CHEMICAL STUDIES ON VIRUSES — STANLEY 363 



of fact, this virus has a preponderance of dicarboxylic amino acids ; 

 hence the protein is acidic in nature. The nucleic acid, comprising 6 

 percent of the virus structure, was fomid to be composed entirely of 

 ribonucleic acid. Because of the acidic nature of the protein com- 

 ponent the nature of the linkage between nucleic acid and protein has 

 elicited much interest. It appears likely that the guanidino groups 

 of arginine are so arranged as to be available for linkage to nucleic 

 acid and that this arrangement is much more specific than similar 

 linkages in the sperm nucleoproteins. 



When the nucleic acid is split from the virus under conditions in 

 which presumably only hydrogen bonds are broken, the protein com- 

 ponent can yield a homogeneous protein subunit which appears to have 

 a molecular weight of about 18,000 to 20,000. It has also proved pos- 

 sible to obtain a nucleic acid-containing degradation product hav- 

 ing a molecular weight of about 360,000 which, although inactive, 

 can nevertheless react serologically in a manner similar to the original 

 virus and can reaggregate to form long rods having the characteristic 

 diameter of 15 mfi. It may also prove meaningful that a subunit 

 preparation of molecular weight about 120,000 and having no nucleic 

 acid can reaggregate in a similar manner to form rods of 15-m/A 

 diameter. This result may indicate that the protein alone can carry 

 the specific information necessary to enable formation of a char- 

 acteristically shaped rod. A noninfectious protein has been isolated 

 from diseased plants which appears to be similar to this subunit in 

 that it can likewise aggregate to a rod having a diameter of 15 mfx. 

 Although none of these aggregation products has been found to 

 possess virus activity, the fact that a protein can aggregate to form 

 a definite and highly specific structure may prove of considerable 

 importance in connection with the elucidation of chemical structure 

 characteristic of this virus. Furthermore, the fact that this subunit 

 appears to consist itself of about 6 to 8 sub-subunits of about 18,000 

 molecular weight provides a direct entree to structural work of im- 

 portance to the organic chemist, for peptide chains of that size have 

 now proved susceptible of direct structural analysis and synthesis. 



Some results of Knight and Harris in the Virus Laboratory are of 

 interest in connection with structural considerations. These workers 

 found that on treatment of tobacco mosaic virus with the enzyme 

 carboxypeptidase about 2,600 threonine residues per mole of virus 

 were released without causing any change in biological activity. If 

 each threonine residue is considered to represent one peptide chain 

 in the virus, the resulting subunit would have a molecular weight 

 of about 18,000. This subunit does not appear to consist of simple 

 chains having an equivalent number of N-terminal amino acid resi- 

 dues, for extensive search for those groupings by Fraenkel-Conrat 



