Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. 2010 Feb 9; 107(6): 2491–2496.
Published online 2010 Jan 21. doi:  10.1073/pnas.0909080107
PMCID: PMC2823903
Cell Biology

Arabidopsis synaptotagmin SYTA regulates endocytosis and virus movement protein cell-to-cell transport


Synaptotagmins are calcium sensors that regulate synaptic vesicle exo/endocytosis. Thought to be exclusive to animals, they have recently been characterized in plants. We show that Arabidopsis synaptotagmin SYTA regulates endosome recycling and movement protein (MP)-mediated trafficking of plant virus genomes through plasmodesmata. SYTA localizes to endosomes in plant cells and directly binds the distinct Cabbage leaf curl virus (CaLCuV) and Tobacco mosaic virus (TMV) cell-to-cell movement proteins. In a SYTA knockdown line, CaLCuV systemic infection is delayed, and cell-to-cell spread of TMV and CaLCuV movement proteins is inhibited. A dominant-negative SYTA mutant causes depletion of plasma membrane-derived endosomes, produces large intracellular vesicles attached to plasma membrane, and inhibits cell-to-cell trafficking of TMV and CaLCuV movement proteins, when tested in an Agrobacterium-based leaf expression assay. Our studies show that SYTA regulates endocytosis, and suggest that distinct virus movement proteins transport their cargos to plasmodesmata for cell-to-cell spread via an endocytic recycling pathway.

Synaptotagmins (Syts) are a large family of Ca2+/lipid binding proteins widely studied in animals due to their role in neurotransmitter release. They are also found in Drosophila and Caenorhabditis elegans and were recently described in plants (1, 2). Syts have a conserved domain structure: a short uncleaved N-terminal signal peptide that overlaps a transmembrane (TM) domain, followed by a cytosolic variable region and two C-terminal C2 domains, C2A and C2B. Whereas C2A and C2B each bind phospholipids in a Ca2+-dependent manner, fold independently and act synergistically, C2B is essential for activity (1). SytI, the best studied Syt, is proposed to act as a Ca2+ sensor to regulate rapid and synchronous synaptic vesicle exocytosis (1). Whether it regulates SNARE complex formation in a temporal and spatial manner, or is itself fusogenic, is unclear. Studies in PC12 cells, and of mouse and Drosophila sytI mutants, suggest that the SNARE complex VAMP1/SNAP25/syntaxin-1 targets the synaptic vesicle to the plasma membrane to create a metastable fusion intermediate. SytI on the vesicle membrane, and perhaps a distinct partner Syt on the plasma membrane, would then interact with phospholipids and the SNARE complex to accelerate SNARE-mediated fusion pore dilation. Liposome studies suggest a direct fusogenic role for SytI, in which shallow insertion of the C2 region into target membranes induces curvature to destabilize the lipid bilayer and form the fusion pore opening (1, 3). Studies in mice, Drosophila, and C. elegans show that SytI also regulates the kinetics of endocytosis at nerve terminals, apparently in a clathrin-mediated manner (1, 4).

Plant virus movement proteins (MPs) mediate the transport of progeny genomes across the cell wall for local and systemic infection. Despite diverse strategies for cell-to-cell movement, two common features have emerged: movement proteins alter plasmodesmata (PD), transwall pores that connect adjacent plant cells; and protein localization and interaction studies implicate the endoplasmic reticulum (ER) and membrane trafficking in this process (5 8). This is typified by the Begomoviruses Cabbage leaf curl virus (CaLCuV) and Squash leaf curl virus (SqLCV) and Tobamovirus Tobacco mosaic virus (TMV), with their respective single strand DNA (ssDNA) or RNA (ssRNA) genomes. CaLCuV and SqLCV encode two movement proteins: the nuclear shuttle protein NSP and the cell-to-cell movement protein MP. NSP binds replicated viral ssDNA in the nucleus and shuttles it to the cytoplasm, where MP traps these complexes to direct them to and across the cell wall via ER-derived transwall tubules. NSP then targets the viral genome to the nucleus for new cycles of replication (7). The ER-derived tubules are proposed to be the analog of the desmotubule, the PD axial membrane component that is first derived from cortical ER “trapped” by the wall during cell division (6). TMV genomes replicate at ER-derived membrane sites in the cytosol. TMV encodes a single 30-kDa movement protein (30K), which binds and targets progeny genomes to cortical ER sites and PD. The 30K protein increases PD size exclusion limits to allow viral ssRNA to move cell to cell (5, 7). Mutational and antisense suppression studies show that interaction of 30K with a cell wall pectin methylesterase (PME) is required for TMV cell-to-cell movement and infection. Hence, PME may direct 30K, complexed with TMV genomes, to PD and/or act to alter PD gating (9). These studies, and those of other movement proteins, link vesicular traffic to virus movement and lead to speculation that viral genomes and other macromolecules may target to and through PD by “grabbing” a receptor or exo/endocytosis (10).

