STRUCTURE OF INSULIN 



Various methods were studied for fractionating the cystine 

 peptides produced. Conditions were required that would not 

 lead to redistribution of the disulfide bonds and that would 

 provide sufficient material for identification of the oxidation 

 products. Paper chromatography seemed unsatisfactory from 

 both these points of view, and the best results were obtained 

 using high-voltage paper ionophoresis (13) in pyridine-acetate 

 buffers atpH 3.3 and 6.4. At the latter /?H a good separation of 

 the neutral peptides may be obtained owing to the unusually 

 low pK. of the amino groups of cystine in peptides (6). Thus 

 whereas most peptides of mono-amino acids are neutral at pH 

 6.4, those which have a free amino group on a cystine residue are 

 slightly acidic and those (such as cystinyl diglycine) with two 

 such groups are even more acidic. In general ionophoresis at 

 two difTerent pH. values yielded peptides of reasonable purity, 

 and their structure was determined by oxidation. 



In Table VI are listed the significant peptides detected in 

 this work. Although the distribution of the disulfide bridges in 

 insulin could be deduced from any one of these peptides, it was 



TABLE VI 



Cystine Peptides Identified in Partial Acid Hydrolvzate of Insulin 



Leu -CyGly 

 S 

 S 

 CyAla 



Ser-Val-Cy 

 S 

 S 

 Glu-Cy 



necessary to investigate all the cystine peptides present in 

 significant amount to be sure that no rearrangement of bonds 

 had occurred. The peptides all agree and establish the struc- 

 ture of insulin given in Figure 4. 



This structure is based on a value of 6000 for the molecular 

 weight but does not prove this value, since the structure shown 



453 



