76 R. MAEKHAM 



accounts, no doubt, for the resistance of the \arus to tryptic digestion when 

 in its native form. 



Leucine accounts for 11 residues; serine, threonine, and glutamic acid 

 each for 14; alanine for 15; and finally aspartic acid has 17 residues. 



It will be noted that there is no histidine, and that as there are 9 arginine 

 residues and 1 free lysine e-amino group, trypsin is likely to produce only 

 11 peptides, when acting on the denatured or disaggregated protem. Nine of 

 these are likely to end with arginine, 1 with lysine and 1 with threonine. 

 This is essentially what has been found experimentally by the Berkeley 

 group, though they anticipated a larger number of peptides (11 arginine, 

 2 lysine). 



This number of peptides has also been found by Schramm and Anderer 

 (1955) who also only found 1 lysine peptide instead of the expected 2. 



The number of carboxyl groups is very large. There is 1 terminal carboxyl 

 group of threonine, and aspartic and glutamic acids provide a further 31. 

 Of these about 19 are amidated (Best and Lugg, 1944; Black and Knight, 

 1953), leaving some 13 carboxyl groups to account for the electrophoretic 

 charge. Oddly enough only 1 of these can bind safranin and 2 can bind 

 acriflavin. After carboxypeptidase action 2 safranin molecules are bound, 

 but the two sites are by no means equivalent. This might suggest that the 

 groupings involved are dissimilar, but a more probable explanation is that 

 one site is less readily accessible. Certainly there is not much likelihood that 

 the acidic groups of the nucleic acid are involved in this reaction. 



On the action of dodecyl sulfate 2 more acidic groups are liberated 

 (Harrington, Levy, and Schachman, quoted by Fraenkel-Conrat, 1957). 

 These are not due to the nucleic acid, and presumably represent carboxyl 

 groups, which may have been hydrogen-bonded, possibly to some of the 

 enormous number of hydroxyl groups (28 alcoholic, 4 phenolic) present in 

 the virus. < 



The carboxylic groups of the virus seem to be the main factors which 

 solubilize it. They are responsible for the interparticle repulsion which is 

 swamped by quite low (0.5 M) ammonium sulfate concentrations, and when 

 the groups are caused to become imdissociated, at pH's about 3, the virus 

 completely loses its solubility in water. The inference from this is that the 

 surface of the virus has relatively few hydrophylic groupings, apart from 

 the carboxyl groups and the 2 amino groups, which account for only 15 of 

 the 145 odd amino acid residues present in each subunit. 



Since the foregoing was written a detailed and careful investigation of 

 the composition of the tobacco mosaic virus protein has been published by 

 Ramachandran (1958). The figures which he has fomid differ slightly from 

 those published by Newmark and Eraser, who used a method which was 

 essentially the same. The results differ primarily in the quantities of glycine, 



