Maataloustieteellinen A ikakauskirja Voi. 59: 193—197, 1987 Immunogold cytochemistry in plant virus research. A review. PER OXELFELT and KARIN TOMENIUS Department of Plant and Forest Protection Swedish University of Agricultural Sciences P.O. Box 7044, S-750 07 UPPSALA, Sweden Abstract. The technique of using antibodies or protein A labelled with colloidal gold for the detection of antigens at the ultrastructural level has only recently come into use in plant virus research. The reports published to date are reviewed and some further possibilities for the use of immunogold techniques in plant virus research are discussed. Index words; immunogold cytochemistry, plant viruses Introduction Over the years many investigators have at- tempted to determine the subcellular sites of virus replication and the sites of synthesis of virus coded proteins in infected plant cells. Plant cells present many problems as com- pared to animal cells because of the limited permeability of the cell wall and the con- siderable asynchrony of infection in intact plant tissue. The immuno-ferritin technique has been used to localize plant viral protein with pre- embedding staining (Shalla and Amici 1967, Shepard et al. 1974, Hatta and Matthews 1976). However, the relatively large ferritin- antibody complex penetrates poorly and binds non-specifically. Under the conditions neces- sary for the diffusion of the labelled anti- bodies redistribution of viral protein cannot be excluded. In recent years techniques have been devel- oped using immunological reagents labelled with colloidal gold. These can be used on ultrathin sections and overcome many of the problems with pre-embedding staining. The immunogold technique has been used in plant virus research only during the last four years. The reports that have appeared to date are briefly reviewed here and the potential of im- munogold techniques in plant virus research is discussed. Methodology The basic principle is that thin sections are treated with antibodies specific for the anti- 193 JOURNAL OF AGRICULTURAL SCIENCE IN FINLAND https://www.c-info.fi/en/info/?token=SKtlWitGtmqc-iNU.xQ5R7NHfGT-DdBupWjdVgg.RPz9syxAYy_rE6EhCT2MbuL6U0JThi0azefY_EqLBUULm7mtfe_QUPhLeoQ0EvlZH5vsrLosnG9_QWYsxoqz4c5QzDjVI9JZ7TD-Z7ulqUzPjz7YhTQX8nR-KTAkRPwH6D5-6q_uD2objB4exPX5Oi-hJT1TpnAGbed7VCFNFg4IFl7iBjeEYaM1jQx2 gen to be localized. These antibodies can be labelled with colloidal gold, but more com- monly a two-step procedure is used in which the specific antibodies are detected by a label- led second antibody (Lin and Langenberg 1983) or labelled protein A (Roth et al. 1978). The gold particles are easily seen in the electron micrographs. Tissue to be used for immunogold studies is fixed in glutaraldehyde. Post-fixation in Os 04 is usually omitted since it considerably reduces antigenicity (Lin and Langenberg 1983). Various embedding media have been tried: Epon (Garnier et al. 1986), Araldite, Lowicryl, and LR Gold (Lin and Langenberg 1983, Tomenius and Clapham 1985 a, b, Lan- genberg 1986). Sections of embeddings in Epon or Araldite are usually etched with hyd- rogen peroxide (Garnier et al. 1986) or alco- holic sodium hydroxide (Lin and Langenberg 1983) in order to improve accessibility of an- tigenic sites. Such treatment has adverse ef- fects on ultrastructure (Tomenius and Clap- ham 1985a). Etching can be dispensed with when Lowicryl or LR Gold is used and these embedding media are therefore preferred (Langenberg 1985, 1986, Tomenius et al. 1987). The sections are mounted on Formvar coated nickel grids. They are first treated with a 1 % solution of bovine serum albumin to reduce non-specific adsorption of immuno- globulins and then incubated with the specific antibody. After washing in buffer they are in- cubated with the gold-labelled second anti- body or protein A followed by a final wash. The grids are kept floating on, or immersed in, drops of the appropriate solution and placed on a magnetic stirrer to ensure gentle movement during all incubation and washing