Contribution of host intracellular transport machineries to intercellular movement of turnip mosaic virus.

The contribution of different host cell transport systems in the intercellular movement of turnip mosaic virus (TuMV) was investigated. To discriminate between primary infections and secondary infections associated with the virus intercellular movement, a gene cassette expressing GFP-HDEL was inserted adjacent to a TuMV infectious cassette expressing 6K₂:mCherry, both within the T-DNA borders of the binary vector pCambia. In this system, both gene cassettes were delivered to the same cell by a single binary vector and primary infection foci emitted green and red fluorescence while secondarily infected cells emitted only red fluorescence. Intercellular movement was measured at 72 hours post infiltration and was estimated to proceed at an average rate of one cell being infected every three hours over an observation period of 17 hours. To determine if the secretory pathway were important for TuMV intercellular movement, chemical and protein inhibitors that blocked both early and late secretory pathways were used. Treatment with Brefeldin A or Concanamycin A or expression of ARF1 or RAB-E1d dominant negative mutants, all of which inhibit pre- or post-Golgi transport, reduced intercellular movement by the virus. These treatments, however, did not inhibit virus replication in primary infected cells. Pharmacological interference assays using Tyrphostin A23 or Wortmannin showed that endocytosis was not important for TuMV intercellular movement. Lack of co-localization by endocytosed FM4-64 and Ara7 (AtRabF2b) with TuMV-induced 6K₂-tagged vesicles further supported this conclusion. Microfilament depolymerizing drugs and silencing expression of myosin XI-2 gene, but not myosin VIII genes, also inhibited TuMV intercellular movement. Expression of dominant negative myosin mutants confirmed the role played by myosin XI-2 as well as by myosin XI-K in TuMV intercellular movement. Using this dual gene cassette expression system and transport inhibitors, components of the secretory and actomyosin machinery were shown to be important for TuMV intercellular spread.

Introduction

Plant viruses move from the initially infected cell to neighboring cells during local spread and then over long distances through vascular tissues to establish a systemic infection in the plant. Transport of viruses between cells first involves the intracellular movement of the viral RNA from the site of replication to plasmodesmata (PDs) and then its delivery into neighboring cells through PDs. PDs are tunnels in the cell wall that connect the cytoplasm, the endoplasmic reticulum (ER) and the plasma membrane between adjoining cells (reviewed in [1]). The size exclusion limit (SEL) of PD is normally too small to allow passive transport of large molecular complexes, but plant viruses encode movement proteins (MPs) that increase the SEL of PDs to allow passage of the viral RNA (reviewed in [2], [3]). Intracellular movement likely involves a membrane-associated viral RNA-host and viral protein complex, but the exact configuration of the viral entity that enters the neighboring cells has not yet been determined (reviewed in [2], [4]). In the case of tobacco mosaic virus (TMV), the viral RNA appears to spread between cells as membrane MP-associated viral replication complexes (VRCs) [5]. For members of the comovirus and caulimovirus genera, viral particles transit through MP-induced tubules that go through PDs for their delivery into non-infected cells [6]–[10].

Although MPs and other viral protein components are important for viral RNA intra- and intercellular movement, it is clear that host factors also are required for these activities. The cytoskeleton is an essential component of organelle trafficking in plant cells (reviewed in [11], [12]) and it has been shown to be involved in vertebrate virus intracellular movement (reviewed in [13]). In the case of TMV, several studies have shown that microtubules and microfilaments are necessary to anchor and release, or aid the movement of the VRC or MP granules often associated with ER (reviewed in [2], [4], [14]). Microfilaments influence the intracellular or intercellular transport of other MPs or viruses [15]–[20]. Myosin motors are also required for MP or viral trafficking [6], [21]–[24]. The secretory pathway is further involved in intra- and intercellular trafficking by several viruses [16], [17], [20], [21], [25], [26]. Finally, recent studies suggest that the endocytic transport pathway may be involved in viral movement [27], [28].

