||<p>This thesis describes the application of an immunocytochemical technique, immunogold labelling, new in the light and electron microscopic study of the plant viral infection. In Chapter 1 the present state of knowledge of the plant viral infection process, as revealed by <u>in</u><u>situ</u> studies of infected cells, is briefly reviewed. Until now, light and electron microscopic studies have merely described morphological changes in cells and tissue as a result of viral infection, but have failed to provide information on the functional role of these structures in the viral infection process and their association with viral components. A common cytopathological feature of many different plant viruses seems to be the induction of membranous vesicles or membranous bodies, which have been implicated in viral replication. However, only in a few cases some evidence was obtained with regard to the Intracellular location of viral replication and the association of replication and membranes. Available cytochemical techniques have apparently failed to provide a tool for the identification of virus particles and virus-encoded proteins within cellular structures. The Impact of a suitable detection techniques to elucidate the molecular processes of viral replication and transport <u>in</u><u>situ</u> is obvious, as it would link findings obtained by <u>in</u><u>vitro</u> experiments to the events observed in the cell.<p>Immunogold labelling seems to provide such a tool for the tracing of antigens in light and electron microscopic preparations of biological specimens. Gold particles are excellent markers for electron microscopy, because of their high electron density which makes them appear as black dots In EM preparations. Furthermore, by a simple silver staining following gold labelling, viral antigens can be dete cted in semi-thin sections with the light microscope. The application of immunogold labelling for the light and electron microscopic localization of antigens is described in Chapters 2, 3, 4, 5, 6 and 7.<p>In Chapter 2 the preparation of homodisperse suspensions of colloidal gold particles is described. By adsorption of protein A to the surface of the gold particles, a marker (protein A-gold, pAg) is obtained which can be used for labelling antigen-antibody complexes. The specificity of the technique was demonstrated by gold labelling of antibodies bound to plant viruses in mixed suspensions of two viruses. Each virus was labelled using its homologous antiserum and pAg, and no significant cross-reaction with the other virus occurred. Simultaneous identification of two different viruses (CCMV and SBMV) with similar morphological appearance was achieved by double labelling with pAg-complexes containing gold particles of 7 and 16 nm, respectively. Immunogold labelling of viral antigens in suspension has been applied to distinguish between different serologically related viruses like strains of TMV (Pares and Whitecross, 1982), and the potyvirus sugarcane mosaic virus and maize dwarf mosaic virus (Alexander and Toler, 1986; 1985). A clear advantage of the immunogold labelling over conventional decoration of antigens is that the discrete gold particles allow quantification of the results.<p>The immunogold labelling of viral antigen in ultrathin sections of infected protoplasts is described in Chapter 3. Best results were obtained when the protoplasts were only mildly fixed with aldehydes, dehydrated and finally embedded in Lowicryl K4M at -30°C. The antigenicity of viral coat protein was well preserved. A disadvantage of the method is the limited preservation of cell structures, especially membranes due to extraction of lipids. Weibull <u>et</u><u>al.</u> (1983) reported that approximately 50% of the lipid content of cells may be extracted, despite the low temperatures used in the Lowicryl K4M embedding procedure. Ashford <u>et</u><u>al.</u> (1986) questioned the low temperature character of Lowicryl embedding, and found that during polymerization of the resin, temperature rises due to the exothermic nature of the reaction. With plant tissue (not protoplasts), low temperature dehydration and infiltration of the embedding resin must be prolonged, to allow sufficient penetration of the chemicals through the thick walls surrounding the plant cells, and this may result in even more extraction than reported by Weibull and colleagues. Rapid dehydration in ethanol and infiltration of plant tissue with a polar resin like LR White at ambient temperatures, therefore, seems to be a good alternative (Newman <u>et</u><u>al.</u> , 1983; Causton, 1984; Newman and Jasani, 1984).<p>Light microscopic localization of viral antigen in semi-thin sections of LR White embedded plant tissue is described in Chapter 6. CCMV was successfully localized in petiolules of systemically inoculated cowpea plants by immunogold labelling and subsequent silver staining (immunogold/silver staining: IGSS). The silver stain could be observed in the light microscope by brightfield, darkfield and phase-contrast illumination. Most sensitive detection, however, was obtained with epi- illumination using polarized light (epipolarization microscopy). Combining epipolarization illumination with brightfield illumination allowed the simultaneous observation of silver stain and cell morphology.