Electron microscopy of satellite tobacco mosaic virus crystals: metal-coated, negatively stained and stereo pairs

Highly purified virions of satellite tobacco mosaic virus (STMV) were found to crystallize at relatively low concentrations (300-500 \ig ml~ l ) in pure water. Small crystals of these preparations were examined in the transmission electron microscope after either being rotary shadowcast with metal or negatively stained with 4% uranyl acetate. Stereo views were also obtained of both types of preparations. Stereo pairs of metal-coated crystals provided good three-dimensional images. When stereo pairs of negatively stained crystals were printed from second negatives, they provided striking images although the three-dimensional aspect was not so pronounced. Images of both types of preparations were compared with a computer-generated model of the virus. This model was based on data obtained in earlier X-ray diffraction crystallographic studies. Measurements of crystal axes on the EM images were somewhat lower than those of the computer model. It is assumed the reason for this is the dehydration of crystals during preparation for electron microscopy. The EM images did verify the type of crystal lattice determined in the X-ray diffraction studies. Conversely, knowing the exact unit cell parameters and the distribution of virions in the crystal from X-ray diffraction data aids in the further interpretation of electron micrographs of virus crystals.


Introduction
Satellite tobacco mosaic virus (STMV) is interesting in a number of ways, one of which is the fact that it has isometric virions whereas its helper virus (TMV) has virions with helical symmetry. It has been quite well characterized biologically [1][2][3][4]. Although found naturally occurring in Nicotiana glauca with the U5 strain of TMV (also known as tobacco mild green mosaic virus, TMGMV) acting as the helper virus, several tobamoviruses can also serve as helper viruses [4].
In tobacco leaves co-infected with TMV and STMV, virions of the latter were found only in cells which had helper virus (TMV) virions. The STMV virions were associated with cytopathic structures that were distinct from those induced by the helper virus. One of the most striking of these structures was a crystal of STMV particles which was associated with two granular inclusion bodies. Both of these structures were bound by a common membrane [5].
A number of crystallographic studies utilizing X-ray diffraction and quasi-elastic light scattering techniques have resulted in a very complete knowledge of STMV virion structure [6][7][8][9]. Laboratory produced crystals have permitted elucidation of virion structure at a 2.9 A resolution while more perfect crystals produced in the Microgravity Laboratory have allowed a 1.8 A resolution. Recently, a structural comparison of STMV was made with satellite tobacco necrosis virus (STNV) and satellite panicum mosaic virus (SPMV) [10]. Although the structural analysis showed a common function and the preservation of a 'jelly-roll' motif in the protein subunit for the three viruses, there were remarkable differences in the arrangement of secondary structural elements, interaction of adjacent subunits and the disposition of the subunits with respect to the icosahedral axes of symmetry. In all three viruses it was shown that the narrow end of the 'jelly-roll' forms fivefold contacts although the nature of these contacts was shown to be different for each of the three viruses. Visibility of nucleic acid in the electron density maps was found only in the case of STMV, and this was manifested as double-helical segments of RNA associated with each coat protein dimer.
The nucleotide sequence has been determined for STMV and the translation of its RNA has been accomplished [11]. Several naturally occurring isolates of STMV have been obtained from the field which differ in characteristics such as mobility of their dsRNAs [2,12] and genotype [13,14]. One of these isolates designated as strain 8, was of special interest for EM since it readily formed crystals in pure water at moderate concentrations (500 ng ml" J ). Such crystalline preparations of this strain were utilized in the present studies. Wyckoff first published electron micrographs of metal-coated virus crystals in 1949 [15]. In this report we present electron micrographs of metalcoated (rotary shadowcast) and negatively stained crystals of STMV including stereo pairs of both preparations. We also present stereo pictures of a computer model of a STMV crystal.

