrast, no remarkable palmitoylation was detected in the NbREM1.1-C206A, NbREM1.3-C177A, and NbREM1.8-C195A point mutant proteins, in which the cysteine residues at the predicted palmitoylation sites were substituted by alanine residues (Figure 1e,f,i,j), indicating that the Cys206, Cys177, or Cys195 residue in the NbREM1.1, NbREM1.3, or NbREM1.8 protein is the essential target residue for palmitoylation, respectively. While in NbREM1.5, single-site mutation of Cys172 in NbREM1.5 could not completely abolish its palmitoylation (Figure 1g,j), but palmitoylation of NbREM1.5 was remarkably reduced when both Cys172 and Cys175 were simultaneously mutated (Figure 1h,j), indicating that NbREM1.5 was palmitoylated at both the Cys172 and Cys175 residues, in line with the results of the GPS-Palm prediction that NbREM1.5 has two predicted palmitoylation cysteines.  Data are mean ± SD (n = 3). Immunoblotting was performed using anti-flag antibody. NH 2 OH indicates presence (+) or absence (−) of hydroxylamine required for acyl group cleavage during the thiopropyl Sepharose 6b capture step. Thiopropyl Sepharose 6b enriched proteins were eluted and represent palmitoylated proteins (palmitoylation). Prior to thiopropyl Sepharose 6b capture the samples were removed as an input loading control (input).

Palmitoylation Contributes to Remorin Protein Accumulation
To test whether palmitoylation broadly contributes to remorin protein turnover, we compared the expression level of NbREM1.1, NbREM1.3, NbREM1.5, and NbREM1.8 with their respective palmitoylation-deficient mutants. The A. tumefaciens expressing the flag epitope-tagged NbREMs and their palmitoylation-deficient mutants were separately infiltrated into the opposite halves of the N. benthamiana leaves. Immunoblotting showed that accumulation levels of NbREM1.1-C206A, NbREM1.3-C177A, and NbREM1.8-C195A were significantly lower than those of their wild-type proteins (Figure 2a-c). However, for NbREM1.5, the single-site mutant in Cys172 was insufficient to affect NbREM1.5 protein accumulation (Figure 2d), whereas its double site mutant in Cys172 and Cys175 accumulated to a remarkably lower level, compared to wild-type NbREM1.5 (Figure 2e). These results indicate that palmitoylation is broadly required for remorin protein accumulation. represent palmitoylated proteins (palmitoylation). Prior to thiopropyl Sepharose 6b capture the samples were removed as an input loading control (input).

Palmitoylation Contributes to Remorin Protein Accumulation
To test whether palmitoylation broadly contributes to remorin protein turnover, we compared the expression level of NbREM1.1, NbREM1.3, NbREM1.5, and NbREM1.8 with their respective palmitoylation-deficient mutants. The A. tumefaciens expressing the flag epitope-tagged NbREMs and their palmitoylation-deficient mutants were separately infiltrated into the opposite halves of the N. benthamiana leaves. Immunoblotting showed that accumulation levels of NbREM1.1-C206A, NbREM1.3-C177A, and NbREM1.8-C195A were significantly lower than those of their wild-type proteins ( Figure  2a-c). However, for NbREM1.5, the single-site mutant in Cys172 was insufficient to affect NbREM1.5 protein accumulation (Figure 2d), whereas its double site mutant in Cys172 and Cys175 accumulated to a remarkably lower level, compared to wild-type NbREM1.5 (Figure 2e). These results indicate that palmitoylation is broadly required for remorin protein accumulation. NbREM1.5-C172A (d), NbREM1.5 and NbREM1.5-C172/175A (e) in N. benthamiana leaves. Data are mean ± SD (n = 4). Asterisks mark significant differences according to two-tailed Student's t-test; * p < 0.05; ** p < 0.01; ns, no significant difference. The opposite halves of N. benthamiana leaves separately expressed flag-tagged NbREMs or their palmitoylation-defective mutants by agroinfiltration. Immunoblotting with anti-actin antibody was used as a loading control. Immunoblotting with anti-flag antibody was used to detect protein accumulation of NbREMs and its palmitoylationdefective mutants.

Palmitoylation Contributes to Plasma Membrane Localization for Remorin Proteins
To test whether palmitoylation contributes to plasma membrane localization for NbREMs, wild-type GFP-NbREMs and their palmitoylation-deficient mutants were transiently expressed in N. benthamiana protoplast or epidermal cells. Confocal microscopy indicates wild-type GFP-NbREMs localized at the plasma membrane ( Figure 3). In contrast, the plasma membrane localization of all the four NbREMs palmitoylation-deficient mutants was remarkably weakened, and their fluorescent signals were observed in nucleus and cytoplasm (Figure 3 and Figure S4). For NbREM1.5, both the single-site mutant in Cys172 or double-site mutant in Cys172 and Cys175 remarkably weakened its membrane localization, suggesting membrane localization of NbREM1.5 is more sensitive to palmitoylation deficiency (Figure 3d lower row and Figure S4c). These results demonstrated that palmitoylation is strictly required for the plasma membrane localization. Data are mean ± SD (n = 4). Asterisks mark significant differences according to two-tailed Student's t-test; * p < 0.05; ** p < 0.01; ns, no significant difference. The opposite halves of N. benthamiana leaves separately expressed flag-tagged NbREMs or their palmitoylation-defective mutants by agro-infiltration. Immunoblotting with anti-actin antibody was used as a loading control. Immunoblotting with anti-flag antibody was used to detect protein accumulation of NbREMs and its palmitoylation-defective mutants.

