436 H. FRAENKEL-CONRAT 



protein, which aggregates below pH 5 to rods of uniform diameter but of 

 greatly varying lengths, often exceeding that of the virus and thus consist- 

 ing of more than the approximately 2300 peptide chains which build up 

 one virus particle. This aggregation of the protein is associated with a 

 marked increase m its electrophoretic mobility, which then approaches that 

 of the complete virus. 



While the aggregation of the protein is readily reversible by changes in 

 pH or ionic strength, it becomes stabilized when viral nucleic acid (5 % of 

 the protein) is also present, and becomes incorporated. In regard to heat 

 denaturation, proteolytic enzymes, bacterial attack, and other agents, the 

 resultant virus resembles the undegraded virus in being appreciably more 

 resistant than the isolated protein. Thus, the result of aggregation is not 

 only a protective shell for the nucleic acid but a self-stabilizing protein 

 structure — a most fascinating case of biological adaptation, and a challenge 

 to the protein chemist. 



Some studies of the nature of the linkages which hold the subimits together 

 have been initiated. It appears that the — SH groups (one per peptide chain) 

 are masked in the sense that they react with some but not with other thiol 

 reagents. The finding (Fraenkel-Conrat, 1955, 1957a) that these groups 

 undergo substitution when they do react, 



[— SH + O3N— Hg— CH3-> — S— Hg— CH3(-f HNO3), and 



— SH + I2 ^ SI( + HI)] 



and that these reactions proceed without loss in the stability of the native 

 protein structure, mdicates that their H atom is piesent and unlinked, and 

 that the bonding which prevents them from reacting like typical thiols must 

 connect the sulfur with some other protein group, Tliis finding is of general 

 importance for the problem of the nature of masked — SH groups. However, 

 there is no definite evidence that bonds involving the sulfur atom actually 

 participate in inter-subunit Imking. If they do participate in the interaction 

 of native iS submiits, then they must form new bonds within those units 

 once the virus is degraded to this state, for they retain their masked char- 

 acter, althougli somewhat weakened, in the iS protein, unless it be irrevers- 

 ibly denatured by detergents, heat, or other agents. 



A similar conclusion can be reached concerning certain phenoHc groups. 

 For there is chemical and/or spectrophotometric evidence that these are 

 masked in both the intact virus and the isolated protein; in the latter case, 

 the masking is again less pronomiced, and is a reversible, pH-dependent 

 phenomenon. Thus, it is either not due to the same bonds as in the original 

 virus, or it does not contribute to the inter-subunit bonds stabilizing the 

 architecture of the virus. 



As has long been known, all methods of degrading the virus architecture 