We report here the functional analysis of a plant synaptotagmin, Arabidopsis SYTA, which we identified in a yeast interactive screen using CaLCuV MP (MPCaLCuV) as bait. SYTA directly binds to MPCaLCuV in vitro, and to the related SqLCV MP (MPSqLCV) and the distinct TMV 30K. We found that SYTA localizes to endosomes, using FM4-64 and compartment-specific markers. To establish the functions of SYTA in virus movement and in plant cells, we showed that CaLCuV infection is delayed, and TMV 30K and MPCaLCuV cell-to-cell spread are inhibited, in an SYTA knockdown line; and a dominant-negative form of SYTA inhibited endocytosis and the recycling of an endosome marker at the plasma membrane, and the cell-to-cell trafficking of TMV 30K and MPCaLCuV in an Agrobacterium tumefaciens-based leaf transient expression assay. We conclude that SYTA regulates both endosome recycling and the activities of the diverse MPCaLCuV and TMV 30K in virus cell-to-cell movement, and suggest that distinct virus movement proteins transport their cargos to PD for cell-to-cell spread via an endocytic recycling pathway.


Synaptotagmin SYTA Interacts with Begomovirus MP and TMV 30K.

We used the yeast SOS recruitment screen to identify Arabidopsis proteins that interacted with MPCaLCuV and could be involved in virus movement (SI Text). We independently identified SYTA six times in our Arabidopsis cDNA library from ˜300 clones (Table S1). In each clone, the SYTA coding sequence extended from the 3′-UTR at least through the C2B domain. Our longest cDNA began just after the start of the C2A domain and corresponded to the 3′ 1,001 nt of a predicted synaptotagmin gene on Arabidopsis chromosome 2, SYTA (At2g20990). We cloned full-length SYTA cDNA from Col-0 poly(A)-containing RNA by 5′- and 3′-RACE. On the basis of our annotation, this was 1,891 nt in length with a 265-nt 5′-UTR, a 1,626-nt coding region, and a 267-nt 3′-UTR (Fig. 1A). The SYTA locus has 11 exons and 10 introns and would encode a 541-aa protein with one 31-aa TM region and a domain structure characteristic of animal Syts (Fig. 1A). SYTA is one of five putative synaptotagmin genes in Arabidopsis, which like animal Syts are highly divergent in sequence (Table S2). SYTA is predicted to localize to plasma membrane or ER, with a cytosolic C-terminal tail.

Fig. 1.
SYTA binds Begomovirus MP and TMV 30K in vitro. (A) Annotated SYTA locus (Upper) and predicted SYTA domain structure (Lower). Shown are 5′- and 3′-UTRs, exons (boxes), nucleotides in genomic locus (Upper) and in full-length cDNA (Lower ...

MPCaLCuV directly interacted in vitro with the fusion protein GST-SYTAΔTM, in which SYTAΔTM lacked the N-terminal signal peptide and TM. When we incubated 35S-met-labeled wild-type (WT) MPCaLCuV in vitro with GST-SYTAΔTM or with GST, each purified from Escherichia coli and tethered to glutathione-sepharose, MPCaLCuV bound to GST-SYTAΔTM (5–15%), but not to GST (Fig. 1B). MPSqLCV (˜75% similar to MPCaLCuV, as diverse as New World Begomovirus MPs are) and the distinct TMV 30K also specifically bound to GST-SYTAΔTM (5–15%) vs. GST in vitro (Fig. 1B). To further control for specificity, we tested luciferase. In contrast to the three movement proteins, luciferase did not bind to GST-SYTAΔTM or GST (Fig. 1B). Thus, MPCaLCuV, MPSqLCV, and the distinct TMV 30K all interacted with SYTA in vitro, suggesting that SYTA might be involved in Begomovirus and Tobamovirus cell-to-cell movement.

CaLCuV Infection Is Delayed in a SYTA Knockdown Line.

To show that SYTA-MPCaLCuV interaction was important for virus infection and reveal possible roles for SYTA in Arabidopsis, we sought T-DNA insertions in SYTA. We found only one line, Sail 775A08 (syta-1, ecotype Col-0) (11), and verified that the T-DNA was in exon 10 (nt 1,728 in the cDNA) just 3′ to the C2B coding sequence (Fig. 1A). This was the sole T-DNA insertion, based on 3:1 segregation when we germinated seed on 2 mg/L bialaphos (χ2 test: P < 0.01). We tested six plants in a segregating population by PCR-based genotyping and identified three homozygous and three heterozygous lines.

Homozygous T3 lines 1, 11, and 16 were ˜50% knocked down in SYTA transcript levels. They produced a predicted truncated SYTA transcript (Fig. 2C; Fig. S1), which encoded a chimeric SYTA protein that was smaller (59.1 kDa) than WT SYTA (61.7 kDa) due to a stop codon in the T-DNA. To show this and assess SYTA levels, we analyzed protein extracts from 10-day-old syta-1 line 1 and WT seedlings on immunoblots using polyclonal antisera against the SYTA variable domain. We detected full-length ˜61.7-kDa SYTA in WT Col-0, and the ˜2-kDa smaller SYTA in syta-1 line 1, where it accumulated to only ˜10% of WT SYTA (Fig. 2D). These results confirmed the specificity of our antisera for SYTA. Our three syta-1 homozygous lines did not exhibit striking phenotypes when grown under long or short day conditions, or continuous light, and had normal fecundity. This suggested that the truncated SYTA, with C2A and C2B intact, did function, or different SYTs had overlapping functions.

