INFECTIVITY OF TOBACCO MOSAIC VIRUS 453 



iiifectivity of TMV-RNA. It would thus seem tliat tliese experiments are iu 

 accord with the existence of active subunits, which can interact witli added 

 RNA upon addition of salt, instead of aggregating with like molecules, and 

 thereby lose much of their efficiency in reconstituting active rods. From tliat 

 finding one must further conclude that packing with largely, if not exclus- 

 ively, TMV-RNA is required for the formation of active virus rods. 



When HR-RNA was used instead of yeast nucleic acids, it acted in a 

 similar manner. This nucleic acid alone reconstitutes poorly with TMV- 

 protein and, like yeast nucleic acid, etc., it acted as a depressant in mixtures 

 with active TMV-RNA. The use of P^^-labeled virus nucleic acid preparations 

 has supplied an additional tool to ascertain wiiether an infective nucleic 

 acid actually favored the joint incorporation of another inactive RNA into 

 a virus particle. However, no definite evidence for this was found. The 

 single strand hypothesis thus seems indirectly supported. 



C. Search for in Yitio-Produced Mutants 



Since it has become possible to demonstrate infectivity in degraded and 

 reconstituted virus preparations, the aim has been to produce at will a new 

 genetic (i.e., replicating) species of molecules. Reconstitution experiments 

 with protein and nucleic acid of two different TMV strains were initiated 

 prior to our realization that the nucleic acid alone was infectious and carried 

 all the genetic information (Fraenkel-Conrat and Singer, 1957). These ex- 

 periments were thus actually done in the hope of producing particles of an 

 intermediate character, and obviously had to fail in this regard. Although an 

 occasional mutant was observed, this also occurred iii similarly rare instances 

 with the progeny of the nucleic acid alone. For it seems that the chemical 

 handling of the nucleic acid per se may be slightly mutagenic. 



Contrary to these conclusions. Commoner (1957) has reported evidence 

 that virus reconstituted from TMV-protein and HR -nucleic acid produced 

 lesions of a size intermediate between those characteristic for the two strains. 

 This effect is described as "temporary, long-term infections resulting in the 

 predominance of virus which appears to follow the character of the nucleic 

 acid donor." Neither Bawden (1957) nor this author have observed such 

 fleeting intermediate effects with the same and other mixed virus systems. 



It might be advisable to interject here a brief description of the procedure 

 used by us in the search for in vitro-])Todu.ced mutants, since this differs from 

 the customary way of looking for "normal" or spontaneous plant virus 

 mutants. The virus preparation suspected of containing genetically mixed 

 particles is first applied to a local lesion host {N. qlutinosa), and the nature 

 of these lesions is observed. Single lesions (6-24) are then excised, ground up, 

 and aliquots of the homogenate are transferred to several different varieties 



