X-ray analysis of haemoglobin 



M. F. PERUTZ 



Medical Research Council Unit for Molecular Biology, Cavendish Laboratory, 

 University of Cambridge 



When Pauling and Corey first noted the correspondence between the three- 

 dimensional vector structure of haemoglobin and the vector distribution to 

 be expected from the a-helix/ and I subsequently found a weak, but detect- 

 able, 1-5 Â reflexion in haemoglobin, ^ I beheved that a solution of the 

 haemoglobin structure in terms of parallel polypeptide chains might be found 

 by trial. This led to an attempt by Bragg, Howells and myself^ to calculate 

 the electron density distribution viewed along the suspected chain direc- 

 tion of the molecule, which coincided with the a axis of the haemoglobin 

 crystal. A Fourier projection on the a plane, corresponding to an end-on 

 view of a-helices, was indeed obtained, but its validity rested on the assump- 

 tion that the interpretation of the structure in terms of parallel a-helices 

 was correct. When the absolute heights of the electron density peaks were 

 measured, they were found to amount to only one-third to a half of the 

 heights to be expected from the initial assumptions, which showed that the 

 model was oversimplified. Crick arrived at the same conclusion by a dif- 

 ferent argument,* and also pointed out that a-helices are likely to pack 

 together at an angle of about 20°, which speaks against their being parallel 

 in haemoglobin.^ 



The failure of this attempt at analysis by trial pointed to the need for 

 a different approach, and one that was free from any preconceived notion 

 about the structure of the molecule. In practice this meant that the phases 

 of the diff"racted rays would have to be determined by direct methods. In 

 a series of papers Bragg, myself and our collaborators have now described 

 four different methods of phase determination. The most useful of these 

 was based on the method of isomorphous substitution described by Dr 

 Kendrew (p. 125). Green, Ingram and Perutz^ crystallized compounds of 

 horse haemoglobin in which the sulphydryl groups were blocked by /j-chloro- 

 mercuribenzoate or by silver ions. These compounds proved to be iso- 

 morphous with normal horse methaemoglobin. From the changes in the 

 intensities of the X-ray diff'raction pattern produced by the presence of the 

 heavy metals. Green, Ingram and Perutz' were able to determine the signs 

 of 87 out of the 94 hOl reflexions covered by their X-ray photograph. From 

 this and other, supplementary, information Bragg and Perutz then calculated 



