J. gen. Virol. (1976), :3I, 265-269 Printed in Great Britain 265 Electron Microscopy of Tobacco Mosaic Virus Prepared with the Aid of Negative Staining-Carbon Film Techniques (Accepted 13 January I976) SUMMARY The negative staining-carbon technique has been applied to suspensions of freshly prepared type strain tobacco mosaic virus particles in high concentrations. Electron microscope images show that paracrystalline arrays of the virus were formed, in which large areas of rods could be viewed along their long axes and in parallel arrays in horizontal positions. High-resolution micrographs showed the protein structure units in rods photographed in both vertical and horizontal orientations. Recent applications of the negative staining-carbon film technique to several icosahedral viruses have shown that two-dimensional and three-dimensional crystalline arrays can be rapidly formed (Horne& Pasquali-Ronchetti, 1974; Horne, Pasquali-Ronchetti & Hobart, I975a; Horne, Hobart & Pasquali-Ronchetti, 1975b ). Studies of filamentous viruses have been carried out on both native particles and reconstituted products with the aid of the above technique (Horne et al. I975b; Goodman, Horne & Hobart, I975a; Goodman et ai. t975b ). The advantages of the method over conventional negative staining are threefold: (a) it is possible to use highly concentrated virus suspensions (1.o to t5 mg/ml virus) for direct examination in the electron microscope, (b) the ratio between the specimen thickness and the supporting film can be arranged to provide optimal contrast conditions and (c) the specimens are ideally suited to optical diffraction and low-angle electron diffraction. We have applied the negative staining-carbon film method to suspensions of native tobacco mosaic virus (TMV) at various concentrations and under different conditions of negative staining. The samples used in our experiments were prepared for electron microscopy as soon as possible after the final centrifugation and resuspension in water. The concentration of virus was estimated to be about 9"o mg/ml. For electron microscopy the concentrated TMV suspensions were diluted 1:5, I :~o and i:2o with glass distilled water to give a range of samples for negative staining and spreading on to freshly cleft mica surfaces (Horne& Pasquali-Ronchetti, 1974). Ammonium molybdate (3 ~) was used as the first negative stain and mixed with an equal volume of diluted TMV from the above suspensions. To study the effects of the negative stain on the TMV rods and spreading properties at the mica surface, the ammonium molybdate was adjusted to give a ph range of 5"2, 6.o, 6"5, 7"o and 8.o. The second negative stain was freshly prepared solutions of 0"5 or 2.o ~ uranyl acetate in glass distilled water. Following the results reported by Unwin & Henderson (1975), we carried out several experiments on TMV suspensions diluted 1:6 with I ~o glucose in water and then mixed with an equal volume of 3 ~ ammonium molybdate. We also studied the effects on particle aggregation and stain penetration of using mixtures of virus, glucose and ammonium molybdate made up with deuterium oxide (D20) in the same ratios as the distilled water samples mentioned above. The electron micrographs were recorded on a JEOL JEM IooB instrument fitted with a pointed filament and 'cool beam' gun system. 18-3
266 Short communications Fig. I. (a) Parts of two TMV rods at high magnification following preparation with I ~ glucose and 3 ~ ammonium molybdate in D20. The rod on the left shows the penetration of stain into the central hollow region of the particle whereas the rod on the right has remained unpenetrated. Although the 2"3 nm periodicity (arrows 'a' in Fig. r b) is visible along the long axis of the rods the presence of structure units forming the surface lattice of the helix is limited to a few areas (arrows 'b'). The two TMV rods appear to be well separated from each other. (b) Photographically integrated image of the rods shown in (a), produced by superimposing the negative at distances corresponding to 6"9 nm (see page 419 in Beeston et al. ~973). Considerable reinforcement of the basic helix is shown together with the tertiary helices. The particle on the right (not penetrated by stain), shows a distribution of structure units at the intersections of the primary and secondary helices. In several areas there is a single structure unit surrounded by six neighbours (arrows 'c') which corresponds closely to the surface lattice of the TMV model constructed from X-ray diffraction patterns. Some asymmetry in the appearance of the reinforced helices and structure units can be seen and results from the possible tilting of the rods with respect to the image plane (Finch, ~972). Approx. 7 structure units are visible in each striation across the rod long axis. The samples diluted to give final virus concentrations of a b o u t ~.5 m g / m l produced areas of evenly spread rods in parallel arrays. W h e n p h o t o g r a p h e d at high magnification a n d in suitably focused images, the m a j o r i t y of rods showed striations along the rod axis corres p o n d i n g to the basic 2"3 n m helix. The spectra recorded in optical diffraction patterns also indicated a layer line of 6"9 n m produced b y the repeat at every three t u r n s of the helix