<p>Immunogold labelling and IGSS in combination with appropriate fixation and embedding of biological specimens, appear to be efficient and simple techniques for the <u>in</u><u>situ</u> identification and localization of antigens, with many advantages over other immunochemical and cytochemical techniques, like ferritin- labelling, peroxidase-anti-peroxidase, immunofluorescence and autoradiography, which have only incidentally been used in plant virus research. Recently, Patterson and Verduin (1987) have reviewed the literature on the use of immunogold labelling in animal and plant virology, showing numerous fields of applications and discussing progress made in virus research. With respect to the technique the authors rightly concluded that immunogold labelling is a flexible technique with little limitation for the improvement of existing assays and the development of new ones.<p>Using immunogold labelling to identify and localize virus particles and coat protein, CCMV- infection in cowpea protoplasts was studied as function of the infection time. Observations with regard to virus entry into protoplasts are reported in Chapter 3. Upon inoculation aggregates of virus particles were observed attached to the plasmamembrane, or sometimes penetrating the plasmamembrane at places where the membrane appeared to be damaged. Virus was also found inside vesicles formed by invagination of the plasmamembrane. These vesicles with inoculum-virus particles were stable over long periods of time. Large vesicles (vacuoles) containing viral antigen were also detected at 24 h post-inoculation in protoplasts which were not infected by CCMV.<p>The mechanism by which plant viruses enter their host cells is still disputed (Shaw, 1986). Passage of the plasmalemma by endocytosis was suggested by Takebe (1975), and through pores or lesions by Burgess <u>et</u><u>al.</u> (1973) and Watts <u>et</u><u>al.</u> (1981). Our observations do not favour endocytosis to be the mechanism of virus entry leading to infection of the protoplasts as virus containing vesicles are stable. Recently, Roenhorst <u>et</u><u>al.</u> (1988) presented data supporting a mechanism of virus entry by initial physical association of virus particles with the protoplast membrane and subsequent invasion of virus particles through membrane lesions. Such a mechanism may be also applicable to the cytoplasmic extrusions observed by Laidlaw (1987) after puncturing plant epidermal cells. The author suggested that virus particles may adsorb to the plasmalemma covering the extrusions, which are then withdrawn into the cell. Invasion of whole particles through membrane lesions may then be followed by a uncoating and initial translation (cotranslational disassembly) at the cytoplasmic ribosomes as suggested by Wilson (1985).<p>Ultrastructure of RNA-inoculated protoplasts was studied in sections of aldehyde- and osmium-fixed protoplasts (Chapter 4). Cytological alterations attributed to virus infection consisted of dilation of the endoplasmic reticulum (ER) and the formation of vesicles early in infection. Distended ER and vesicles seemed to form a kind of membranous area in the cytoplasm. In protoplasts fixed and embedded in Lowicryl K4M newly synthesized virus particles or coat protein were first localized in restricted areas of the cytoplasm at 6-9 h post-inoculation. The rough appearance of the cytoplasm in these areas suggested the presence of membranous structures like observed in osmium-fixed protoplasts. However, due to poor membrane preservation in Lowicryl embedded material this could not be proven. Within one protoplast several of these labelled areas were identified. At later stages of infection viral antigen was located throughout the cytoplasm, but also in the nucleus and in particular the nucleolus. No viral antigen was detected in or specifically associated with chloroplasts, mitochondria, microbodies and vacuoles. The specificity of gold labelling was demonstrated by quantification of the labelling density on sections of infected and non-infected protoplasts. These results indicate that CCMV coat protein synthesis and virus assembly take place in the cytoplasm of plant cells, but the involvement of cellular structures, in particular membranes, remains to be established. Protein synthesis and virus assembly may occur in certain restricted sites (compartments) in the cytoplasm possibly formed by the membranous bodies. Compartmentalization of the cytoplasm, creating different environments in the cell, may explain the occurrence of both disassembly and assembly in the same cell, and furthermore account for the phenomenon of specific assembly of viral RUA and homologous coat protein in cells infected with two related viruses like CCMV and BMV (Sakai <u>et</u><u>al.</u> , 1983 ; Zaitlin and Hull, 1987). Whether RNA-replication also occurs in the same location as coat protein synthesis and virus assembly could be established by localization of non-structural virus encoded proteins involved in viral replication. However, antisera against these products of the CCMV-genome were not available. The function of CCMV coat protein or virus in the nucleus and especially the nucleolus is not known. Coat protein may have an affinity for ribosomal proteins and/or fulfill some functional role in the viral replication. Kim 1977 described the occurrence of filamentous inclusions (FI) in the nucleus often associated with the nucleolus. These FI were not found in the nuclei of cowpea protoplasts (this study) or tobacco protoplasts (Burgess <u>et</u><u>al.