Fig. 2.
CaLCuV infection is delayed in syta-1. (A) Wild-type (WT) Col-0 (red) and syta-1 line 1 (blue) were inoculated with equal amounts of CaLCuV (Table S3, Trial 2). (B) CaLCuV or mock inoculated (GV3101) WT Col-0 or syta-1 line 1 at 30 dpi. (C) sqRT-PCR of ...

To test whether reduced SYTA levels affected virus infection, we inoculated syta-1 lines 1 and 16, or WT Col-0, with CaLCuV. CaLCuV infection progressed more slowly in syta-1, in contrast to Col-0 with its 95–100% infected plants and typical disease symptoms first evident at ˜7 d postinoculation. In syta-1, disease symptoms were attenuated (milder epinasty, less chlorosis, and larger rosettes) and delayed 1–3 days in onset, and infectivity levels were lower (Fig. 2 A and B; Table S3). We confirmed this by semiquantitative PCR (sqPCR), using viral NSP-specific primers to quantify CaLCuV DNA levels in extracts from systemically infected leaves at 15 d postinoculation (12). CaLCuV DNA accumulated in symptomatic systemic leaves from syta-1 to ≤50% of the levels in WT Col-0 plants (Fig. 2E). Thus, SYTA was necessary for CaLCuV systemic spread and infection in Arabidopsis, consistent with SYTA-MPCaLCuV interaction being important for virus movement.

SYTA Regulates TMV 30K and CaLCuV MP Cell-to-Cell Trafficking.

To demonstrate a role for SYTA in viral movement protein function, we developed two leaf-based transient expressions assays. We adapted agroinfiltration (13) to test whether a dominant-negative form of SYTA interfered with TMV 30K and MPCaLCuV cell-to-cell trafficking, when these were expressed as functional GFP fusions (30K-GFP and GFP-MPCaLCuV) in Nicotiana benthamiana (5, 6). We also used biolistic bombardment to transiently express 30K-YFP and GFP-MPCaLCuV in Arabidopsis.

We cloned two potential dominant-negative SYTA mutants for agroinfiltration: an N-terminal tagged Myc-SYTAΔTM that lacked the signal peptide and TM, and a C-terminal tagged SYTAΔC2B-Myc lacking the C2B domain. Myc-SYTAΔTM should be cytosolic and could compete with endogenous SYTA to bind and mislocalize TMV 30K or MPCaLCuV. SYTAΔC2B-Myc would target properly, but be inactive as C2B is essential for function. To test these predictions, we expressed SYTA, SYTAΔTM, and SYTAΔC2B in N. benthamiana leaf protoplasts, each as C-terminal GFP fusions from the constitutive 35S promoter. At ˜20 h posttransfection, the earliest we detected each fusion, full-length SYTA-GFP localized to distinct patches in the cortical region of the cell (Fig. 3 A and C), and by ˜40 h, was in tight vesicle-like structures at or near the cell periphery (Fig. 3B). This resembled the pattern of RabF1, a unique plant Rab5 that appears to be on early/sorting endosomes (14). To examine this localization more precisely, we labeled protoplasts for 10 min with FM4-64, a dye that fluoresces when inserted into the plasma membrane and labels endosomes (15). Like RabF1, FM4-64 labeled most of the SYTA-GFP-labeled vesicle-like structures (Fig. 3 CE), suggesting that SYTA localized to plasma membrane-derived endosomes. Fitting with this observation, SYTA-GFP labeled localized areas of the plasma membrane and endosomes when transiently expressed in N. benthamiana leaves (Fig. 3N).

Fig. 3.
SYTA localizes to endosomes. CLSM projected Z-series of (AE and N) SYTA-GFP or (FM) SYTAΔC2B-GFP expressed in N. benthamiana protoplasts (AM) or leaf epidermal cell (N). Cells imaged at 20 h or 40 h posttransfection ...

At ˜20 h, SYTAΔTM-GFP was diffuse in the cytosol and in a few irregular patches; by ˜40 h, it was in large irregular patches, possibly aggregates or associated with membrane (Fig. S2), which were distinct from SYTA-GFP vesicles at 40 h. In contrast, based on FM4-64 labeling and coexpression of the ER marker KDEL-CFP, SYTAΔC2B-GFP accumulated at the plasma membrane and in reticulate ER, with a few vesicles rarely seen at the cell periphery by ˜40 h (Fig. 3 FM). This showed that SYTA-GFP was functional and that the C2B domain was required for SYTA to translocate from the plasma membrane onto endosomes. We concluded that SYTAΔTM or SYTAΔC2B could act as a dominant-negative mutant.

To assay SYTA function, we coinfiltrated two A. tumefaciens GV2260 cultures. One would express high levels of Myc-SYTAΔTM, SYTAΔC2B-Myc, or C-terminal Myc-tagged intact SYTA from the 35S promoter. The other expressed either TMV 30K-GFP or GFP-MPCaLCuV from the estradiol-inducible promoter (16). In this way, once the SYTA construct accumulated, we induced synchronous expression of 30K-GFP or GFP-MPCaLCuV with estradiol. To visualize cell-to-cell movement, we optimized ratios so that most infiltrated cells expressed the SYTA construct, but ˜1 in 10 cells expressed the movement protein at 15–25 h. On the basis of immunoblots of protein extracts from infiltrated sectors probed with Myc antibody, all SYTA constructs accumulated to comparable levels (Fig. S3). To control for leaf variation in each trial, we compared plasmid pairs that were agroinfiltrated into different sectors on the same leaf. We also agroinfiltrated TMV 30K-GFP alone or with untransformed A. tumefaciens to assess the competence of the leaf for, and the baseline level of, movement for the amount of coinfiltrated bacteria.