Short communications 267 (Horne et al. ~975b). Under the preparative conditions we have used together with the JEM IooB electron optical system, it was possible to record the same optical diffraction spectra showing 2. 3 nm and 6"9 nm layer lines from electron micrographs of the TMV specimens after exposure to the electron beam for periods of up to about 5 min. No precautions were taken to minimize the radiation damage in any of our experiments (Williams & Fisher, ~97o; Hart & Yoshiyama, ~975). However, considerable care was taken to ensure that the contamination rate in the region of the specimen was kept to about o-~ to o'2 nm[ rain. When the contamination rate was greater than this, the visible detail present in both electron microscope images and the optical diffraction patterns decreased rapidly. The best results recorded in our electron micrographs came from TMV samples prepared with mixtures of virus diluted with deuterium oxide (D20) and mixed with I ~ glucose and an equal volume of 3 ~ ammonium molybdate, ph 6"5, also dissolved in D20. One interesting feature was the tendency of the rods to remain separated from their neighbours when seen in positions parallel to each other (Fig. I). This was in contrast to the more closely packed arrays showing evidence of the interdigitation in specimens prepared in distilled water and negative stain. There was some evidence from the electron micrographs to suggest that substructure was associated with the striations along the rod axis (Fig. I a). With the aid of the photographic integration techniques of Markham et al. 0964), we superimposed individual images of the rods as shown in Fig. I (a), at intervals of 6"9 nm, corresponding to a repeat distance of the basic helix at approx, every three turns. One of the resulting images is illustrated in Fig. I (b) and shows considerable reinforcement of the basic and tertiary helices. Individual structure units were seen associated with the primary and secondary helices as illustrated in Fig. I (a) and (b), and in certain positions along the rod axis single structure units could be discerned, surrounded by six neighbours as indicated in Fig. I (b). The average number of structure units in each striation across the axis was seven. Markham et al. 0964) obtained reinforcement of the helical components on photographically integrated TMV rods, but the primary helical pitch was given as 1.8 nm. The positions of the structure units in TMV at intersections of the primary helix (n = 0 and other dominant helices (n = 16 and 17) were previously described by Finch 0964) from negatively stained specimens. Optically filtered images of TMV rods (Klug & De Rosier, I966) also showed particles distributed over large areas of the rod which were interpreted as being subunits projecting from the body of the virus particle. The structural features shown in Fig. l (a) and (b) do not necessarily contribute any new information relating to the structure of TMV, but the technique does allow high-resolution images to be obtained over relatively long periods (~ to 5 rain), without precautions against radiation damage. Moreover, the micrographs can be recorded at magnifications sufficiently high to allow repeating features to be reinforced by very simple photographic averaging methods. At final concentrations of about 2 to 3 mg/ml, TMV formed both large and small paracrystalline arrays when mixed with 3 ~ ammonium molybdate, ph 6"5, dissolved in water or ~ ~ glucose. The electron micrographs from the higher concentrations showed many arrays of the type illustrated in Fig. 2(a), with some TMV rods viewed end on and in hexagonal array. Other rods within the arrays were seen at various angles with respect to the specimen plane. Some small isolated groups of TMV rods when seen end on showed evidence for radially arranged structure units (Fig. 2b). The techniques and results described above for the rapid formation of paracrystalline arrays of TMV and for high-resolution studies have obvious applications to optical and low-angle electron diffraction studies and will be the subject of a separate publication. The high-resolution image of the TMV rods showing the presence of structure units is certainly
268 Short communications Fig. 2. (a) Paracrystalline arrays formed at a TMV final concentration of 2 to 3 mg/ml when mixed with 3 ~ ammonium molybdate in glucose. Stereo images showed that these arrays were composed of long rods and not of protein discs. Many areas can be seen to contain rods packed hexagonally. (b) TMV particles viewed end on with regularly arranged projections (arrow). The size and distribution of the projections is consistent with those of protein structure units forming the T M V nucleocapsid.