Palmitoylation Contributes to Plasma Membrane Localization for Remorin Proteins
To test whether palmitoylation contributes to plasma membrane localization for NbREMs, wild-type GFP-NbREMs and their palmitoylation-deficient mutants were transiently expressed in N. benthamiana protoplast or epidermal cells. Confocal microscopy indicates wild-type GFP-NbREMs localized at the plasma membrane ( Figure 3). In contrast, the plasma membrane localization of all the four NbREMs palmitoylation-deficient mutants was remarkably weakened, and their fluorescent signals were observed in nucleus and cytoplasm (Figures 3 and S4). For NbREM1.5, both the single-site mutant in Cys172 or double-site mutant in Cys172 and Cys175 remarkably weakened its membrane localization, suggesting membrane localization of NbREM1.5 is more sensitive to palmitoylation deficiency (Figure 3d lower row and Figure S4c). These results demonstrated that palmitoylation is strictly required for the plasma membrane localization.

NbREM1.5 Interacts with TMV-MP and Interferes with Its Ability for Expanding Plasmodesmata Size Exclusion Limit
BiFC assays were used to test whether NbREM1.5 could interact with the TMV-MP in vivo. Strong plasma membrane-localized yellow fluorescence was observed upon coexpression of NbREM1.5-nYFP with TMV-MP-cYFP, whereas no fluorescence was observed in the combinations of NbREM1.5-nYFP with RSV-NSvc4-cYFP, or NbREM1.1-nYFP with TMV-MP-cYFP (Figure 5a). In vivo interaction between NbREM1.1 and RSV-NSvc4 was found in our previous study [8] and was used as a positive control; the interaction between NbREM1.1-nYFP and TMV-MP-cYFP was used as a negative control (Figure 5a). These results indicate that NbREM1.5 specifically interacts with TMV-MP in vivo.
TMV-MP is known to expand the plasmodesmata size exclusion limit (SEL) [28]; we then tested whether NbREM1.5 could affect the ability of TMV-MP to expand the plasmodesmata size exclusion limit. Highly diluted Agrobacterium (1000×) expressing GFP was co-infiltrated with Agrobacterium expressing TMV-MP and the tomato bushy stunt virus encoded gene silencing suppressor P19, and then the diffusion of GFP from an individual expressing cell to adjacent cells through plasmodesmata was monitored at 3 dpi. P19 can enhance the transient expression of proteins by agroinfiltration [29]. We found that overexpression of NbREM1.5 together with TMV-MP significantly restricted the GFP diffusion to neighboring cells, as compared with the overexpression of GUS-flag with TMV-MP, suggesting that NbREM1.5 directly affects the ability of TMV-MP to expand the plasmodesmata size exclusion limit (Figure 5b,c).
The callose deposition assay by using aniline blue staining showed that callose deposition at the plasmodesmata was significantly increased in N. benthamiana epidermal cells after overexpression of Flag-NbREM1.5 by agroinfiltration, compared with overexpression of GUS as a control ( Figure S6). These results indicate that NbREM1.5 restricts TMV cellto-cell movement by interacting with TMV-MP in vivo and interfering with its ability to expand the plasmodesmata size exclusion limit.

Palmitoylation Is Required for NbREM1.5 to Inhibit TMV Cell-to-Cell Movement
To understand if palmitoylation of NbREM1.5 is required to inhibit TMV-MP expanding the plasmodesmata size exclusion limit, we co-expressed palmitoylation-deficient mutant NbREM1.5-C172/175A with TMV-MP, as compared with overexpression of GUS-flag with TMV-MP. The results showed that the rate of GFP diffusion showed no remarkable difference when TMV-MP was co-expressed with the deficient mutant NbREM1.5-C172/175A or GUS-flag (Figure 5b,c), indicating that palmitoylation is required for NbREM1.5 to restrict the ability of TMV-MP to increase the size exclusion limit of plasmodesmata.
In N. benthamiana epidermal cells, there was no significant difference in callose deposition at the plasmodesmata between the overexpression of the palmitoylation-deficient mutant Flag-NbREM1.5-C172/175A and the overexpression of the GUS as a negative control ( Figure S6), indicating that palmitoylation is indispensable for NbREM1.5 to directly hamper plasmodesmata permeability. We then tested whether palmitoylation plays a role in the restriction of TMV cell-to-cell movement by NbREM1.5. Infection assays in N. benthamiana showed that the NbREM1.5 palmitoylation-deficient mutant NbREM1.5-C172/175A completely lost the ability to restrict TMV-GFP cell-to-cell movement, even promoting TMV-GFP cell-to-cell movement (Figure 5d,e). Altogether, these results provide strong evidence that palmitoylation is crucial for NbREM1.5 to inhibit TMV cell-to-cell movement.