As with microinjection or biolistic bombardment (5, 6), TMV 30K-GFP, imaged by confocal laser scanning microscopy (CLSM), localized to punctate fluorescent spots at the cell wall in agroinfiltrated N. benthamiana leaves, and spread cell-to-cell over time, based on the number of foci having two or more adjacent vs. single fluorescent cells. We first detected 30K-GFP at 15–18 h postinduction, at which time it was confined to isolated single cells in ≥90% of foci. Cell-to-cell spread of 30K-GFP to adjacent cells was first evident ˜30 h later than biolistic assays at 46–52 h postinduction (Table S4) (5, 6). To further show that this was cell-to-cell movement, we coinfiltrated the ER marker HDEL-GFP, which does not move between cells (17), with an empty binary vector or with SYTA using the same amounts of bacteria as in our movement assays. Like 30K-GFP at 15–26 h postinduction, ˜90% of HDEL-GFP, examined up to 96 h postinfiltration, was confined to single cells, with ˜10% in two adjacent cells. It was rarely seen in clusters of three cells (Table S4).

Neither SYTA-Myc nor Myc-SYTAΔTM inhibited 30K-GFP cell-to-cell trafficking. In a typical assay, TMV 30K-GFP spread to adjacent cells ˜60% of the time when coexpressed with SYTA-Myc or Myc-SYTAΔTM, or coinfiltrated with GV2260 (Table 1). Hence, overexpressing SYTA did not nonspecifically inhibit cell-to-cell trafficking by overloading the secretory pathway or sequestering Ca2+ to impair cell function. In contrast, SYTAΔC2B-Myc did inhibit 30K-GFP cell-to-cell trafficking ˜35–50%. In the same trial, 30K-GFP spread to adjacent cells in only 36% of foci when coexpressed with SYTAΔC2B-Myc, compared to ˜60% of foci when coexpressed with SYTA-Myc, Myc-SYTAΔTM, or GV2260 (Table 1). In addition, 30K-GFP spread beyond two cells less frequently when coexpressed with SYTAΔC2B-Myc (Table S4). We obtained comparable results for GFP-MPCaLCuV cell-to-cell trafficking in this assay (Table 1).

Table 1.
Effects of SYTA variants or SYTA knockdown on movement protein spread

On the basis of χ2 analysis, SYTAΔC2B-Myc significantly inhibited cell-to-cell transport compared to GV2260, SYTA-Myc, and Myc-SYTAΔTM; the latter three did not vary significantly (Table 1). Thus, SYTAΔC2B-Myc specifically interfered with TMV 30K and MPCaLCuV cell-to-cell trafficking. Expressing SYTAΔC2B-Myc did not notably alter the subcellular localization of 30K-GFP (Fig. S3) or MPCaLCuV retained in single cells, suggesting that SYTA did not recruit either to the cell wall region. To extend these results, we used biolistic bombardment to transiently express 30K-YFP or GFP-MPCaLCuV in Arabidopsis. TMV 30K-YFP and GFP-MPCaLCuV moved to adjacent cells 80–90% of the time in WT Col-0 rosette leaves. In contrast, the cell-to-cell trafficking of both proteins was strikingly inhibited in syta-1 knockdown lines 1 and 16: 30K-YFP and GFP-MPCaLCuV spread to adjacent cells only 20% or 36% of the time, respectively (Table 1). Fitting with our infectivity studies, these findings suggest that SYTA regulates cell-to-cell trafficking of these two viral movement proteins via PD, and that the mutated form of SYTA must be targeted through the secretory pathway to inhibit movement protein transport.

SYTA Regulates Endocytosis.

To establish the role of SYTA in membrane trafficking in plant cells and thereby understand its role in virus movement, we examined the effects of SYTA, SYTAΔTM, and SYTAΔC2B on the localization of GFP markers specific for secretory and endosomal compartments, and on the uptake of FM4-64, using our N. benthamiana transient assay. If SYTA regulates vesicular traffic at a stage along the secretory pathway, then SYTAΔC2B should block a stage in trafficking, visualized by specific GFP markers accumulating in inappropriate upstream compartments (17). To assess this, we used GFP-HDEL (reticulate ER), N-St-GFP (motile Golgi and to a lesser extent ER), Q8 (plasma membrane), or secreted SecGFP (apoplast) (17, 18). If SYTA regulates a stage in endocytosis, then SYTAΔC2B should alter the pattern of FM4-64 labeling and of the endosomal marker RabF1-GFP.

