Some Observations on the Structure of Tobacco Mosaic Virus 



H. E. Huxley 



Medical Research Council, Department of Biophysics, University College, London 



X-RAY diffraction studies of oriented preparations of 

 tobacco mosaic virus have made it possible to 

 elucidate the internal structure of the molecule with 

 a very high degree of detail (1, 2, 3, 6). Electron 

 microscope observations, on the other hand, have, 

 until recently, revealed only that the molecules are 

 rod shaped, about 150 A in diameter and 3000 A 

 long, and have shown nothing of any internal struc- 

 ture in these molecules. During the initial period 

 of operation of the Siemens Elmiskop I at University 

 College, the question was considered of whether the 

 very high resolution now available with that instru- 

 ment would make it immediately possible to see 

 structural regularities within the TMV molecules, 

 and indeed, within other large biological molecules 

 and filaments which previously had appeared struc- 

 tureless. 



A very dilute solution of tobacco mosaic virus 

 was placed on carbon-filmed electron-microscope 

 grids, and stained with 40 % phosphotungstic acid 

 in the manner described by Hall (4). The results at 

 first seemed disappointing, for even when the resolu- 

 tion was better than 10 A, no regular internal struc- 

 ture was visible in the particles. However, a curious 

 effect came to light which did reveal one feature of 

 the internal structure. 



In some areas of the grids it was sometimes appa- 

 rent that the excess stain had not been completely 

 removed by washing before the preparation was 

 allowed to dry. In these areas, the particles became 

 outlined by stain in a very distinctive manner. A 

 similar effect has previously been noted by Hall (4) 

 with bushy stunt virus. In addition to the outline, 



however, many of the particles now showed a dark 

 line running along their long axes (fig. I). In the 

 longer particles, this line was frequently much better 

 defined at the end of the particle than near the 

 centre (fig. I). The line was indistinct or invisible in 

 particles which had not been outlined by the stain. 



It was found that this outlining effect could be 

 produced not only by phosphotungstic acid, but 

 also by 0.1 Af potassium chloride, or 0.1 M bicar- 

 bonate buffer; it seemed to be sufficient merely to 

 have a little salt of some sort present when the prepa- 

 ration was allowed to dry. The extent of the effect 

 was found to vary very considerably from area to 

 area on the grids, and from one batch of grids to 

 the next, and is presumably related to the wetting 

 properties of the supporting film. 



Preparations of virus degraded into short lengths 

 in bicarbonate buffer at pH 10.3 (5) also showed 

 particles with a dark line down their centre when 

 treated by the above techniques. Very short rods, 

 standing on end, are often observed in such prepara- 

 tions; these rods appear to be accurately cylindrical 

 in section, and the central core is readily seen (figs. 2 

 and 3). 



Preparations of virus which have been very thor- 

 oughly washed (fig. 4) do not appear "outlined" in 

 this way and it seems that both the PTA or the salt 

 is easily removable. 



The most likely explanation of this effect is in 

 terms of the structure, proposed by Franklin on the 

 basis of x-ray studies, in which the virus rods have 

 a hollow core. Thus, if a TMV particle is allowed to 

 dry in a pool of dilute salt solution, the salt will tend 



Fig. 1. Tobacco mosaic virus, outlined by drying in very 

 dilute solution of potassium chloride (see text). 

 Magnification 140,000. 



Fig. 2. TMV degraded in alkaline solution and outlined in 

 phosphotungstic acid. Magnification 140,000. 



