Interaction between the Alfalfa mosaic virus movement protein and plasmodesmata

For a full infection of a host, plant viruses should be able to multiply in the initially infected cell and to spread to neighbouring cells as to eventually invade the entire plant. The viral transport pathway can in principle be divided into two steps, i.e. cell-to-cell movement within tissues, and long-distance transport via the vascular system. The first process, movement from cell-to-cell, is an "active" process, in terms of that a viral gene product, the so-called "movement protein" (MP) is essentially involved by modifying the host's plasmodesmata to allow viral infectivity to pass. So far at least two different mechanisms of cell-to-cell movement have been described, i.e. the "tobamovirus-type" mechanism, involving transport of virus in a non-virion form through (cytologically) slightly modified plasmodesmata, and the "comovirus-type" mechanism, involving the formation of transport tubules in heavily modified plasmodesmata and allowing mature virus particles to pass. In both mechanisms the viral MP plays a role in the modification of the host's plasmodesmata, and in case of the tobamovirus movement, an increased size exclusion limit of the plasmodesmata is detected.

The PhD research project described in this thesis was part of a programme aiming to unravel the role of the plant's plasmodesmata in virus transport and in cell-cell communication. An important approach in this programme was to use the MPs of two different viruses, each representing both distinct viral transport mechanisms, as probes to identify which host proteins are involved in viral movement and possibly investigate the molecular make up of the plasmodesma. For this purpose the viruses Alfalfa mosaic virus (AMV) and Cowpea mosaic virus (CPMV) were selected, which moreover share a common host, i.e. Nicotiana benthamiana , allowing us to study viral MP-plasmodesma interactions in a single host. Previous work had demonstrated that CPMV uses a tubule-guided movement mechanism, whereas it was anticipated that AMV, in view of its genetic relationship to tomaboviruses, would use the tobamo-type transport. The rational behind the choice for this combination was that the MPs of these two viruses would interact with plasmodesmal constituents or proteins in a distinct way and thus would represent different probes for identifying plasmodesmal proteins and other MP-interacting host proteins in e.g. the yeast two-hybrid system.

This thesis covers the cytological studies within the research programme. The first part of this thesis (Chapters 2-5) deals with the unravelling of the cell-to-cell movement mechanism as utilised by AMV, and in the last part (Chapter 6) cytological studies are presented focussing on a host protein (denoted AD3) interacting with the AMV MP in the yeast two-hybrid system (PhD thesis L. Jongejan, UL, The Netherlands, in preparation).

At the onset of this PhD research hardly any information on the cell-to-cell movement mechanism of AMV was available. Using cowpea protoplasts as test system, it was first shown (Chapter 2) that AMV, like CPMV, is able to form tubular structures on infected plant cells, unexpectedly indicating that AMV would also use a tubule-guided movement mechanism like CPMV. Further analysis indeed demonstrated that these MP-constructed protrusions contain mature virus particles.

To confirm whether AMV, similar to CPMV, would use these structures for a tubule-guided cell-to-cell mechanism, different mutants, defective in systemic spread, were tested for their (in)capability to induce virus-filled tubules. The CP mutant CP4P was previously reported to be unable to form stable virions and was now shown to induce empty tubules on protoplasts (Chapter 3). Moreover, both MP mutants SP6 and SP7 appeared unable to form tubules on infected protoplasts. Hence, the results with all three mutants indicated that the inability to produce virion-filled tubules on single plant protoplasts coincides with a transport-deficient phenotype. These three mutants were completely deficient in cell-cell movement in parenchyma tissues, a restricted spread of these viruses to neighbouring cells was observed in cortex cells surrounding the vascular system. Whether this indicated a second, tissue-dependent minor movement mechanism of AMV or whether this spread reflected distinct properties of plasmodesmata of cortex cells, was not further investigated.

During intensive cytological searches in systemically AMV-infected N. benthamiana plants no elongated tubules could be discerned, despite an earlier report. At the front of infection modified plasmodesmata were found containing both the AMV CP and MP, and having a significantly wider diameter than those in non-infected as well as fully infected tissues (Chapter 4). This finding implied that the modification of the plasmodesmata by AMV is only transient, restricted to the front of infection. Cryo-sectioning of such plasmodesmata revealed the presence of rows of virus particles within the interior of the plasmodesmal pore, which might suggest that only short, not-extending tubules are formed (Chapter 4 and Figure 1 of General Discussion)

Furthermore, not only a transient modification of plasmodesmata was observed in AMV-infected leaf tissue, also a temporary increase of their number at the infection front was noted (Chapter 5). By inspecting cell walls in mesophyll tissue respectively in front of, at, and after the infection zone it was calculated that the density of plasmodesmata in the infection zone at least doubled. It remains to be resolved whether this increase is due to specific viral induction of new plasmodesmata to become engaged in viral transport or whether this reflects a host response to restore cell-to-cell communication after a major part of the pre-existing plasmodesmata have become modified.

In a simultaneously executed PhD project within the same programme (by L. Jongejan, PhD thesis, University of Leiden, in preparation), the AMV MP was used as bait in the yeast two-hybrid system to identify host ( N. benthamiana ) proteins which are specifically targeted by this viral protein during the infection process. Chapter 6 deals with the cytological analysis of the AD3 protein, an AMV MP-binding plant protein of which the expression could be verified in plant tissue. Specific antibodies were raised against heterologously expressed AD3 and used in immuno-cytological localisation studies. These studies revealed that AD3 is specifically found in membrane fractions of both leaf and root tissues of N. benthamiana , and immuno-gold EM demonstrated its localisation in the plasma membrane, which is not in contradiction with a potential function of this protein to support intercellular movement of viruses. Probably due to the low expression levels of AD3 and the transient presence of AMV MP solely at the infection front, a co-localisation between these two proteins in situ could not be established