However, not all viruses or their components use the cytoskeleton or the secretory pathway for movement. For example, PD targeting of the tubule-forming MP of cowpea mosaic virus (CPMV) is not affected by either the disruption of ER-Golgi transport or by cytoskeleton disruption [29]. Similarly, the targeting of the triple gene block protein 3 (TGBp3) of poa semilatent virus to PD does not require a functional cytoskeleton or the secretory pathway for its intracellular transport [30].

The genome of potyviruses is a single ∼10 kb RNA molecule that codes for a polyprotein, which is processed into ten mature proteins. In addition to polyprotein-derived polypeptides, an ∼7 kDa protein termed PIPO is produced in infected cells [31] and is also found as a trans-frame protein consisting of the amino-terminal half of P3 fused to PIPO (P3N-PIPO) [32]. Potyviruses have no designated MP but many viral proteins have been reported to have MP-related functions. For instance, HCPro and the coat protein (CP) can increase PD SEL [33]. In addition, CP and cylindrical inclusion (CI) protein are required for virus intercellular movement [34]–[36] and are associated with PD [37], [38]. Recently, the targeting of CI to PD was shown to be mediated by P3N-PIPO [39], which itself is targeted to the plasma membrane through an interaction with the host protein PCaP1 [32]. One last protein involved in viral movement is the 6 kDa membrane-associated 6K2 protein. It induces the production of motile vesicles that contain viral RNA and have been proposed to be the vehicle for intracellular trafficking of potyviral RNA [40], [41]. These findings, while providing a partial understanding of the mechanism of TuMV intracellular and intercellular movement, do not sufficiently explain the involvement of the host secretory pathway or the cytoskeleton in potyvirus movement.

In this study, we used a novel dual gene cassette construct that differentiated primary infected cells from cells infected after virus intercellular movement to show that the early as well as the late secretory pathway, but not endocytosis, was important for TuMV transport. We also determined that myosin XI-2 and XI-K, but not myosin XI-F and VIIIs, influenced TuMV intercellular movement. Although these cellular components were required for intercellular movement of TuMV, they did not appear to be involved in the virus protein production in primary infected cells.

Results

An in vivo quantitative assay for TuMV intercellular movement

A recombinant tobacco etch virus (TEV) (genus Potyvirus) engineered to express the reporter protein β-glucuronidase (GUS) allowed direct observation of virus spread in leaves [34]. Virus spread is influenced by the rates of virus RNA replication and virus intercellular movement. Hence, the use of the above TEV-GUS construct to fully interpret results from virus spread studies is limited since virus replication in live tissue cannot be quantitated through GUS staining. In order to discriminate initially infected cells from later infected cells in live tissue, we introduced within the T-DNA borders of a binary vector a gene cassette expressing the ER-localized GFP-HDEL adjacent to a TuMV infectious genome cassette expressing 6K2:mCherry (Fig. 1A). Since both gene cassettes are delivered to the same cells and GFP-HDEL does not move between cells [28], primary infected cells should display concomitant green and red fluorescence while secondary infected cells should display red-only fluorescence. This system also allows differentiation between virus RNA replication and virus intercellular movement.

Figure 1

TuMV intercellular movement time course.

(A) Schematic representation of the plasmid pCambiaTuMV/6K2:mCherry//GFP-HDEL used to discriminate primary infected cells from secondary infected cells after agroinfiltration. cDNA coding for 6K2:mCherry was inserted between P1 and HCPro cistrons. (B) Image of a leaf under uv illumination. Green fluorescing zone shows agroinfiltrated area. (C) Three-dimensional rendering of 35 1 µm thick confocal image sections that overlap by 0.5 µm showing the distribution of TuMV-induced 6K2:mCherry-tagged structures and GFP-HDEL labeled at 72 hpinf. The TuMV-induced 6K2:mCherry-tagged structures represent the viral factories, and the circular green structure in the center of the cell is the nucleus. Scale bar = 20 µm. (D) Three-dimensional Tile imaging rendering of N. benthamiana leaf agroinfiltrated 60 hrs before with A. tumefaciens strain Agl1 containing the above plasmid. The confocal image tiles was formed using the 10× objective by assembly 12×12 images in xy and the three-dimensional rendering was created by 5 z stacks of 90 µm thick confocal images that overlap by 45 µm. Left panel, red fluorescence channel imaging TuMV producing 6K2:mCherry; middle panel, green fluorescence channel imaging GFP-HDEL; and right panel, merged images. (E–G) Same infiltrated area as in D but confocal images were taken at 4, 5 and 6 dpinf. Scale bar = 2.5 mm.