All SYTA constructs were expressed at comparable levels (Fig. S4). SYTA, SYTAΔTM, or control GV2260 did not affect any stage along the secretory pathway; nor did they affect the pattern of endosomes, visualized after brief 0.5- to 1-h labeling with FM4-64. ER to the plasma membrane trafficking, compartment morphology, secretion, and endocytosis all appeared to be normal, on the basis of the pattern of GFP-HDEL-labeled ER, the number and motility of N-St-GFP-labeled Golgi, the uniform Q8 labeling and morphology of the plasma membrane, secGFP accumulation in the apoplast, and the large numbers of FM4-64-labeled endosomal vesicles (Fig. 4; Figs. S4 and S5). These last were smaller than Golgi and not distributed along ER. N-St-GFP did not increase in ER, Q8 and SecGFP did not back up in the Golgi or ER, and SecGFP did not accumulate in the vacuole (17).

Fig. 4.
SYTAΔC2B inhibits the formation of plasma membrane-derived endosomes. (AE) CLSM images (C and D) or projected Z-series (A, B, and E) of FM4-64 labeled N. benthamiana leaf sectors infiltrated with (A) GV2260, (B) SYTA-Myc, or (C ...

SYTAΔC2B also did not affect the secretory pathway; however, it did lead to the depletion of endosomes and accumulation of large FM4-64-labeled vesicles that appeared to be attached to the plasma membrane (Fig. 4 AE; Fig. S4). Thus, SYTAΔC2B inhibited the formation of plasma membrane-derived endosomes. To show that these were early/sorting endosomes, and not at the vacuole membrane, we coexpressed RabF1-GFP with SYTA or SYTAΔC2B. RabF1-GFP expressed alone, or with SYTA, specifically labeled endosomes. In contrast, SYTAΔC2B caused RabF1-GFP to accumulate in large vesicles at the plasma membrane (Fig. 4 FJ). We conclude that SYTA regulates endocytosis (FM4-64) and endosome recycling at the plasma membrane (RabF1-GFP) in plant cells, but not membrane traffic along the secretory pathway. SYTAΔC2B specifically inhibited both these processes and the cell-to-cell trafficking of TMV 30K and MPCaLCuV, thereby providing a functional link between these two pathways. This suggests that TMV 30K and MPCaLCuV transport their cargos to PD for cell-to-cell spread via an endocytic recycling pathway.


Viral movement proteins coordinate genome replication and movement to ensure that replication precedes virus genome transport across the plant cell wall. Thus, a central premise is that movement proteins can identify cellular pathways regulating macromolecular traffic to and through PD. Here we identify a plant synaptotagmin SYTA that interacts with viral movement proteins, show that SYTA regulates endocytosis and endosome recycling, and provide evidence for an endocytic recycling pathway in plant cells to traffic movement proteins to PD.

SYTA both localizes to endosomes and regulates endocytosis at the plant plasma membrane. The latter follows from the functional consequences of inactivating SYTA by deleting the C2B domain. SYTAΔC2B is not on endosomes. It accumulated at the ER and plasma membrane in protoplasts, presumably remaining at the latter because it could not mediate fusion to form endosomes (Fig. 3). Agroinfiltration studies underscore this conclusion. SYTAΔC2B did not alter trafficking in the secretory pathway, but it did inhibit the formation of endosomes at the plasma membrane. This was evident from the depletion of FM4-64-labeled endosomes and the accumulation of large FM4-64- and RabF1-GFP-labeled vesicles at the plasma membrane (Fig. 4). The presence of RabF1 in these vesicles suggests that SYTAΔC2B disrupted endosome recycling. SYTA or SYTAΔTM did not affect trafficking in the secretory or endocytic pathways, or of movement proteins, in this functional assay (Figs. 3 and and4;4; Figs. S4 and S5; and Table 1), showing that high levels of SYTA did not inhibit cell function by sequestering Ca2+. Like SYTAΔC2B in plant cells, an analogous SytI C2B mutant inhibits receptor-mediated endocytosis in HeLa cells and Drosophila (19, 20). We propose that SYTA traffics from the ER to plasma membrane, where it likely acts to regulate endocytosis and endosome recycling at the plasma membrane to maintain homeostasis.

SYTA regulates MPCaLCuV and TMV 30K cell-to-cell spread, as this trafficking is inhibited in the syta-1 knockdown line and by SYTAΔC2B (Table 1). Here, SYTAΔC2B provides the key functional link between endocytosis and virus cell-to-cell movement. Recent studies on other viruses also implicate the ER and membrane trafficking in cell-to-cell movement, although these viruses appear to use diverse strategies to move cell to cell. A functional role for ER is clearly shown in studies of an enhanced TMV 30K that accelerates virus movement, and the interaction of TMV 30K with PME (6, 9). Cauliflower mosaic virus (CaMV) P1 and Grapevine fanleaf virus MP interact with an apparent rat PRA homolog and the cytokinesis-specific syntaxin KNOLLE, respectively (10). The movement proteins TGB2 and TGB3 of Potato virus X (PVX) and Potato mop top virus (PMTV) can, depending on the virus and host, affect PD gating to move the viral ssRNA genome (PMTV) or virions (PVX) cell to cell. GFP fusions of TGB2 and TGB3 localize to ER, and PMTV TGB2 associates with, and recruits, TGB3 to what appear to be plasma membrane-derived endosomes, on the basis of FM4-64 labeling. But, it was not shown whether TGB2, with TGB3, traffic from endosomes to PD or to the vacuole (8, 15). In our studies, if SYTA and early endosomes targeted MPCaLCuV and 30K to the vacuole for degradation, then suppressing SYTA (syta-1 mutant) or expressing SYTAΔC2B would increase movement protein levels and accelerate, not inhibit, CaLCuV systemic spread and/or MPCaLCuV and TMV 30K cell-to-cell movement (Table 1). We conclude that the interaction with SYTA and early endosomes targets 30K and MPCaLCuV via a recapture pathway to PD for intercellular transport. SYTAΔC2B, but not SYTAΔTM, inhibited this process, yet neither obviously altered the subcellular localization of 30K-GFP or GFP-MPCaLCuV. Thus, SYTA must traffic via the secretory pathway to reach the plasma membrane and exert its effect on movement protein trafficking. At our level of resolution we cannot say whether SYTAΔC2B alters the cell wall distribution of 30K or MPCaLCuV.

