- Top of page
- Materials and methods
- Supporting Information
Fluorescence recovery after photobleaching (FRAP) was used to study the mechanism by which fluorescent-protein-tagged movement protein (MP) of tobacco mosaic virus (TMV) is targeted to plasmodesmata (PD). The data show that fluorescence recovery in PD at the leading edge of an infection requires elements of the cortical actin/endoplasmic reticulum (ER) network and can occur in the absence of an intact microtubule (MT) cytoskeleton. Inhibitors of the actin cytoskeleton (latrunculin and cytochalasin) significantly inhibited MP targeting, while MT inhibitors (colchicine and oryzalin) did not. Application of sodium azide to infected cells implicated an active component of MP transfer to PD. Treatment of cells with Brefeldin A (BFA) at a concentration that caused reabsorption of the Golgi bodies into the ER (precluding secretion of viral MP) had no effect on MP targeting, while disruption of the cortical ER with higher concentrations of BFA caused significant inhibition. Our results support a model of TMV MP function in which targeting of MP to PD during infection is mediated by the actin/ER network.
Tobacco mosaic virus (TMV), type member of the tobamoviruses, is one of the most extensively studied viral pathogens of plants. Despite over 100 years of research into the interaction of TMV with its hosts, the mechanism by which TMV moves between plant cells remains uncertain (1–6). However, it has been speculated that during coevolution with host plants, viruses ‘hijacked’ one or more host proteins for trafficking molecules through plasmodesmata (PD). As a result, studies of the interaction of TMV with host plants are of particular relevance because they form the basis for models of macromolecular trafficking. TMV, in common with a wide range of viruses, encodes a viral movement protein (MP) that is involved in a number of interrelated functions. These include targeting and accumulation within PD (7–10), an increase in the size exclusion limit (SEL) of the PD pore (9,11,12) and the ability to bind to single-stranded RNA (13). Although the MP accumulates within PD (14), and increases the SEL of the PD pore in transgenic plants (11), it appears that during a natural viral infection, the ‘gating’ of PD is transitory and is restricted to the leading edge of infection (9), while the loss of gating function possibly involves the phosphorylation of the MP within the PD pore (15).
In recent years, considerable attention has focused on the mechanism by which the TMV MP is targeted to PD, although data remain conflicting as to the precise intracellular pathway used by MP to reach the PD pore. Over the past decade, a series of reports have shown the association of the TMV MP with elements of the cytoskeleton, including actin (16), but more predominantly, microtubules (MTs) (17–23), although the latter is under spatiotemporal control, being more pronounced several cell layers behind the leading edge of the infection. In keeping with the growing body of evidence that suggests that MTs may play a general role in RNA trafficking in eukaryotic cells (24–27), a number of studies have correlated MP binding to MT with the efficiency of TMV spread (21,22,28,29). Support for a role of MTs in the intracellular trafficking of MP was obtained by Boyko et al. (21) who identified a conserved tobamovirus MP sequence with homology to a tubulin motif and from observations that TMV mutants with point mutations in the putative tubulin-binding domain of the MP showed reduced cell-to-cell spread. Also, a mutant form of MP, MPNT-1, which interferes with the association of MP–green fluorescent protein (MP–GFP) with MTs, reduced the trafficking of MP through PD, although targeting of MP to PD was not reduced (30). However, in a previous report, we examined the putative role of MTs in delivery of the TMV MP to PD and found that viral cell-to-cell movement was unimpeded by MT inhibitors, such as colchicine or oryzalin, or by silencing of the α-tubulin (TUA) gene (23,31). Furthermore, a TMV vector expressing a DNA-shuffled MP gene (32) that showed enhanced cell-to-cell movement showed greatly reduced association with MTs (23,33), suggesting that MTs are not essential for TMV cell-to-cell movement. While MTs may not play a direct role in the TMV cell-to-cell movement process, the pathway of MP to the PD pore remains unresolved. It has been shown that the TMV MP interacts with a number of intracellular components during infection. For example, an early event during TMV infection is the association of MP with the cortical endoplasmic reticulum (ER) (18,20,23,34), leading to the suggestion that the ER might provide the functional pathway for MP targeting to PD (23,34).
