MOVEMENT OF VIRUSES INSIDE PLANTS

After forming copies of its genome, virus must move to the non-infected cells to spread infection to the entire plant. This is usually done via plasmodesmata (PD) or phloem.

Plasmodesmata: Gateways to Local and Systemic Virus Infection

PD: structure and function. They are plasma membrane–lined channels that traverse the cell wall between adjacent cells (Figure 7). They are formed during cytokinesis when a component of the endoplasmic reticulum (ER) becomes trapped in the developing phragmoplast. The ER forms a  central  axial  membranous  element termed  the  desmotubule.  The  space between the plasma

 membrane and the desmotubuleis called the cytoplasmic sleeve and, from electron micrographs, is seen to be interrupted by proteinaceous spokes that create a network of microchannels through which small molecules are judged to diffuse. Therefore, PD have plasma membrane, ER membrane, and soluble phase continuity across the new cell wall, providing opportunities for soluble and membrane-associated molecules to translocate from cell-to-cell.

The extracellular cell wall domain surrounding PD is also distinctive, having callose deposits close to the neck aperture and pectin-rich areas in the central region. PD are present at the interfaces of all symplastically connected cells but become structurally and functionally modified to create specific symplastic domains (Ehlers and van Bel 2010; Oparka and Turgeon 1999); for example, distinguishing different inter- cellular boundaries in the vasculature including in extremis for the creation of phloem element sieve pores. Our current knowledge of the molecular composition of PD is still very limited. Very few proteins have been shown definitively to be located at PD, although a significantly larger number may have a transient or indirect association (Maule 2008). Although it is speculation, reference to the analogous functions of the nuclear pore complex in macromolecular transport might suggest that there will be many tens or hundreds of proteins associated with the structure and function of PD. There is evidence for cytoskeletal components (actin, myosinVIII, centrin, and tropomyosin) (Faulkner et al. 2009), a kinase capable of phosphorylating Tobacco mosaic virus (TMV) MP, a family of receptor-like proteins (PDLP1-8) (Thomas et al. 2008), β- 1,3-glucanase, a family of callose-binding proteins (PDCB) (Simpson et al. 2009), a reversibly glycosylated polypeptide associated with Golgi and PD (RGP2), lipid raft-associated remorin (Raffaele et al. 2009a), and ER-associated calreticulin all being directly associated with the structure (Maule 2008).

Interestingly, these collectively represent the various structural domains within the PD. Hence, components of the cytoskeleton, RGP2, and the kinase probably occur within the cytoplasmic sleeve; calreticulin is an ER luminal protein; PDLP is a type I integral membrane protein of the plasma membrane; and β-1,3-glucanase and PDCB are associated with the extracellular face of the plasma membrane to interact with callose (β-1,3- glucan) at the neck region. The presence of remorin is important because it indicates that the plasma membrane within PD may contain lipid raft microdomains. These support the presence of glycosylphosphatidylinositol (GPI)-anchored proteins, such as PDCB, that are typically associated with lipid raft microdomains (Grennan 2007). Regulation of molecular flux through PD and the roles of the components listed above remains a matter of conjecture. Most of the attention has focused on physical constriction of the PD aperture by callose deposition in the near-wall (Chen and Kim 2009; Epel 2009; Hofmann et al. 2010) or on the importance of actin, probably surrounding the desmotubule, which could place tensional constraints upon the PD structure to limit molecular flux (Ding et al. 1996; Su et al. 2010; White et al. 1994).