A single infiltration with an A. tumefaciens suspension containing the above plasmid was performed on leaves of three-week old N. benthamiana plants, resulting in an agroinfiltrated area of 5–10 mm in diameter (Fig. 1B). Fluorescence emitted by GFP-HDEL was generally observed at approximately 36 hrs post infiltration (hpinf) and mCherry fluorescence resulting from TuMV replication was detected at approximately 60 hpinf. Systemic TuMV infection was observed at 4–5 days post infiltration (dpinf) in leaves above the infiltrated one by Western blot analysis using a rabbit serum against the CP of TuMV (data not shown). A similar systemic movement was obtained when more dilute agrobacterium suspensions (e.g. 0.01–0.001) were infiltrated, indicating the bacterial load did not elicit a plant defense response that might have affected virus infection rate. N. benthamiana cells displayed the expected green polygonal ER pattern and virus-induced 6K2-tagged red vesicles (Fig. 1C). Virus intercellular movement was assayed by observing red and green fluorescence at the perimeter of the infiltrated area. At 72 hpinf, a majority of cells in the infiltrated area emitted both green and red fluorescence and just a few red-only cells were observed (Fig. 1D), indicating that viral movement was just starting. Viral movement was followed in the same agroinfiltrated region at 4, 5, and 6 dpinf (Fig. 1E–G). At the end of the observation period, the surface area of green fluorescence did not changed, indicating that GFP-HDEL did not moved into neighboring cells.

Intercellular movement of fluorescent signal was also followed by observing spread of green and red fluorescence for 17 consecutive hours starting at 72 hpinf in order to evaluate the rate of cell-to-cell movement. This was achieved by securing an infiltrated leaf still attached to the plant on the confocal microscope stage for observation. When the perimeter of the agroinfiltrated area was initially imaged, intercellular movement have already begun as indicated by the presence of red-only fluorescent patches (Movie S1). To calculate the rate of cell-to-cell movement in cells/hour, we first measured the number of leaf epidermal cells for every linear 1 mm in three-week-old N. benthamiana plants and we calculated that 1 mm corresponded to 17.6 cells (n = 20). We then measured the distance from the agroinfiltrated front to the limit of expansion of the red fluorescence at the end of the observation period. In this experiment, the red-only fluorescence, indicative of TuMV secondary infection, had spread an average distance of 311 µm (n = 20) in the xy plane in 17 h. This expansion corresponded to a rate of one new infected cell every 3 hours. We repeated this experiment five times and we observed the same rate of cell-to-cell movement. This rate of intercellular movement was similar to that observed for TEV expressing GUS and for TMV [5], [34]. The increase in red fluorescence surface area was not due to the diffusion of 6K2:mCherry from agroinfected cells in the absence of virus spread because replacement of the TuMV cassette with a cassette expressing only 6K2-mCherry did not produce red-only fluorescent foci (Fig. S1A). These findings validated the use of the double cassette, GFP-HDEL and TuMV 6K2:mCherry to follow intercellular movement by TuMV.