Methods
Virus purification STMV (strain 8) was extracted from leaves of Nicotiana tabacum L. Xanthi co-infected with TMV-U5 by the purification protocol described in reference 4. However, to obtain a highly purified virus preparation, sucrose density gradient centrifugation (10-40%) was performed, fol- lowed by high speed centrifugation of the gradient fraction containing primarily STMV and the final pellet was resuspended in sterile water. This preparation was used in the electron microscope studies at concentrations of 300-500 Hg ml" 1 in pure water. Since strain 8 of STMV formed crystals readily in pure water, no crystallization technique per se was used.

Electron microscope grids
Copper grids (100 mesh) with heavy carbon coated formvar films were used. The grids were exposed to a plasma glow treatment [16] shortly before use to facilitate deposition of the virus crystals which were pipetted onto the carbon coated surface of the grid. Since the STMV crystals were in pure water, it was not necessary to include a wash step before either rotary shadow casting with platinum-palladium alloy or negative staining with 4% aqueous uranyl acetate [17,18].

Electron microscope calibration
After pictures of experimental preparations were taken at a given magnification, additional pictures of a twodirectional replica grating [19,20] were taken to provide a size standard. The replica grating used had 2150 lines mm" 1 and lines were in two direaions each perpendicular to the other. Use of the manufacturer's scale on the resulting photographs provided an accurate determination of the magnification.

Photography of stereo pairs
The goniometer stage on the Hitachi H600 transmission electron microscope was used to obtain stereo pairs of photographs. To obtain the 'right'-and 'left'-hand views of the stereo pairs, an angle of 4° was used between the two pictures of the pair. A crystal to be photographed was first found at zero tilt of the stage. The stage was then tilted either plus 2° or minus 2°, recentred, refocused, and photographed. The stage was then tilted 2° from zero in the opposite direction (either plus or minus), the crystal recentred, refocused, and photographed. By observing the pair with a stereo viewer one could readily determine the 'right' and 'left' views of the pair [21,22].
Photography of final electron microscope images Second negatives of pictures of metal-coated specimens are commonly utilized to give a desirable perspective view of the specimen in the final photographic print. This is obtained by making a 'contact print' of the original negative from the electron microscope onto another sheet of EM negative film. This 'second' negative is then used for the final photographic print. Pictures of negatively stained electron microscope preparations are generally printed from the original negative obtained in the microscope. However, we found that a more striking photographic image of the negatively stained crystals was obtained if a second negative was prepared. Therefore, such negatives were used for the final photographs of negatively stained virus crystals.
Computational techniques A computer model of the STMV crystal was generated on an SGI 320 VGX workstation. A file containing coordinates for each of the virus panicles in the crystal composed of 500 unit cells (10X10X5 cells) was prepared in a RIBBONS atom file format [23]. Each virion was, thus, represented as a sphere or atom. In the model, the spheres are separated according to the packing and space group parameters determined by X-ray diffraction [24].

Results and discussion
Metal-coated crystals Large crystals, not shown, were too dense to the electron beam to permit the definition of individual virions on the surface of the crystal. The virus suspension that was deposited on the grids often had clusters of small crystals closely associated with one another. Figures 1 and 2a are of two such small crystals of STMV strain 8 which have been rotary shadow cast with a platinum-palladium alloy. These might be considered as 'incomplete' crystals as virions appear to be still depositing on the surface of the crystal. By careful observation one can note that virions in a layer below the top layer appear to be in lines that are orientated at an angle to the lines of virions in the top layer. Perhaps the virions in the lower layer are in the interior of the crystal unit cell.
Negatively stained crystals The micrographs in Fig. 2b illustrate the images of negatively stained crystals when the photograph is printed from a second negative. Sets of white lines perpendicular to one another represent virions in one 'layer' of the crystal. The second set of white lines which are at an angle of 45° to the upper layer represent the next lower 'layer' of virions in the crystal. Since the picture was made from a second negative, one must keep in mind that the white lines are actually deposits of uranyl acetate between the rows of virions.

