436 



CORRADO BACILIONl 



base chaiijic in the pi-oposcd code. Smith (]9()21)) lias shown that if an 

 arbitrary so([iK'ni'o is assif^nt'd to the triplet cotliny for a given amino 

 acid (i.e., the ba.sc sequcnee for glutamic acid is a.ssumed to be IWCJ). 

 the se(|Uence of the trijilets coding for 18 amino acids can be established 

 utilizing the known amino acid sul)stitutions in abnormal hemoglobins 

 and in nitrous acid mutants of tobacco mosaic virus (see Table III). 

 It is possible to construct in this way six sets of codes, based on the six 

 permutations of the seciuencc arbitrarily assigned; one of these is likely 

 to represent the correct code in messenger RNA. However, for some 

 amino acids two or three alternative coding units have been rejjorted 

 (Matthaei et al., 1962; Speyer et al., 1962). The code appears thus to be 

 degenerate, as suggested also by the elegant genetic exjieriments of 

 Crick et al. (1961). 



In view of tlie degeneracy of the code it is of some interest to examine 

 in detail instances of multiple amino acid replacements of the same resi- 

 due. These replacements are evidently produced by different changes 

 of the same coding units. Three double replacements have been found in 

 the ^-peptide chain of Hb-A (for references and explanations see Tables 

 II and nil: 



Glu 



Val 



Lvs 



UAG 



UUG 



UAA 





^ Zurich g3 



His 



Tyr 



Arg 



UAC 



UAU 



UGC 



D Punjab 

 /3 121 



Arabia 



121 



Glu 



Glu-NH., 



UAG 



-UCG 



UAA 



These rejilacements are in complete agreement with the jiroposed genetic 

 code (Smith, 1962b). 