Short communications 269 consistent with the X-ray diffraction model and electron microscope results from optically reconstructed images reported in earlier publications (cf. Caspar, 1963; Finch, I964; Klug & Berger, I964; Klug & De Rosier, 1966; Finch & Holmes, I967; Finch, 1972). We wish to thank Mr L. S. Clarke and Mr S. Frey for their help with the photographic image integration and processing. John Innes Institute Colney Lane Norwich NR 4 7UH, U.K. REFERENCES R.W. HORNE J.M. HOBART R. MARKHAM BEESFON, B. E. P., HORNE, R. W. & MARKHAM, R. (1973). In Electron Diffraction and Optical Diffraction Techniques: Practical Methods in Electron Microscopy. Edited by Audrey M. Glauert. Amsterdam: North- Holland. CASPAR, D. L. D. 0963). Assembly and stability of the tobacco mosaic virus particle. Advances in Protein Chemistry x8, 37-121. FINCH, J. T. (I964). Resolution of the substructure of tobacco mosaic virus in the electron microscope. Journal of Molecular Biology 8, 872-874. FINCH, J. T. 0972). The hand of the helix of tobacco mosaic virus. Journal of Molecular Biology 66, 291-294. FINCH, J. T. & HOLMES, K. C. (I967). Structural studies of viruses. In Methods in Virology, vol. In, pp. 35I. Edited by K. Maramorosch & H. Koprowski. New York and London: Academic Press. GOODMAN, R. M., HORNE, R. W. & HOBART, J. M. (I975a). Reconstitution of potato virus X in vitro. Characterization of the reconstituted product. Virology 68, 299-3o3. GOODMAN, R. M., MACDONALD, J. G., HORNE, R. W. & BANCROFT, J.B. (1975b). Assembly of flexuous plant viruses and other proteins. Philosophical Transactions of the Royal Society, London, B (in the press). HART, R. G. & YOSHIYAMA, J. i. (I975)- Electron microscopy with reduced beam damage to the specimen: a retractable image intensifier. Journal of Ultrastructure Research 51, 4o-45. HORNE, R. W. & PASQVALI-RONCHETTI, I. (I974). A negative stain-carbon film technique for studying viruses in the electron microscope. I. Preparative procedures for examining icosahedral and filamentous viruses. Journal of Ultrastructure Research 47, 361-383. HORNE, R. W., PASQUALI-RONCHETTI, I. & HOBART, J. M. (I975a). A negative stain-carbon film technique for studying viruses in the electron microscope. ]1. Application to adenovirus type 5. Journal of Ultrastructure Research 5x, 233-252. HORNE, R. W., HOBART, J.M. & PASQUALI-RONCHETTI, I. (1975b). Application of the negative stain-carbon technique to the study of virus particles and their components by electron microscopy. Micron 5, 233-26I. K.LUG, A. & BERGER, J. E. (I964). An optical method for the analysis of periodicities in electron micrographs, and some observations on the mechanism of negative staining. Journal of Molecular Biology xo, 564-569. KLUG, A. & DE ROSIER, D. J. (1966). Optical filtering of electron micro~aphs: Reconstruction of one-sided images. Nature, London 2x2, 29-32. MARKHAM, R., HITCHBORN, J. H., HILLS, G. J. & FREY, S. (I964). The anatomy of the tobacco mosaic virus. Virology 22, 342-359. UNWIN, P. N. T. & HENDERSON, R. (1975). Molecular structure determination by electron microscopy of unstained crystalline specimens. Journal of Molecular Biology 94, 425-440. WtLLIAMS, R. C. & FISHER, H. W. (I970). Electron microscopy of tobacco mosaic virus under conditions of minimal beam exposure. Journal of Molecular Biology 52, I2I-I23. (Received 3 November I975)