BARBARA W. LOW AND JOHN T. EDSALL 



hydrogen bonds between peptide CONH residues cannot 

 be formed. If we ask whether there is any configuration 

 for this region of the chain which would permit some 

 intrachain hydrogen bonding, or whether it is possible 

 to maintain the a-helix configuration in at least some part of the 

 Ag^] loop, then these problems can be studied using scale atomic 

 models. 



Lindley and Rollett (62) have shown that this region of the 

 chain may be twisted into a more regular configuration if it 

 joins two segments of a chain which have different screw direc- 

 tions. If the Ai-9 part of the chain is a left-handed coil and the 

 A9-21 part is a right-handed coil, then the A"] region forms 

 the figure of eight required by this change of hand in the coil. 

 In the detailed model, which they have developed, only two of 

 the main chain NH. . . .OC hydrogen bonds are broken. The 

 CO group of residue 8 could, in beef insulin, form an OH . . . . O 

 bond with the hydroxyl group of the seryl residue 9 (see Figures 

 1 and 2). In this model the helical axis of the right-handed 

 A9-21 sequence is parallel to a right-handed B chain a-helix. 

 The helical axis of the A1-7 segment is in the same plane as the 

 B and A9-21 axes; it is tilted away from B at an angle of 30 °. 



Low (68) has constructed a model using a similar change 

 of hand from a left-handed Ai-g segment to a right-handed 

 All- 21 segment. The B chain in her model is a left-handed 

 a-helix. In this structure the two parts of the A chain do not lie 

 in the same plane. There is a lateral shift between them so 

 that the model appears in "end-on" projection as an area of 

 three close-packed helices. Three intrachain hydrogen bonds 

 are broken in this model. Both structures are plausible. The 

 Lindley and Rollett model is the more satisfying solution of the 

 simple stereochemical problem: to find a model structure for 

 the insulin molecule in which all of the B chain and as much as 

 possible of the A chain are in the a-helix configuration. The 

 helix structure in this model is predominantly right-handed, 

 whereas the Low model is predominantly left-handed. 



Insulin was included in Arndt and Riley's (5) detailed 



422 