Two recent studies suggest a role for SYTA in membrane resealing, on the basis of SYTA RNAi knockdown lines or the syta-1 line we used exhibiting reduced tolerance to salt stress or freezing (21, 22). syta-1 was misidentified as a null mutant: the RT-PCR primers used will not detect the truncated SYTA transcript, and antibodies against the C2A/C2B region did not detect SYTA on immunoblots (21). Our antibodies against the SYTA variable domain did detect truncated SYTA protein when sufficient extract was loaded on the gel (Fig. 2). Consistent with our results, one study concluded that at least a subset of SYTA was at the plasma membrane (21). Membrane resealing in animal cells requires Ca2+-dependent endocytosis of damaged membranes (23), but the published studies did not determine the actual function of SYTA in plant cells. Our studies show that the regulation of endocytosis at the plasma membrane by SYTA underlies its roles in biotic (virus spread and infection) and abiotic (21, 22) stress.

We propose a model based on animal viruses, which if they acquire an envelope from an internal membrane will transit through the remainder of the secretory pathway for release. TMV 30K and MPCaLCuV reach cortical ER, from which they will transit to the plasma membrane. Interaction with SYTA at the plasma membrane directs 30K and MPCaLCuV onto early endosomes, which traffic them via a recapture pathway to dock at PD for intercellular transport. Perhaps 30K and MPCaLCuV pirate SYTA and the early endosome machinery for virus cell-to-cell movement, just as HIV and other retroviruses redirect the machinery that forms late endosomes to the plasma membrane for virus budding (24). Whether SYTA cooperates with other SYTs in this process remains to be shown, as does a role for SYTA in 30K transport through PD or in MPCaLCuV forming tubules to alter PD. PME may act to stabilize 30K at the plasma membrane and facilitate 30K interaction with SYTA. The ability of SYTA to bind to MPCaLCuV, MPSqLCV, and TMV 30K is reminiscent of PME binding to movement proteins as dissimilar as TMV 30K and CaMV P1 (9). Most likely, SYTA and PME bind structural motifs within these viral proteins. PMTV utilizes a third distinct mode of movement compared to TMV and CaLCuV. As TGB2 localizes to endosomes, it may be that SYTA and the endosome recycling pathway play a key role in regulating virus cell-to-cell trafficking and, by implication, that of macromolecular complexes via PD.

Plant homologs of clathrin and its associated proteins, and uptake studies using tracers like FM4-64, suggest that plant endocytosis may parallel clathrin-dependent endocytosis in animal cells, where clathrin-coated vesicles at the plasma membrane are internalized and directed to early/sorting endosomes. A recent study provided direct evidence for clathrin-coated vesicles at plant plasma membranes (25), but sorting and recycling endosomes have yet to be fully characterized. Our findings, and recent reports that the potassium channel KAT1 in guard cells, or the auxin efflux carrier protein PIN1, can exchange between FM4-64-labeled vesicles and the plasma membrane (25, 26), provide compelling evidence for an endocytic recycling pathway in plant cells and its crucial role in regulating plant cell-to-cell communication via PD.

Materials and Methods

Unless noted, Arabidopsis thaliana ecotype Columbia (Col-0) was grown at 22 °C with 16 h/8 h day/night cycles. RNA was isolated with TRIzol (Invitrogen). All clones were sequenced.

SYTA cDNA Cloning.

Total RNA was isolated from 10-day-old Col-0 seedlings. GeneRacer (Invitrogen) or SYTA-specific primers were used to reverse transcribe the 3′ and 5′ templates, respectively, and amplify SYTA 5′ and 3′ ends by touchdown PCR. Full-length SYTA cDNA with unique NotI and SalI sites was cloned by crossover PCR using the 161-bp overlapping 5′- and 3′-RACE products as templates (27), and cloned into pBSII SK– (Stratagene). We used PSORT, SMART, Pfam, or WoLF PSORT to predict the SYTA signal peptide, TM, variable domain (VD), and C2 domains or topology (SI Text).

Pulldown Assays.

