THE BIOCHEMISTRY OF PLANT VIRUSES 109 



number of viruses which seem to have little in common except that they 

 belong to the "spherical," small viruses and have similar host ranges. The 

 group is very interesting but rather difficult to work with, because the quantity 

 of virus obtained from infected plants is small, being of the order of 1 to 

 50 mg. /liter of sap of tobacco leaves. 



The tobacco necrosis viruses have a very wide host range, and are able to 

 infect the roots of a large number of plants without causing much damage. 

 It is only under exceptional climatic conditions that they infect the plants 

 above the root region, but in that case they produce a very severe necrotic 

 disease. 



A. Purification 



The virus was first purified by Pirie and associates (1938), who took 

 advantage of the stability of the virus to ethanol for clarification and pre- 

 cipitation (75 % ethanol was used, which is probably not suitable for all 

 isolates of the virus) and this was followed by precipitation with ammonium 

 sulfate. Finally the virus was crystallized in part as flat plates by the cautious 

 addition of ammonium sulfate at 0°C. On standing at this temperature the 

 virus crystallizes out (Fig. 20). The material was examined by the ultra- 

 centrifuge, and was found to contain several components. Chemical analysis 

 showed that it was nucleoprotein in nature and contained some 14-18 % of 

 nucleic acid of the ribose type. It also had the expected ultraviolet absorption 

 of a nucleoprotein. 



Some time later the problem was reinvestigated by Bawden and Pirie 

 (1942), who multiplied up local lesion isolates obtained from various sources. 

 These workers found that they were working with a group of different 

 viruses, several of which could be traced to the earlier preparations because 

 of the existence of antisera made to the latter. So undoubtedly the virus of 

 Pirie and associates (1938) was a mixture. This is not generally found in 

 plant virus work, where under such circumstances one strain will dominate the 

 others after a while; but in this case the viruses in the mixture were unrelated, 

 and as they caused only local lesions no interference between strains could 

 take place. 



Bawden and Pirie (1942) managed to characterize several of their isolates. 

 They were named according to their origin, for example. Tobacco 1, 2, 3, 

 and so on, Potato, Princeton, Eothamsted. The isolates Tobacco 1 and 2 are 

 probably very similar, and crystallize as rhombic dodecahedra and also as 

 bipyramids. Princeton, like many of the Cambridge isolates, crystallizes as 

 flat plates, and most of these strains have sedimentation coefficients of 

 about 115 S. (Price and Wyckoff, 1939) Rothamsted and the Dutch stipple 

 streak are smaller and have s = 50 ^. 