Callose production is regulated by a large family of callose synthases and β-1,3-glucanases. Callose synthases operate within the context of larger molecular complexes (Verma and Hong 2001), implicating yet further proteins in regulating callose synthesis. Recently, it has been shown that mutations in the callose synthase GSL8 led to an increased PD size exclusion limit (SEL), allowing the diffusion of the cell autonomous protein SPEECHLESS to neighboring cells (Guseman et al. 2010). On the other hand, Arabidopsis knockout lines for the PD-located β-1,3-glucanase AtBG_pap showed increased callose at PD and reduced PD permeability (Levy et al. 2007a and b). Collectively, these proteins and their paralogues provide huge potential for specificity with respect to tissues and induced responses. The lack of single mutant phenotypes for many of these genes, however, suggests that their functions are at least partially redundant. The observation that overexpression of PDCB (which has callose-binding activity but no known catalytic activity) or RGP2 (which has no known function) increased callose accumulation at PD and reduced molecular flux (Simpson et al. 2009; Zavaliev et al. 2010) shows that regulation of callose accumulation may be significantly more complicated than currently proposed. This is reinforced by the finding that other intracellular and extracellular signals seem to play a role in this regulation. Reactive oxygen species (ROS) induced by treatment with oxidants or by a defective maintenance of cell redox homeostasis have been shown to increase callose concentration at PD or to modulate intercellular transport (Benitez-Alfonso and Jackson 2009). This may be important in the response to stresses and for some developmental transitions where cell-to-cell communication is critical. For example, exposure to aluminum (Jones et al. 2006; Sivaguru et al. 2000) and cadmium (Ueki and Citovsky 2005) induce ROS production and callose deposition at PD, reducing intercellular movement of symplastic dyes and viral MP.

During some developmental processes, such as leaf transition to senescence and fiber cell elongation in cotton, a reduction in PD SEL by callose deposition correlates with an increase in intracellular ROS and in the activity of ROS-scavenging enzymes (Jongebloed et al. 2004; Li et al. 2007; Ruan et al. 2004). More direct evidence of this regulatory mechanism is provided by the study of mutants in the synthesis of tocopherol (the main antioxidants in leaves) or in ROS-detoxifying enzymes (e.g., the thioredoxinGAT1). These mutants accumulate increased callose at PD, which reduces symplastic transport and results in severe developmental phenotypes such as accelerated senescence and impaired meristem development (Abbasi et al. 2009; Benitez- Alfonso et al. 2009). From the predicted impact of callose on flux through PD, we might expect a correlated impact on virus infection, and this is, indeed, the case. Hence, increasing callose turnover in transgenic tobacco expressing specific antisense RNA to β-1,3- glucanase delayed the spread of TMV (Beffa et al. 1996). Conversely, overexpression of the gene from a TMV vector increased the size of necrotic lesions of the virus (Bucher et al. 2001). Furthermore, overexpression of some viral MP—for example,TMV MP, (Olesinski et al. 1996), Tomato spotted wilt virus (TSWV) MP (Rinne et al. 2005), and Potato leaf roll virus (PLRV) MP (Kronberg et al. 2007)—led to increased accumulation of callose, altered carbohydrate partitioning, and altered plant phenotypes, reminiscent, in part, of virus-infection phenotypes.

Interestingly, plants overexpressing RGP2and showing increased callose accumulation show similar phenotypes to those described above for plants transgenic for viral MP. In contrast to the latter plants, however, the RGP2 transgenics showed a decrease in cell-to-cell movement of TMV (Zavaliev et al. 2010). RGP2 is a member of the C1RGP that accumulate as large homomultimeric complexes of approximately 400 kDa in the cytoplasmic sleeve of PD (De Pino et al. 2007; Dhugga 2006), perhaps pointing to a physical blockage of the channel. With respect to actin, treatment of tissue with the chemical disruptors of actin fibers was found to increase PD SEL (Ding et al. 1996) and to increase the size of the PD pore at the neck region of PD (White et al. 1994). This has functional implications because MP seems to interact with actin filaments and mediate virus spreading to neighboring cells (Heinlein et al. 1995; McLean et al. 1995; Su et al. 2010). There is also evidence that TMV MP increases the cross-sectional dimension of the desmotubule by swelling of the ER lumen, regulating the transport by diffusion of TMV replication complex in the ER desmotubule continuum. An additional level of flux control was shown recently where PD at the base of tobacco trichomes showed selectivity for the direction of trafficking and, further, selective specificity in that TMV could move whereas diffusion of green fluorescent protein (GFP) was restricted, even when viral MP was present (Christensen et al. 2009a,b). Currently, there is no explanation for how this might be achieved. Hence, we have a picture of PD as dynamic but highly regulated structures that respond to intra- and extracellular stimuli to modulate the passage of large and small molecules from cell to cell and through the vasculature.