</u> , 1974), but may be formed later in the infection by excess coat protein. Bancroft <u>et</u><u>al.</u> (1969) showed the ability of CCMV-coat protein to form narrow tubules under specific conditions. The (FI) described by Kim (1977) may represent this type of coat protein aggregation, although the chemical composition of the (FI) is not yet known.<p>In Chapter 5 preliminary observations are reported on the localization of sites of CPMV replication in cowpea protoplasts, by in situ detection of coat proteins and non-structural proteins involved in viral replication and proteolytic processing. With regard to virus entry and subsequent locations of inoculum virus inside vesicles, similar phenomena were observed as in infection with CCMV. Infection of CPMV generates large inclusion bodies in the cytoplasm, consisting of membranous vesicles with fibrillary material and adjoining amorphous electron-dense material which have been observed as early as 12 h post- inoculation. Virus particles and/or coat protein were first detected 24 h after inoculation throughout the entire cytoplasm and in between the membranous vesicles and electron dense material. The 24K, 170K and their precursor proteins were exclusively localized in the electron dense material and not in association with the membranous vesicles or any other location in the cell. These results show that the electron-dense material consists at least in part of CPMV-encoded non-structural proteins and may represent a site for accumulation of non-functional proteins. The membranous vesicles have been implicated in viral RNA synthesis (Goldbach and Van Kammen, 1985). The failure to detect non- structural proteins in association with these membranes may be explained by either a low concentration of these proteins at the site of replication or by extraction of these proteins during the fixation and embedding procedure, despite the low temperature.<p>With IGSS the distribution of CCMV in cowpea plants was monitored at different times after systemic inoculation according to Dawson and Sehlegel (1976) (Chapters 6 and 7). No virus was detected at the time of temperature shift (t=0) in petiolule and leaves of plants subjected to 3 days of differential temperature treatment. Virus was first localized in phloem parenchyma cells of petiolule and veins at t=3 h and from there it spread to neighbouring tissues. Twenty four hours after systemic inoculation virus was located in the phloem, bundle sheath, cortex, but also in the cambium and some xylem cells. These results show that CCMV is transported from the inoculated primary leaves to the secondary leaves through the phloem, apparently following the route of metabolites. This finding is in agreement and further supports the generally accepted concept of plant virus long-distance transport through phloem. tissue (Matthews, 1982; Atabekov and Dorokhov, 1984). The failure to detect CCMV in differentiated sieve elements may indicate that the form in which the infectious entity is transported is another than virus particles (Atabekov and Dorokhov, 1984), or that the amount of virus transported through the sieve elements is below detectable levels. The true character of the synchrony of infection of leaf mesophyll cells obtained by differential temperature treatment is disputed. Infection of mesophyll tells may have been accomplished after shifting the plants to higher temperature by fast transport of infectious particles from the vascular tissue, as was also suggested by Dorokhov <u>et</u><u>al.</u> (1981).<p>For the first time a suitable method for localization of antigens is available, which can be routinely applied for both light and electron microscopic study of the plant viral infection process. The application of the gold labelling technique in the localization of viral structural and non-structural proteins has been demonstrated, using CCMV- and CPMV-infections of plant cells as model system.<p>With regard to the technique, future work must be done on the improvement of the preservation of cellular structures, especially membranes, as these appear only poorly in Lowicryl embedded plant tissue even with dehydration, infiltration and polymerization at low temperatures. Alternatives, may be found in cryofixation and cryosectioning or freeze-substitution techniques.<p>With regard to the study of the plant viral infection process, the localization of virus-encoded proteins involved in replication and transport, but also the localization of plant viral nucleic acids by <u>in</u><u>situ</u> hybridization, will contribute to the understanding of the mechanisms underlying these events. New biochemical techniques like the production of infectious transcripts from cloned viral cDNA (Ahlquist <u>et</u><u>al.</u> 1984) enabling genetic manipulation of the viral genome, and integration of plant viral genes into the plant genome (Gardner <u>et</u><u>al.</u> , 1984; Abel <u>et</u><u>al.</u> , 1986) will supply future model systems for the study of virus-host interactions.