One of the major problems in identifying the very early stages in the trafficking of MP to the PD pore lies in the examination of static images during the infection cycle. Thus, early events in the targeting process are assumed to be reflected by events at the leading edge of TMV infection site (17,19,21–23,31) or by the first observable events in infected protoplasts (16,20,35). These limitations are exacerbated by the fact that GFP and similar fluorophores, which have been used extensively as tags for the viral MP, may take several hours to mature (36,37), during which time the earliest events in the infection process are likely to have occurred or are unobservable due to the imposed detection limits of the fluorescent reporter. To overcome this problem, we used the technique of fluorescence recovery after photobleaching (FRAP) to selectively photobleach MP–GFP-labeled PD at the leading edge of the infection site, and subsequently followed the recovery of fluorescence within the PD, reflecting movement of new fluorescent protein into the photobleached area (38). Fluorescence recovery in PD was significantly affected by pharmacological agents that disrupted the ER and/or the actin cytoskeleton, and strongly inhibited by the less specific metabolic inhibitor azide. In contrast, disruption of MTs had no effect on the trafficking of MP to PD. Our results support the view that the TMV MP is targeted to PD by means of the cortical ER/actin network.
- Top of page
- Materials and methods
- Supporting Information
Although GFP is a powerful tag for MP studies (43), real-time imaging of MP–GFP kinetics at the true leading edge of infection is hindered by the maturation time of the GFP fluorophore, which can take a period of hours even in rapidly maturing GFP variants (36,37). It is therefore likely that some of the very early events in MP trafficking to and through PD will already have occurred before the MP–GFP fusion becomes detectable. Although MP–GFP accumulation in PD has been reported to be a very early event in TMV infection (18,19,23,44), it is also possible that viral RNA trafficking to adjoining cells has occurred prior to maturation of the fluorophore. However, in a previous study of TMV infection, it was shown that excision of unlabeled cells at the leading edge of an infection site prevented virus movement beyond these cells, suggesting that accumulation of MP within PD is indeed an early event in the infection process (9). As levels of mobile MP–GFP in the cell may be too low for detection by confocal microscopy, we adopted an FRAP approach to examine the kinetics of MP targeting to PD in TMV-infected cells.
Following photobleaching, MP–GFP fluorescence was restored in PD, which reached 40% of prebleach levels within 40 min. The recovery rate is considerably slower than the rate observed for the recovery of photobleached Golgi bodies [in the order of 100 seconds (45)], intralumenal diffusion as reported for the ER (46), or the remodeling and diffusion of plant ER membranes (47) and other membrane-bound plant compartments such as chloroplast stromules (48). However, in our experiments, the amount of MP available for movement to the PD was considerably lower at the infection front than in the systems analyzed above, which could explain the slow recovery rates.
An early role for the actin/ER network in MP trafficking
The cortical ER network in plants is intimately associated with the actin cytoskeleton (39,49,50), the latter providing a cytoskeletal scaffold for the movement of specific macromolecular complexes, including the Golgi bodies, along the ER in cortical regions of the cell (39,51,52). Previous studies of MP targeting during TMV infection have implicated both the cortical ER (34) and the actin cytoskeleton (16) in MP trafficking to PD, based on an association of MP with both of these components at various stages of the infection cycle (18,23). To examine the role of the ER/secretory system in MP trafficking to PD, we used the drug BFA. At low concentrations, this drug has been shown to block vesicle-mediated secretion in plant cells and to cause the Golgi bodies to become reabsorbed back into the ER (41). As some viral MPs have been suggested to be delivered to PD by Golgi-body-derived vesicles (53), we were interested to determine whether this might be a feasible pathway for trafficking of the TMV MP. Following confirmation that low BFA concentrations (10 μg/mL) caused the Golgi bodies to disappear, but did not severely disrupt the ER, subsequent monitoring showed that fluorescence recovery at PD was unaffected relative to untreated controls, indicating that Golgi-body-mediated secretory events are not required for MP delivery to PD.
In contrast, at the higher BFA concentration (100 μg/mL), the ER was severely disrupted (54) and fluorescence recovery into PD was significantly reduced. Similarly, MP trafficking into PD was significantly decreased by inhibitors of the actin cytoskeleton, cytochalasin B and latrunculin. While causing complete disruption of the actin cytoskeleton, these compounds also affect the structure and movement of the ER. Although it is possible to disrupt the ER structure and maintain the actin cytoskeleton, disruption of the actin scaffold leads to considerable alteration of the overlying ER network. These data therefore suggest that MP trafficking to PD requires components of the ER network closely linked to a functional actin network.