A translational fusion of GST-SYTAΔTM, lacking the first 32 residues of SYTA, was constructed in pET24a+ (SI Text). GST-SYTAΔTM and GST were purified from E. coli Rosetta(DE3) lysates by binding to glutathione-sepharose (GE Healthcare) in GSTB (50 mM Hepes, pH 7.5, 150 mM NaCl, 1 mM AEBSF) with 2.5 mM CaCl2. Oligonucleotide contaminants were removed by washing in GSTB with 50 mM CaCl2 to an OD260 of zero. MPCaLCuV, MPSqLCV, and TMV 30K were expressed from the SP6 promoter, and luciferase from the T7 promoter, in pGEM7z– (SI Text). Each was labeled with 35S-met in vitro using reticulocyte lysates (TnT system; Promega), and incubated for 2 h at 4 °C with 2 μg GST-SYTAΔTM or GST bound to glutathione-sepharose in GSTB with 2.5 mM CaCl2, 0.1% Igepal CA-300 and 30% glycerol. After three washes in this buffer and one wash in GSTB with 2.5 mM CaCl2, bound protein was eluted in SDS buffer and analyzed by SDS/PAGE (12%) and autoradiography.


GFP-tagged SYTA constructs, cloned as transcriptional fusions to the 35S promoter in pTEX(E) (28), were transformed into GV2260 (SI Text). N. benthamiana protoplasts were prepared and transfected as described (12). FM4-64 (Invitrogen) was used at 20 μM.


A Myc tag was fused to the N terminus of SYTAΔTM, or the C terminus of full-length SYTA or SYTAΔC2B lacking the C-terminal 177 residues. We cloned transcriptional fusions of these to the 35S promoter in pCAMBIA1302, or of TMV 30K-GFP and GFP-MPCaLCuV to the estradiol-inducible promoter in pER8 (16) (SI Text). pVKHEn6 clones of SecGFP, GFP-HDEL, and N-St-GFP, and pBI121-RabF1, pEGAD-Q8, and infiltration of A. tumefaciens GV2260 have been described (14, 17, 18, 28). TMV 30K-GFP and GFP-MPCaLCuV were induced with 5 mM estradiol at 1–2 h postinfiltration and imaged at 46–52 h postinduction. FM4-64 (Invitrogen) at 50 μM in water was infiltrated into leaves 0.5–1 h prior to imaging. We counted 12 fields (coinfiltrations) or 5–10 fields (single infiltrations) in each trial.

Biolistic Bombardment.

pTEX clones for TMV 30K-YFP or GFP-MPCaLCuV were precipitated onto 0.6-μm gold particles (Bio-Rad). Leaves from 12 h/12 h light/dark grown WT Col-0 or syta-1 plants were bombarded using the Helios Gene Gun (Bio-Rad) (29) and imaged by CLSM at 36–44 h postbombardment.


Live protoplasts or leaf pieces were imaged on a Leica SP2 microscope. Z-series were collected at ˜1-μm intervals and imaged with Leica software (Leica Microsystems).

SYTA Antisera.

To create the His-thrombin-SYTAVD fusion, SYTA VD (nt 362–739) was PCR amplified to contain unique NdeI and NotI sites, and fused in-frame to the C terminus of the HIS6 tag encoded in pET28a+ (Novagen). His-thrombin-SYTAVD (17 kDa) was obtained as an insoluble pellet from E. coli Rosetta(DE3) lysates and purified by SDS/PAGE (15%) (SI Text). Rabbit polyclonal antisera were prepared at Cocalico Biologicals, and stored and used as described.


Leaf tissue was frozen in liquid nitrogen and ground at 4° C in 10 mM Tris-HCl pH 8.0, 10 mM EDTA, 1 mM AEBSF, 2% SDS. Proteins were resolved by SDS/PAGE (12%), blotted onto nitrocellulose, probed with SYTA antisera or Myc antibody (Sigma Aldrich), and detected by chemiluminescence (GE Healthcare) (12).

Supplementary Material

Supporting Information:


We thank J. Nasrallah, C. Schopfer, R. van der Hoeven, G. Martin, R. Abramovitch, I. Kaplan, V. Citovsky, S. Ueki, D. Erhardt, T. Ueda, and I. Moore for help, advice and/or plasmids. The work was supported by National Science Foundation (NSF) MCB-9982622, National Institutes of Health AI066054, and Cornell research funds (S.G.L.); and a National Sciences and Engineering Research Council fellowship, a NSF/Department of Energy/US Department of Agriculture training grant, and the Willets Fund (J.D.L.).


The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

This article contains supporting information online at www.pnas.org/cgi/content/full/0909080107/DCSupplemental.