steps. The sections can then be stained with uranyl acetate and lead citrate (Langenberg 1985, Tomenius et al. 1987). Applications The first reports on the use of immunogold cytochemistry in the study of plant viruses ap- peared in 1983. Lin and Langenberg (1983) used the technique to localize barley stripe mosaic virus in infected wheat cells. After applying the gold-IgG complex they found strong uniform specific labelling over crystal- line aggregates of virus particles. A quantita- tive evaluation of the specificity was made. Although the labelling was not very intense on theaverage 1.5 gold particles per virus rod there was a linearrelationship between gold particles and virus rods. The low level of labelling is explained by the fact that only antigenic sites on the surface of the sections are accessible, the resin being impermeable to the antibodies. Tomenius et al. (1983) used immunogold cytochemistry to localize red clover mottle virus (RCMV) antigen in infected pea leaf cells. Label was intense in the cytoplasm and among the membranes of the virus-induced membranous inclusions but not over the ve- sicles. Due to the relatively poor contrast in tissue fixed with glutaraldehyde alone virus particles were generally not seen. However, since gel diffusion tests showed that the anti- serum used was specific for whole virus and did not react with coat protein subunits it was concluded that the gold label showed the location of virus particles. In further studies (Tomenius et al. 1984) using antibodies specific for the larger (40K) capsid protein of RCMV it was shown that label was generally distributed over the cyto- plasm and among the membranes of the virus- induced inclusions. The capsid protein was de- tectable simultaneously with intact virus but could not be detected before the membraneous inclusion had developed. There was no indi- cation of accumulation of the capsid protein at any specific site or in any organelle. Anti- bodies specific for the smaller (22K) capsid protein have also been produced and they are being used in studies with the immunogold technique (unpublished). Langenberg (1985) studied wheat doubly infected with wheat spindle streak mosaic virus (WSSMV) and soil-borne wheat mosaic virus (SBWMV). He found that SBWMV 194 4 antibodies reacted with virions and inclusion bodies of both viruses in thin sections, al- though when tested with virions in leaf dip preparations, the two antisera were comple- tely specific. The conclusion was that SBWMV capsid protein or inclusion body protein apparently binds to WSSMV pin- wheels and virions aggregates in vivo in doubly infected cells. Langenberg (1986) demonstrated the pre- sence of capsid protein or virions in cylindri- cal inclusions induced by two viruses that in- fect wheat, WSSMV and wheat streak mosaic virus. He hypothesized that the cylindrical in- clusions play a role in the cell-to-cell move- ment of virions through the plasmodesmata. Langenberg and Van Der Wal (1986) could indentify barley yellow mosaic virus in roots of infected barley by immunogold label- ling. They also stained sections containing various stages of Polymyxa graminis which is known to be a vector of BYMV and Lagena radicicola which is a possible vector. In no case did they find label above background. In recent years a number of virus-coded proteins other than the structural proteins have been identified and antibodies to them have been produced. It is thus possible to study such proteins with the immunogold technique. Giband et al. (1984) used immu- nogold cytochemistry for the in situ localiza- tion of cauliflower mosaic virus and the major protein of the virus-induced inclusion bodies. Garnier et al. (1986) studied cells infected with turnip yellow mosaic virus. They used antibodies to the RNA replicase and its 115K virus-coded subunit. Label was found at the periphery of chloroplasts, which confirmed earlier findings where the replicase was loca- lized by its synthetic activity. However, they also found label in the