X. TMV STUDIES IN GENETIC CODING 501 



better than this (less than ±2%) can generally be obtained with the 

 analyzer and repeat analyses performed on each of two hydrolyzates 

 can supply enough data to establish a reasonable degree of certainty. 



In contrast to our primary use of direct analysis (Tsugita and 

 Fraenkel-Conrat, 1960) Wittmann continues to determine the composi- 

 tion of strains by the summation of the analyses of the peptide fragments 

 obtained by trypsin digestion. That approach is less dependent on high 

 accuracy in the analytical techniques, but the requirement that each 

 peptide be obtained in a state of high purity makes this method laborious 

 and expensive. The fact that the analytical results for common TMV 

 obtained with the whole protein are in agreement with those obtained by 

 peptide summation (see Section II,B) validates both methods; the direct 

 method, being the simpler, thus appears preferable for the primary sur- 

 vey of a large number of mutant proteins. 



For mutants showing differences in composition from the wild type 

 the question arises as to which particular amino acids are exchanged 

 along the peptide sequence of the protein. Obviously it is important to 

 establish the nature and the location of an exchange of amino acids, 

 including an identification of glutamine and asparagine as contrasted 

 to glutamic and aspartic acid. 



For this purpose 20-300 mg of the proteins were digested by trypsin 

 in the usual manner. From the digest the isoelectrically insoluble peptide 

 was separated, and the soluble peptides were fractionated on Dowex 

 1X2, using a volatile solvent system similar to that used by Wittmann 

 and Braunitzer (1959), as shown in the caption of Fig. 6. A typical pep- 

 tide elution pattern is shown in that figure. A few of the peaks appeared 

 to consist of a single peptide as shown by rechromatography and analyti- 

 cal composition. Others were mixtures, but upon analysis of aliquots 

 from the rising or falling edge of the curve the composition of the pure 

 components was often closely approximated. For preparative isolation 

 and definitive analyses such mixtures were rechromatographed. For the 

 purpose of amino acid analysis of peptides the 50-cm and 5-cm columns 

 of the analyzer were usually used rather than the 1 50-cm and 15-cm 

 columns. This procedure has the advantage that the time for analysis is 

 but one-third that of the normal procedure and that only one-tenth as 

 large a sample is required (0.01-0.3 ^umoles), without much loss in 

 accuracy (d=5%) (Tsugita, 1962a). 



Any peptide showing a change in composition from that of the 

 corresponding peptide of the normal strain was subjected to the usual 

 variety of techniques of protein chemistry, i.e., digestion by proteinases 

 or peptidases and chemical methods including dinitrophenylation, phen- 

 ylisothiocyanation. and hydrazinolysis. Thus, the localization of the 