1. Chapman ER. How does synaptotagmin trigger neurotransmitter release? Annu Rev Biochem. 2008;77:615–641. [PubMed]
2. Fukuda M. Molecular cloning, expression, and characterization of a novel class of synaptotagmin (Syt XIV) conserved from Drosophila to humans. J Biol Chem. 2003;133:641–649. [PubMed]
3. Martens S, McMahon HT. Mechanisms of membrane fusion: Disparate players and common principles. Nat Rev Mol Cell Biol. 2008;9:543–556. [PubMed]
4. Jarousse N, Wilson JD, Arac D, Rizo J, Kelly RB. Endocytosis of synaptotagmin 1 is mediated by a novel, tryptophan-containing motif. Traffic. 2003;4:468–478. [PubMed]
5. Waigmann E, Ueki S, Trutnyeva K, Citovsky V. The ins and outs of nondestructive cell-to-cell and systemic movement of plant viruses. Crit Rev Plant Sci. 2004;23:195–250.
6. Roberts AG, Oparka KJ. Plasmodesmata and the control of symplastic transport. Plant Cell Environ. 2003;26:103–124.
7. Lazarowitz SG, Beachy RN. Viral movement proteins as probes for intracellular and intercellular trafficking in plants. Plant Cell. 1999;11:535–548. [PMC free article] [PubMed]
8. Verchot-Lubicz J. A new cell-to-cell transport model for Potexviruses. Mol Plant Microbe Interact. 2005;18:283–290. [PubMed]
9. Chen MH, Citovsky V. Systemic movement of a tobamovirus requires host cell pectin methylesterase. Plant J. 2003;35:386–392. [PubMed]
10. Oparka KJ. Getting the message across: How do plant cells exchange macromolecular complexes? Trends Plant Sci. 2004;9:33–41. [PubMed]
11. Sessions A, et al. A high-throughput Arabidopsis reverse genetics system. Plant Cell. 2002;14:2985–2994. [PMC free article] [PubMed]
12. Carvalho MF, Lazarowitz SG. Interaction of the movement protein NSP and the Arabidopsis acetyltransferase AtNSI is necessary for cabbage leaf curl geminivirus infection and pathogenicity. J Virol. 2004;78:11161–11171. [PMC free article] [PubMed]
13. Goodin MM, Dietzgen RG, Schichnes D, Ruzin S, Jackson AO. pGD vectors: Versatile tools for the expression of green and red fluorescent protein fusions in agroinfiltrated plant leaves. Plant J. 2002;31:375–383. [PubMed]
14. Ebine K, Ueda T. Unique mechanism of plant endocytic/vacuolar transport pathways. J Plant Res. 2009;122:21–30. [PubMed]
15. Haupt S, et al. Two plant-viral movement proteins traffic in the endocytic recycling pathway. Plant Cell. 2005;17:164–181. [PMC free article] [PubMed]
16. Zuo JR, Niu QW, Chua NH. Technical advance: An estrogen receptor-based transactivator XVE mediates highly inducible gene expression in transgenic plants. Plant J. 2000;24:265–273. [PubMed]
17. Batoko H, Zheng HQ, Hawes C, Moore I. A Rab1 GTPase is required for transport between the endoplasmic reticulum and Golgi apparatus and for normal Golgi movement in plants. Plant Cell. 2000;12:2201–2218. [PMC free article] [PubMed]
18. Cutler SR, Ehrhardt DW, Griffitts JS, Somerville CR. Random GFP:cDNA fusions enable visualization of subcellular structures in cells of Arabidopsis at a high frequency. Proc Natl Acad Sci USA. 2000;97:3718–3723. [PMC free article] [PubMed]
19. von Poser C, et al. Synaptotagmin regulation of coated pit assembly. J Biol Chem. 2000;275:30916–30924. [PubMed]
20. Littleton JT, et al. synaptotagmin mutants reveal essential functions for the C2B domain in Ca2+-triggered fusion and recycling of synaptic vesicles in vivo . J Neurosci. 2001;21:1421–1433. [PubMed]
21. Schapire AL, et al. Arabidopsis synaptotagmin 1 is required for the maintenance of plasma membrane integrity and cell viability. Plant Cell. 2008;20:3374–3388. [PMC free article] [PubMed]
22. Yamazaki T, Kawamura Y, Minami A, Uemura M. Calcium-dependent freezing tolerance in Arabidopsis involves membrane resealing via synaptotagmin SYT1. Plant Cell. 2008;20:3389–3404. [PMC free article] [PubMed]
23. Idone V, et al. Repair of injured plasma membrane by rapid Ca2+-dependent endocytosis. J Cell Biol. 2008;180:905–914. [PMC free article] [PubMed]
24. Morita E, Sundquist WI. Retrovirus budding. Annu Rev Cell Dev Biol. 2004;20:395–425. [PubMed]
25. Dhonukshe P, et al. Clathrin-mediated constitutive endocytosis of PIN auxin efflux carriers in Arabidopsis . Curr Biol. 2007;17:520–527. [PubMed]
26. Sutter JU, et al. Abscisic acid triggers the endocytosis of the Arabidopsis KAT1 K+ channel and its recycling to the plasma membrane. Curr Biol. 2007;17:1396–1402. [PubMed]
27. Link AJ, Phillips D, Church GM. Methods for generating precise deletions and insertions in the genome of wild-type Escherichia coli: Application to open reading frame characterization. J Bacteriol. 1997;179:6228–6237. [PMC free article] [PubMed]
28. Frederick RD, Thilmony RL, Sessa G, Martin GB. Recognition specificity for the bacterial avirulence protein AvrPto is determined by Thr-204 in the activation loop of the tomato Pto kinase. Mol Cell. 1998;2:241–245. [PubMed]
29. Ueki S, Lacroix B, Krichevsky A, Lazarowitz SG, Citovsky V. Functional transient genetic transformation of Arabidopsis leaves by biolistic bombardment. Nat Protoc. 2008;4:71–77. [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


  • Gene
    Gene links
  • GEO Profiles
    GEO Profiles
    Related GEO records
  • HomoloGene
    HomoloGene links
  • MedGen
    Related information in MedGen
  • Pathways + GO
    Pathways + GO
    Pathways, annotations and biological systems (BioSystems) that cite the current article.
  • Protein
    Published protein sequences
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...