cytoplasm and con- cluded that this probably represented newly synthesized replicase or its 115 K subunit. Tomenius et al. (1987) used immunogold cytochemistry to localize the virus-coded 30K protein in tobacco mosaic virus infected leaves. This protein has been proposed to have a function in the cell-to-cell movement of the virus. The 30K protein was localized to the plasmodesmata. There was an accumulation inside the plasmodesmata with a maximum about 24 hours after inoculation. No specific label was found in the nucleus or at any other site in the cells. Although single-stranded nucleic acids are not immunogenic, double-stranded nucleic acids are. Lin and Langenberg (1985) used a poly(l):poly(C) antiserum to study root tips of wheat systemically infected with barley stripe mosaic virus. They found gold label in vesic- les in proplastids thus supporting the notion that proplastids play a role in BSMV replica- tion. Work in progress in this laboratory (unpub- lished) using a poly(A):poly(U) monoclonal antibody (Delage et al. 1984) has shown that double-stranded RNA can be localized in plant cells infected with TMV and RCMV, respectively. In addition to applications in cytochemistry, gold-labelled antibodies can be used for en- hancement of decoration of virus particles in immunosorbent electron microscopy (Lan- genberg 1985, Tomenius and Gidlund 1986). This should prove useful in particular with small spherical viruses present in low concen- tration. Discussion The immunogold technique, although it still has its limitations, has a considerable poten- tial in plant virus research. A limitation is that tissue prepared without 0s04 fixation has relatively poor contrast: membranes are not visible and small spherical virus particles are difficult to see unless present in crystalline aggregates. Attempts have been made to over- come these difficulties (Tomenius and Clap- ham 1985 a, b,) but it is still necessary to compromise between good contrast and well preserved ultrastructure on the one hand, and retained antigenicity on the other. As pointed out earlier only antigen at the very surface of the section can be labelled. However, as shown in several of the reports 195 reviewed here, the specificity is good, and even though the label is not intense, it is still suffi- cient for a quantitative evaluation. By careful adjustment of the conditions in the preparation of colloidal gold it is possible to obtain gold particles of defined size (Frens 1973). This opens the possibility to do double labelling experiments. Antibodies specific for two different antigens, e.g. two virus-coded proteins can be applied and in the second step detected by second antibodies labelled with gold particles of different sizes. Work in progress in this laboratory (unpublished) with antibodies to RCMV and the two capsid pro- teins has shown that this is indeed possible. A possibility that has yet to be exploited is to use cDNA probes for the detection of viral nucleic acids on thin sections. cDNA can be biotinylated and detected by antibodies to biotin; this technique is used in filter hybridi- zation tests for the detection of viruses and viral nucleic acids. In principle, it should be possible to use it in combination with im- munogold cytochemistry. References Deface, G., Nahon, E., Huynh, T., Jeusset, J. & La- cour, F. 1984. A monoclonal antibody to the double- stranded polyribonucleotide complex poly(A):poly(U). Molecular Immunology 21: 939—944. Frens, G. 1973. Controlled nucleation for the regulation of the particle size in monodisperse gold suspensions. Nature Physical Science 241; 20—22. Garnier, M,, Candresse, T. & Bove, J. M. 1986. Immu- nocytochemical localization of TYMV-coded structural and nonstructural proteins by the protein A-gold tech- nique. Virology 151: 100—109. Giband, M., Stoeckel, M. G. & Lebeurier, R. G. 1984. Use of the immunogold technique for in situ localiza- tion of cauliflower mosaic virus (CaMV) particles and the major protein of the inclusion bodies. J. Virol. Methods 9: 277—281. Matta, T. & Matthews, R. 1976. Sites of coat protein accumulation in turnip yellow mosaic virus infected cells. Virology 73: 1—l6. Langenberg, W. G. 1985. Immunoelectron microscopy of wheat spindle streak and soil-borne wheat mosaic virus doubly infected wheal. J. Ultrastruct. Res. 84: 16—23. 