X. TMV STUDIES IN GENETIC CODING 483 



group, is permanently stable in the virus rod but decomposes concomi- 

 tantly with disaggregation of the virus. In contrast, the — SH group of 

 the isolated protein is oxidized by iodine (Fraenkel-Conrat, 1955a). The 

 — SH group of the virus does not react with p-chloromercuribenzoate, 

 but is able to bind small mercurials (e.g., methyl mercuric nitrate) and 

 the methyl mercury group is not dissociated from the \-irus by cysteine, 

 in contrast to that bound to the isolated protein (Fraenkel-Conrat, 

 1959). These observations suggest that the — SH is located near one of 

 the protein surfaces which are exposed upon dissociation of tlie virus, 

 and that the masked — SH carries an H atom which can be replaced by 

 mercurials without change in the tertiary structure. Such an — SH group 

 may be engaged in hydrogen bonding by acting as acceptor to another 

 X — H donor group. This would account for the unusual iiroperties of the 

 sulfur. 



Another type of bond probably contributing to the tendency of the 

 protein to aggregate to rod-shaped particles involves certain carboxyl 

 groups. The disaggregation of the virus has been found to be accompa- 

 nied by the release of one to two H" ions per protein unit, and, conversely, 

 H"" ions are consumed in the aggregation of the protein with or without 

 RNA. A further interesting observation supporting the existence of this 

 type of hydrogen atom is that about 2000 lead atoms are bound by the 

 virus and that the introduction of lead seems to stabilize the inter- 

 subunit bonding. It has also been observed by X-ray scattering that this 

 lead is located at two distinct cross-sectional sites in the virus particle 

 (Caspar, 1956; Klug and Caspar, 1960). On the basis of the above 

 experiments it has been suggested that there exists a particular H 

 bonding involving undissociated — COOH groups which may play an 

 important role in the build-up of the cylindrical virus structure (Fraen- 

 kel-Conrat and Narita, 1958). 



The attractive forces between the nucleic acid and the protein units 

 may be largely ionic bonds between the basic residues of the protein and 

 the phosphate groups of the nucleic acid, but hydrogen bonds and other 

 forces probably contribute to the stability of the coaggregate of the two 

 components. No definite experimental approach has yet been applied 

 to the study of these forces. 



Thus, to summarize, the architecture of the TAIV particle is the 

 consequence of protein aggregation reactions at two levels: the first, 

 the establishment of a specific chain conformation of the protein unit, 

 and the second, the aggregation of these protein units, at first by side- 

 to-side attachment and then by helical stacking into stable rods with a 

 minimal protein surface area and maximal bonding of each unit to its 

 six neighbors. 



