VIII. PROTEIN SYNTHESIS AND GENE ACTION 361 



However, the possibility of a DNA-like structure in localized regions of 

 all T-RNA molecules still remains. From the studies cited above it 

 appears that attempts to study the code by isolating specific T-RNA 

 fractions will be difficult. 



Several other complications reinforce this viewpoint. While the 

 evidence points to a particular T-RNA chain being able to accept only 

 a single type of natural amino acid, it appears that more than one kind 

 of chain can code for the same amino acid. This was demonstrated most 

 clearly by the finding that at least 20% of E. coli leucyl-RNA had a 

 different base sequence than the rest (Berg and Lagerkvist, 1962). 

 Separation of two fractions from rat liver soluble RNA with leucine- 

 acceptor activity has also been reported (Doctor et al., 1961). Similar 

 results have been reported when T-RNA's from different species have 

 been compared. In general, these experiments have involved studies of 

 the formation of various amino acyl-RNA compounds using T-RNA 

 from one species and crude activating enzyme preparations from another. 

 Thus, different amounts of methionine-RNA were formed using E. coli 

 T-RNA depending on whether E. coli or yeast enzymes were used (Berg 

 et al., 1961), and guinea pig liver enzymes were quite poor for forming 

 leucyl-RNA with E. coli T-RNA, although RNA preparations from a 

 number of mammalian sources were approximately equivalent (Allen 

 et al., 1960). The most detailed study of the species specificity of T-RNA 

 has shown that the amount of a particular amino acid attached to a 

 particular RNA varies greatly, depending on the sj^ecics from which the 

 activating enzymes are derived and which amino acid is studied (Benzer 

 and Weisblum, 1961). These results with crude enzyme systems are less 

 certain than those involving separation of RNA types. However, it is 

 clear that differences do exist among RNA chains specific for a single 

 amino acid. 



These differences may result from a non-universal code. That is, the 

 "coding" base sequence for leucine in T-RNA from rabbits and E. coli 

 may be different. A more likely possibility is that the specificity meas- 

 ured in the ''species" experiments is for the reaction of T-RNA with the 

 specific amino acyl-AMP-enzyme complex. Species specificity in this 

 case might be expected. The discussion above indicates that a large part, 

 if not the complete T-RNA molecule, is required for the formation of 

 amino acyl-RNA. A smaller part of the T-RNA molecule is expected to 

 form the coding unit (see below), and this may not be the only factor 

 in the reaction with activating enzymes. This explanation for hetero- 

 geneity in T-RNA chains specific for a single amino acid does not apply 

 in the case where two leucyl-RNA chains are found in the T-RNA of 

 the same cell and both of these appear to react equally well with a single 



