THE BIOCHEMISTRY OF PLANT VIRUSES 47 



on tliis same virus. Carboxypeptidase is an enzyme having ea:;opeptid.ase 

 activity, that is to say, it attacks the terminal amino acids in a peptide 

 chain in sequence, and this particular enzyme attacks the end of the chain 

 which has a free — COOH group. Rather to their surprise, Harris and Knight 

 (1955) found that only threonine was liberated, and in fact this amounted to 

 some 2900 residues per 50,000,000 molecular weight or, if the somewhat 

 lower estimate of molecular weight favored by some authorities is taken, 

 some 2300-2400 residues of this one amino acid are liberated. Assuming that 

 each submiit of the virus yields one and only one residue of threonine, this 

 gives a subunit size of some 17,000 molecular weight, a size which although 

 large is still within the range which one might visualize tackling by the 

 methods either available at present or being developed. Moreover, this size 

 agrees well with the values estimated from the cysteine content of the virus 

 protein (M = 18,000). The main difficulty which arises is that one has to 

 assume that the virus is made up from only one type of subunit and that all 

 these submiits are identical. That this is an unlikely assumption is, of course, 

 obvious because the isotopic distribution of the atoms constituting the 

 structure precludes absolute identity. But if one ignores tliis obvious limita- 

 tion, what are the chances that the subunits are closely similar? In fact, it 

 might appear extremely likely that the polypeptide subunits of a virus 

 should be limited in variety; it is not unlikely that the subunits may be of 

 one type only, and, if one considers that the only obvious function of the 

 protein part of, say, tobacco mosaic virus is protective, it is likely that they 

 are. Furthermore, as has been noted, the chemical evidence to date indicates 

 that this is so, as well as do the X-ray diffraction data. However, it is not 

 unthinkable that minor irregularities do exist. In fact, it would be 

 surprising if they did not. Also, it is possible that the very ends of the virus 

 rods may be different in some degree from the rest, although the great 

 tendency of the virus to aggregate in a linear direction might suggest the 

 opposite. 



There is, of course, a certain indication that the virus rods are in fact not 

 uniform. For example, Harrington and Schachman (1956), studying the 

 degradation of tobacco mosaic virus by alkali, found that a small proportion 

 of particles in any one preparation are much more resistant to the action of 

 the alkali than are the remaining particles. Furthermore, they found that 

 the alkalme degradation of the particles was more rapid at the beginning, and 

 that when the particles had been degraded to about one-third of their original 

 length, the remainder was much more resistant to the chemical action. This 

 latter obser\'ation is, of course, by no means conclusive evidence of the 

 essential difference of the remaining fragment of the rod, but it is suggestive. 

 However, if one were to have to give a dispassionate summing-up of the 

 present situation, it would be that there is a chance that by making the 