Stereo micrographs of crystals
The stereo pair represented in Fig. 2a (metal-coated  crystal), when viewed with a stereo viewer, provides a three-dimensional view of a crystal. An impression that virions are still being deposited on the surface of the crystal is best seen when the crystal is viewed in stereo.
The three-dimensional aspect of the stereo pair in Fig.  2a is more apparent than in the pair of a negatively stained crystal in Fig. 2b. However, the depth of the stereo image of the negatively stained preparation makes interpretation of the angles and depths between successive layers of virions easier. Stereo views of the crystals were only possible with small crystals. Replica techniques, which were not used, may have allowed for stereo images of electron dense large crystals.
can actually be seen in Fig. 2b, and it is schematically represented in Fig. 3.
We have carefully measured distances between viral particles as seen on electron micrographs. We have taken into account the fact that during these measurements we can only see two of the unit cell edges. The distances between viral particles could usually be correlated with the crystallographic data according to the ratios of their lengths. Systematically, the measured values have been approximately 10-15% lower than those obtained by crystallographic analysis. We presume that this is because of the evacuation and dehydration procedure employed to prepare samples on grids for electron-microscopic analysis. Comparisons of computer model with electron The ima 8 es of metal-coated crystals resemble somewhat micrograph images those obtained from tobacco necrosis virus crystals and , -,, ..-•.-southern bean mosaic virus crystals by Wyckoff [15]. The Knowing the exact unit cell parameters and distribution . , ... , ,, ,., , " , . , , ,.,..., , , ,, ," . micrographs published by Wyckoff were obtained by using of viral particles in the crystal obtained by X-ray diffraction , ,. , , ' • . ,. , ,." , ,.
. , larger intact crystals, seen to assume regular repetitive crystalline arrangement. Furthermore, it can be seen that in every case the crystals are viewed perpendicular to one of the crystallographic Concluding remarks faces. Since we are dealing with the body centred crystal-Horne [17] used the neg ative stain technique to elucidate line lattice, all particles in the second layer should be the ultrastructure of crystal lattices of a number of specishifted with respect to the upper layer by j of the me ns, and he also used the technique to demonstrate the crystalline period in the direction of all crystallographic production of crystalline arrays of virus particles on the axes. This arrangement is illustrated in Fig. 3. This results surface of carbon-coated mica support films [18]. We in a particle being positioned right in the centre and believe that this report is the first to utilize the negative underneath each of the four particles seen in the upper sta i n technique on actual virus crystals. As stated above, layer of the crystal. This can be seen on the metal-coated tne production of a second negative of images with electron micrographs (Figs 1 and 2a) where several single this technique provides striking contrast between various viral particles are located on the surface of the crystal, layers of small crystals. The technique does, however. These particles always touch four of the other particles po int out the great need to make every effort to avoid on the layer beneath. These particles appear larger since photographic and interpretive artifacts when applied to they are exposed and accumulate more metal during the biological specimens as pointed out by Wergin and Pooley coating procedure. In the case of negatively stained crystals [25]. Although Hayat and Miller [26] have described (Fig. 2b), regular patterns can be seen. These patterns several methods for making carbon-coated support films appear as four sets of intersecting parallel lines rotated more hydrophilic, they did not seem to consider the approximately 45° with respect to each other. These plasma glow technique recommended by Jakstys [16]. striations can be seen as a consequence of the distribution We have found that the latter method provides the best of the virions in the crystal lattice. The particles in the hydrophilic surfaces for deposition of specimens which crystal can be imagined to lay on rectangular nets (almost are to be negatively stained, square, since the ratios of the three crystallographic axes are 1:1.1:1.2). If the space between virus particles in the f crystal receives a small amount of uranyl acetate, then _. ". dense areas under the electron microscope will appear as ' Valverde R A and Dodds J A (1986) Evidence for a satellite RNA , ,.,... are connected along the square edge or the square diagonal > ££,' ^j^^^E^^^f *"*** (which would make the distance 1.4 times greater). In 7vaive r deRX'He^J A7™^t he second case the series of uranyl acetate deposits would satellite tobacco mosaic virus, helper virus, and their hosts, be further apart, and lines would appear fainter. This .