F. SANGER 



Both oxytocin and partial hydrolyzates of insulin exhibit 

 strong activity as strepogenin, a growth factor for Lactobacillus 

 casei (29). Oxytocin and insulin have no amino acid sequences 

 in common, the only common feature being the disulfide ring, 

 which may thus again assume a biological role. 



The structure given in Figure 4 is deduced as the only one 

 that is consistent with all the experimental findings, assuming 

 certain general principles of protein chemistry such as the 

 peptide hypothesis. Although our results do not prove these 

 principles, they do lend considerable support to them. If 

 indeed insulin were not composed solely of a-amino acids bound 

 together by peptide linkages and disulfide bonds, one would have 

 expected to find some inconsistency in the results, and it would 

 probably not have been possible to deduce a unique structure. 



Various workers have suggested from time to time that other 

 types of structures and bonds such as 7-glutamyl bonds and 

 diketopiperazine structures may be present in small amounts in 

 proteins. Such structures do exist in some of the bacterial 

 polypeptides, but the evidence for their occurrence in proteins is 

 based largely on the results of rather violent degradative pro- 

 cedures where it is impossible to predict all the ways in which the 

 various amino acid side chains would react. Indeed it would 

 seem unlikely that a mechanism of protein synthesis which 

 specifically joins together amino acids by the peptide bond 

 would produce other types of bonds in occasional positions in 

 protein chains. Thus, until more definite evidence to the con- 

 trary is produced, we are probably justified in concluding from 

 the results with insulin that proteins in general will be found to 

 obey the simple laws of protein chemistry and that they are 

 composed of polypeptide chains having a unique arrangement of 

 amino acids. This arrangement differs from protein to protein 

 but does not differ from one molecule to another molecule of the 

 same protein. Each position in the insulin chains is occupied 

 by one and by only one residue, at least to within the limits of the 

 methods used. This implies that all proteins are synthesized by 

 a process which joins the amino acids together by some common 



456 



