THE BIOCHEMISTRY OF PLANT VIRUSES 65 



C. Purified Preparations 

 When the plants are ground up and the sap expressed, the uniform rods 

 of virus may be isolated by several cycles of high-speed centrifugation 

 alternated with low-speed spins. Such a preparation is characteristic of 

 modern tobacco mosaic virus as it is usually investigated. If this type of 

 preparation is allowed to stand at room temperature for any time as an 

 aqueous solution, it almost invariably becomes contaminated with bacteria, 

 whereas a similar preparation made by salt precipitation, although obviously 

 aggregated, usually remains sterile. Apparently the rods, although uniform 

 in appearance, retain a superficial layer of contaminating plant material. 

 This layer is exceedingly difficult to remove, and contains, among other 

 things, plant ribonuclease, nucleic acids, and host-specific antigens (Holden 

 and Pirie, 1955; Pirie, 1956a). It should be remarked that such material 

 constitutes but little of the total mass of the preparations, but its presence 

 may account for some of the anomalous behavior of the virus suspensions. 

 There is some indication that the contamination tends to be concentrated 

 at the ends of the virus rods, and so effectively prevents end-to-end aggrega- 

 tion. Certainly treatments which tend to remove the contaminants usually 

 result in such an aggregation, which is characteristic of the more drastically 

 treated preparations. This aggregation, which is evidently an extension of 

 the crystalline structure of the virus rods themselves (Bernal and Fankuchen, 

 1941a), is extremely difficult to reverse completely; a uniform suspension of 

 rods cannot be obtained from such preparations, although, of course, greatly 

 elongated rods are unstable even to small hydrodynamic forces. Larger 

 forces will even disrupt the short virus rods (Oster, 1947). 



Aggregation of the virus is the almost invariable consequence of the 

 treatment of preparations with proteolytic enzymes or with salt or acid 

 sufficient to precipitate the virus from solution. 



Some substances, such as the strongly basic protein, pancreatic ribo- 

 nuclease, combine extremely strongly with the virus, which is essentially 

 acidic, and cause the formation of long fibres (Loring, 1942). When such an 

 aggregation takes place, a secondary effect is also noted; that is the formation 

 of factoids (Bernal and Fankuchen, 1941a). These are cigar-shaped droplets 

 of virus which have a higher concentration of protein than that in the bulk 

 of the fluid, and these sink to the bottom of the container, giving rise to 

 the "bottom layer," first noticed by Bawden and Pirie (1937a). These 

 tactoids, which are equivalent to the "crystals" obtained by Stanley, are 

 aggregates of rods having variable lengths, and so ordered that the long 

 axes of the rods tend to be parallel to the long axis of the tactoid, although 

 the shape of the tactoid precludes perfect parallelism throughout. In a 

 plane at right angles to their length the rods are arranged in a two-dimensional, 

 hexagonal, close packing, which is maintained even when the particles are 



VOL. II — 5 