Although it has been claimed that TMV MP lacks an apparent ER signal peptide (2,55), it does behave as an integral membrane protein (34,56,57). In this context, it has been proposed that the interaction between MP and either PME or calreticulin might provide the MP with the ER-targeting function it requires (6,58). It has also been suggested that MP may use an endogenous non-cell-autonomous protein (NCAP) pathway for targeting to PD (59). Mutations in the transmembrane domain of NCAPP1 (non-cell-autonomous pathway protein 1), normally localized to the cortical ER, have been shown to block trafficking of specific NCAPs including TMV MP and CmPP16.
Treatment with either inhibitor of actin function or high concentrations of BFA resulted in some recovery of fluorescence. It has recently been shown that diffusion of protein within the ER membrane is reduced but not abolished by treatment with latrunculin (47). We therefore propose that disruption of the ER results in accumulations of ER membrane at the vertices some of which may associate closely with PD sites. Therefore, under treatment with inhibitors, MP may still be able to diffuse along the ER membranes to the PD at these sites. Whether the MP actively moves along the ER with independent motility using molecular motors, or whether it simply moves with the flow of the cortical ER membrane, the latter showing significant mobility within the cortical layer, remains to be determined (47). Thus, the MP may require only a temporary insertion into the ER membrane, with subsequent actin-driven ER flow, or even diffusion within the ER membrane, carrying it to the vicinity of PD. The proposed role of the ER in the targeting of MP to PD may explain the apparent contradiction between the present work and the recent observations of Prokhnevsky et al. (60) that TMV MP targeting can take place in the presence of actin and myosin inhibitors. This accumulation occurs over a time period in excess of 18 h and may therefore involve diffusion of MP within the ER membrane into PD.
We found that fluorescence recovery was more severely inhibited by treatment of cells with azide than with any of the other treatments used. Because the disruption of cell structure by azide was apparently similar to treatment with either cytochalasin B or latrunculin, that is, disruption of actin filaments and cessation of ER and Golgi body movement, this would suggest that additional factors may be involved. Azide functionally depletes energy levels within cells indicating that one or more energy-dependent processes, for example, phosphorylation of the MP (61), are required for trafficking of viral MP to or its accumulation within PD. Previous studies have suggested that energy depletion (62,63) or actin inhibitors (64) dilate rather than close PD. Our results do not contradict these studies as the observed effects would reflect transport to (or trapping within) PD, rather than movement through the PD pore.
Taken together, the FRAP data obtained for PD indicate that MP traffics to PD in the ER membrane through an actin-dependent, energy-dependent mechanism and does not require Golgi-body-mediated secretion or MTs. This is consistent with the observation that movement of TMV viral replication complexes, labeled with GFP, is inhibited by latrunculin (31) and the recent demonstration that the TMV replication complex, along with the 126-kD protein, a constituent of the viral replication complex, traffics along microfilaments (65). Similarly, it has recently been shown that the actin/myosin network is involved in the targeting of a viral Hsp70 homolog to PD (60).
MTs and MP trafficking
MP trafficking to PD at the leading edge of infection was unaffected by MT inhibitors (oryzalin and colchicine). During later stages of the infection process, we observed that the association of MP with MTs protects the MTs against disruption by either oryzalin or colchicine as also recently shown by Ashby et al. (66). However, we confirmed that disruption of filaments incorporating TUA–GFP does occur at the infection front (66). Although plant MTs may require high concentrations (in the mm range) of colchicine to cause depolymerization, the dinitroaniline herbicide oryzalin has been shown to be more potent and cause the disappearance of most MTs when used at nanomolar concentrations (67). It has been contested that tua–GFP is not a reliable marker for MT and that MT might not be completely disrupted by chemical treatment (68). In an attempt to counter this suggestion, we showed, in parallel experiments, that colchicine and oryzalin, especially the latter, resulted in a marked disruption of MTs when both α-tubulin (indicated by observation of TUA–GFP transgenic plants) and β-tubulin (revealed by antibody labeling) were imaged. If MTs are involved in the early trafficking of MP to PD, we contend that such severe treatment, even if not disrupting every MT, should have resulted in significant differences in the FRAP measurements, which were not observed.
