BARBARA W. LOW AND JOHN T. EDSALL 



together by interactions between side-chain groups, which may 

 be polar or nonpolar. The /? structure is essentially made up of 

 extended peptide chains in parallel array. The two pleated 

 sheet structures provide nicely for the stereochemical require- 

 ments of this mode of packing, and one or other fits the observed 

 x-ray diffraction patterns of most of the j3 proteins. 



In the a-helix structure the side chains bristle out from the 

 chain core. The structure may be thought of as a cylinder of 

 circular cross section. When a group of these structures are 

 arranged in parallel close-packed array an interhelical distance 

 of X angstroms would give rise to an interplanar spacing of 

 X V 3/2 A. The pitch of the a-helix is 5.4 A, and the struc- 

 ture should give rise to a nonmeridional reflection of this spac- 

 ing. Further such parallel arrays of a-helices should give a 

 strong meridional x-ray reflection corresponding to diffraction 

 from planes normal to the helical axis and spaced a unit residue 

 increment apart : 5.4/3.6 = 1.5 A. 



A reflection of 1.5 A was noted in an a-keratin pattern from 

 porcupine quill by MacArthur (71) in 1943, but no especial 

 significance was then attached to it. When the a-helix structure 

 was first proposed, Perutz (86) recognized that the reflection 

 should show up prominently in a-helix structures if the fiber were 

 tilted at the appropriate Bragg angle to the incident x-ray beam. 

 He observed it in synthetic polypeptides, in a-keratin, and in 

 wet crystals of the globular protein hemoglobin. 



The synthetic polypeptides do show the expected non- 

 meridional 5.4 A reflection. In fact the synthetic polypeptide 

 poly-7-methyl-L-glutamate gives an x-ray diffraction pattern 

 which fits excellently the calculated pattern for an a-helix 

 (9,24,110). 



In the fibrous protein a -pattern there are meridional 

 reflections at 5.1 A and 1.5 A. The discrepancy between a 

 predicted 5.4 A nonmeridional spacing and the 5.1 A meridional 

 spacing observed cannot be resolved simply by supposing that 

 the a-helices are inclined to the fiber axis, for this would put the 

 1.5 A spacing off the meridian. Crick (29) has shown that if 



404 



