M. F. PERUTZ 



haemoglobin are transformed into bundles of birefringent needles in 

 the presence of oxygen. X-ray analysis shows the crystals to be 

 rhombohedral with six haemoglobin molecules arranged in a row 

 along an axis of triad symmetry. The dichroic absorption of the 

 crystals suggests that the flat sides of the haem groups are normal to 

 the triad axis, while a vector projection on a plane normal to that 

 axis gives an indication, but no more than an indication, that the 

 polypeptide chains are parallel to that axis. On this basis the triad 

 in reduced haemoglobin would correspond to the X axis in methaemo- 

 globin, as far as the orientation of the haemoglobin molecules is 

 concerned. It is very difficult, however, to fit a molecule into the 

 unit cell of dried reduced haemoglobin which also fits into the unit 

 cell of dried methaemoglobin. This, and the totally different crystal 

 form of reduced haemoglobin indicates that a molecular change more 

 profound than the mere oxygenation and de-oxygenation of the haem 

 groups accompanies the transformation between the two compounds. 

 This had already been suggested by Haurowitz 6 , but so far no other 

 evidence has appeared which would provide any clue as to the nature 

 of the change. Neither Haurowitz nor Gutfreund (private com- 

 munication) could detect any difference in viscosity, nor could the 

 latter find any difference between the sedimentation constants of 

 carboxyhaemoglobin and reduced haemoglobin. There are, of course, 

 great differences between (a) the solubilities of these two derivatives, 

 and (b) the stabilities of the two types of derivatives on storage 

 (Roughton, private communication). It remains to be seen whether 

 these are due to a change in structure on oxygenation. 



Haemoglobins of Other Species — A comparative study of foetal and 

 and adult sheep haemoglobin, undertaken at Sir Joseph Barcroft's 

 suggestion, has recently been published 7 and has provided strong 

 additional evidence for the non-identity of the two proteins. Our 

 study was more concerned with a general survey of the types of crystal 

 structure which occur in the foetal-adult haemoglobin system of sheep 

 than with a detailed analysis of any one of them. We did, however, 

 find indications of a difference in molecular symmetry between the 

 two proteins : in the adult haemoglobin the molecular weight of the 

 asymmetric unit* was 68,000, while in the foetal one it was 34,000. 

 This unit of 34,000 in turn seemed to consist of two at least closely 

 similar sub-units of molecular weight 17,000. These differences in 

 molecular symmetry can be correlated with different splitting properties 

 of the two proteins in solution. Gutfreund (p. 195) who measured 

 the sedimentation constant of the two proteins found that adult sheep 

 haemoglobin does not dissociate, whereas the foetal haemoglobin — 

 * For definition of this term see page 162. 



146 



