THE BIOCHEMISTRY OF PLANT VIRUSES 89 



low phosphorus content of the virus found by these authors (0.45 %), it 

 should be about 2 million. This discrepancy might well be due to the partial 

 breakdown of the nucleic acid during its isolation. For example, Ginoza (1958), 

 examining the stability of the virus nucleic acid to heating, as evidenced by 

 its ability to cause infection, found that it was indeed very unstable, although 

 his results do not preclude the possibility of the inactivation being due to 

 a heat-stable enzyme, like plant ribonuclease, which was breaking down his 

 nucleic acid, rather than a purely chemical type of hydrolysis as he assumed. 

 The possibility that the nucleic acid might be linked to the virus by very 

 weak covalent bonds such as tertiary phosphate ester links is one which 

 has attracted much attention. It does seem very unlikely, now, that such 

 linkages can occur. Koshland et at. (1957) degraded the virus in the presence 

 of HgO^^ by means of detergent and found no 0^^ in the ribonucleic acid 

 released. 



N. The Effect of the Host on the Virus 



Many plant viruses are adapted to parasitize more than one host. Tobacco 

 mosaic virus is no exception, and a considerable time ago it was thought of 

 some interest to determine whether the virus was modified m any way by 

 the association with the host on which it was growing. It is well known, for 

 example, that in the case of other plant viruses a change of host may cause a 

 permanent change in the symptom pattern produced by the virus, and 

 while it was originally thought that this ij\)Q of change was a host-induced 

 modification, it is now fairly certain that the host acts by selecting mutants 

 of a type which are more suited to multiplication on it. This type of change is, 

 of course, certain to be reflected in a change in the chemical composition 

 of the virus, although it may be so small as to escape detection. Similarly, 

 chemical changes might be produced without an appreciable change in the 

 symptom pattern. 



The first attempt to determine whether there were any host-induced 

 variations in a virus were made by Loring and Stanley (1937), who compared 

 tobacco mosiac virus grown in tomato plants and in tobacco plants. Apart 

 from minor differences in the physical properties, which are not altogether 

 unexpected these days, these workers found little difference between their 

 preparations. What chemical analyses were made were limited to elementary 

 analyses, and it is quite evident that only very gross differences could be 

 detected by the techniques used. Moreover, the two plants used in the 

 investigation were closely related botanicaUy, so that the system was not 

 the most likely one to show up any differences due to the host. 



A much more convincing system was studied by Gaw and Stanley (1947), 

 who compared two distinct viruses, the "t}TDe" tobacco mosaic virus and 

 the ribgrass virus, grown on tobacco and on Phlox drmnmondii, a plant only 