1986. Virus protein association with cylindrical inclu- sions of two viruses that infect wheat. J. Gen. Virol. 67; 1161 1168. & Van Der Wal, D. 1986. Identification of barley yellow mosaic virus by immunoelectron microscopy in barley but not in Polymyxa graminis or Lagena radi- cicola. Neth. J. PL Path. 92: 133—136. Lin, N. S. & Langenberg, W. G. 1983. Immunohisto- chemical localization ofbarley stripe mosaic virions in infected wheat cells. J. Ultrastruct. Res. 84: 16—23. & Langenberg, W. G. 1985. Peripheral vesicles in proplastids of barley stripe mosaic virus-infected wheat cells contain double-stranded RNA. Virology 142: 291—298. Roth, J., Bendayan, M. & Orci, L. 1978. Ultrastructural localization of intracellular antigens by the use of pro- tein A-gold complex. J. Histochem. Cytochem, 26: 1074—1078. Shalla, T. A. & Amici, A. 1967. The distribution of viral antigen in cells infected with tobacco mosaic virus as revealed by electron microscopy. Virology 31: 78—91. Shepard, J. F., Gahr, D. G. & Purcifull, D. E, 1974. A study of tobacco etch virus-induced inclusions using indirect immunoferritin procedures. Phytopathology 64: 419—425. Tomenius, K., Clapham, D. & Oxelfelt, P. 1983. Loca- lization by immunogold cytochemistry of viral antigen in sections of plant cells infected with red clover mottle virus. J. Gen. Virol. 64: 2669—2678. —, Clapham, D. & Oxelfelt, P. 1984. Localization of coat protein subunits of red clover mottle virus by im- munogold cytochemistry. Sixth International Congress of Virology, Sendai, 1984. (Abstr.). & Clapham, D. 1985 a. Ultrastructural localization of viral antigen in osmiumfixed plant tissue by immuno- gold cytochemistry. Conference on Developments and applications in virus testing, University of Cambridge, 10—12 April 1985 (Abstr.). & Clapham, D. 1985 b. Improved technique for the detection of viral antigen in thin sections of plant tis- sue by immunogold cytochemistry. Scandem 1985. 38th Annual Meeting of the Scandinavian Society for Elec- tron Microscopy, Linköping, Sweden, June 10—13, 1985 (Abstr.). —, Clapham, D. & Meshi, T. 1987. Localization by im- munogold cytochemistry of the virus-coded 30K pro- tein in plasmodesmata of leaves infected with tobacco mosaic virus. Virology (in press). & Gidlund, A. 1986. ISEM immunosorbent electron microscopy for detection of a small spherical virus at low concentration in infected plants. Scandem 1986, 39th Annual Meeting of the Scandinavian Society for electron Microscopy, Oulu, Finland, June 4 —6, 1986. 196 SELOSTUS Kullaleimaus kasvivirusten immunoelektronimikroskopiassa Per Oxelfelt ja Karin Tomenius Department of Plant and Forest Protection, Swedish University of Agricultural Sciences, P. O. Box 7044, S-750 07 Uppsala, Sweden Vasta-aineiden tai proteiini A;n kultaleimaukseen pe- rustuvilla immunoelektronimikroskooppisillamenetelmil- lä tutkitaan antigeenejä hienorakennetasolla. Vasta vii- me vuosina tätä menetelmää on kokeiltu myös kasvivi- ruksiin. Katsauksessa esitetään menetelmän perusteet ja tarkastellaan sen tärkeimpiä sovellutuksia kasvivirusten tutkimuksessa. Kultaleimauksella on tutkittu mm. virusten sijaintia kas- vin juurissa ja lehdissä. Sillä on paikannettu virusten tuot- tamia proteiineja ja selvitetty niiden merkitystä. Menetelmän heikkoutena on toistaiseksi se, että ilman osmiumtetroksidivärjäystäkasvisolut antavat elektroni- mikroskoopissa heikon kontrastin, jolloin tietyt solun osat ja erityisesti eräät viruspartikkelit erottuvat vain kasau- mina. Lisäksi vain aivan leikkeen pinnalla olevat antigeenit voidaan leimata. Rajoituksista huolimatta kultaleimaus- tekniikalla on laajat sovellutusmahdollisuudet kasvien vi- rusinfektioiden perusmekanismien tutkimisessa. 197