Intercellular movement of TuMV requires both the early and late secretory pathways

Chemical and protein inhibitors were used to evaluate the role of the early and late secretory pathways in the intercellular movement of TuMV using the dual gene cassette system. The plant secretory pathway consists of the ER, the Golgi apparatus, various post-Golgi intermediate compartments (e.g. trans-Golgi network, endosomes), the vacuoles/lysosomes and the small vesicular transport carriers that shuttle between these compartments. The early secretory pathway embraces the ER–Golgi interface while the Golgi apparatus and the various post-Golgi organelles that control plasma membrane or vacuolar sorting is categorized as the late secretory pathway (reviewed in [42]).

Brefeldin A (BFA) is an inhibitor that interferes with protein transport between the ER-Golgi interface [43]. Concanamycin A (CMA) inhibits protein transport at the trans-Golgi network (TGN) [44] by inhibiting the function of TGN-localized proton-ATPases, which leads to the acidification of the TGN lumen [45]. To evaluate the influence of these secretory inhibitors on TuMV intercellular movement, N. benthamiana leaves were treated with DMSO, 10 µg/mL BFA or 0.5 µM CMA 4 h before pCambiaTuMV/6K2:mCherry//GFP-HDEL agroinfiltration. BFA at this concentration effectively blocked the secretory pathway (Fig. S1B). At 4 dpinf, DMSO alone had no inhibitory effect on TuMV movement (Fig. 2A). On the other hand, BFA and CMA treatment reduced cell-to-cell movement of TuMV (Fig. 2B–C). The surface area for mCherry-only expressing foci for each treatment was measured and the statistical analysis confirmed the inhibitory effect of BFA and CMA on TuMV intercellular movement (Fig. 2D). This experiment was repeated two more times and the statistical data are presented in Fig. S2A. To assess whether or not BFA or CMA inhibited TuMV replication, we quantified mCherry fluorescence intensity over GFP fluorescence intensity in primary infection foci for all treatments. Fig. 2E shows that that there was no significant difference in the ratio of red over green fluorescence during BFA and CMA treatments compared with the no inhibitor treatment (TuMV alone) at 4 dpinf, indicating that viral protein production in the primary infected cells was not affected by the drug treatments. Green fluorescence levels were also similar between DMSO- and BFA- or CMA-treated primary cells indicating that steady state level of GFP, which was not associated with virus replication, was not affected by the treatments (data not shown).

Figure 2

The secretory pathway is required for intercellular movement.

N. benthamiana leaves were infiltrated with DMSO (A), 10 µg/ml BFA (B) and 0.5 µM CMA (C) 4 hours before agroinfiltration with A. tumefaciens containing pCambiaTuMV/6K2:mCherry//GFP-HDEL. All images were taken at 4 dpinf. Left panel, red fluorescence channel imaging TuMV producing 6K2:mCherry; middle panel, green fluorescence channel imaging GFP-HDEL; and right panel, merged images. Scale bar = 200 µm. (D) Surface area of red-only fluorescent foci was calculated and expressed in fluorescence units. (E) Fluorescence intensity ratio of red over green foci was calculated and expressed in fluorescence units. Bars represent means and standard errors for 20 replicates per treatment. One-way analysis of variance calculation followed by Tukey's Multiple Comparison Test allowed analysis of differences between means: **, 0.001

N. benthamiana leaves were infiltrated with DMSO (A), 10 µg/ml BFA (B) and 0.5 µM CMA (C) 4 hours before agroinfiltration with A. tumefaciens containing pCambiaTuMV/6K2:mCherry//GFP-HDEL. All images were taken at 4 dpinf. Left panel, red fluorescence channel imaging TuMV producing 6K2:mCherry; middle panel, green fluorescence channel imaging GFP-HDEL; and right panel, merged images. Scale bar = 200 µm. (D) Surface area of red-only fluorescent foci was calculated and expressed in fluorescence units. (E) Fluorescence intensity ratio of red over green foci was calculated and expressed in fluorescence units. Bars represent means and standard errors for 20 replicates per treatment. One-way analysis of variance calculation followed by Tukey's Multiple Comparison Test allowed analysis of differences between means: **, 0.001