In conclusion, we contend that at the leading edge of a TMV infection, the targeting of MP to PD requires components of the ER network closely linked to a functioning actin network and does not involve MTs.
- Top of page
- Materials and methods
- Supporting Information
Video 1: Time series of ER movement. Series of 30 images, taken at 1s 164ms intervals, of ER-GFP epidermal cells showing normal ER movement in tissue pre-treated for 2h with water (control), 500μM colchicine or 20μg ml−1 oryzalin, and the disrupted structure and lack of movement in tissue treated with 0.02% sodium azide, 25μM latrunculin or 0.1mg ml−1 cytochalasin B.
Video 2: Time series of Golgi apparatus movement. Series of 30 images, taken at 1s 164ms intervals, of STtmd-GFP epidermal cells showing the normal movement of Golgi stacks in tissue pre-treated for 2h with water (control), 500μM colchicine or 20μg ml−1 oryzalin, and the lack of movement in tissue treated with 0.02% sodium azide, 25μM latrunculin or 0.1mg ml−1 cytochalasin B.
Figure 1: Appearance of cell organelles in Nicotiana epidermal cells. In all images GFP-associated fluorescence is coloured green and chlorophyll-associated autofluorescence is coloured blue.
(A) Appearance of a sodium azide-treated TMV.MP-EGFP.CP lesion on Nicotiana benthamiana away from the leading edge showing that azide does not alter the accumulation of MP on MTs. Bar = 25μm.
(B) TUA-GFP plants showing MTs and apparent accumulation of GFP into aggregates after treatment with 100μg ml−1 BFA. Bar = 25μm.
(C) Maximum projection of a stack of confocal images taken at the leading edge of a TMV.MP-EGFP.CP lesion on Nicotiana benthamiana (left) showing the targeting of MP to PD, formation of clusters and accumulation on MT after treatment with 100μg ml−1 BFA. Bar = 25μm.
(D) Alexa 488 phalloidin staining of actin, showing an intact actin network after treatment with 100μg ml−1 BFA. Bar = 25μm.
(E) TUA-GFP transgenic plants, 4 days after Agrobacterium inoculation with a strain carrying a binary for YFP-HDEL expression. When treated with cytochalasin B, GFP fluorescence (green) in a non-infected cell (left) forms a similar pattern to ER-associated fluorescence (magenta) in the neighbouring cell (upper right) indicating possible accumulation of GFP-tubulin at the ER vertices. Bar = 25μm.
(F) Maximum projection of a stack of confocal images taken at the leading edge of a TMV.MP-EGFP.CP lesion on Nicotiana benthamiana (lower left) showing the targeting of MP to PD and the formation of clusters after treatment with latrunculin. Bar = 25μm.
(G) Anti-ß-tubulin Cy3 conjugate labeling of Nicotiana benthamiana leaf epidermis under control conditions showing an intact MT network. Bar = 10μm.
(H) and (I) TUA-GFP transgenic plants infected with TMV.MP-mRFP.ΔCP. At the edge of the infection the presence of MP clusters (magenta) (darts) does not prevent the complete disruption of the MTs (green) following treatment with 500μM colchicine (H). Under control conditions (I) MP clusters (magenta) (darts) at the edge of the infection can be seen, as well as MTs (green) (arrows). Bar = 25μm.
(J) Maximum projection of a stack of confocal images of a TMV.MP-EGFP.CP lesion on Nicotiana benthamiana showing that the association of MP with MT prevents the disruption of MT when treated with 500μM colchicine. Bar = 25μm.
(K) and (L). Alexa 488 phalloidin staining of actin, showing no significant change after treatment with colchicine (K) or oryzalin (L). Bar = 25μm.
(M) and (N) ER-GFP transgenic plants showing no obvious alteration to the ER network in the presence of colchicine (M) or oryzalin (N). Bar = 25μm.
(O) and (P) STtmd-GFP transgenic plants treated with colchicine (O) or oryzalin (P) showing no significant effect on Golgi stacks. Bar = 25μm.
Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.