An optical method for finding the molecular orientation in different forms of crystalline haemoglobin; changes in dichroism accompanying oxygenation and reduction.

Abstract

When the amplitudes of the h0l reflexions of horse methaemoglobin at different lattice stages are plotted, an encouraging measure of agreement is found between amplitudes which come to lie in the same region in reciprocal space, confirming the impression that the haemoglobin molecule is a rigid entity whose structure is not affected by swelling and shrinkage of the crystal. The bewildering complexity of the diagram is lessened if the amplitudes are corrected for the effects of the different salt concentrations present at the different shrinkage stages, and if the layer lines are plotted singly. On applying the transform principle to each layer line, there are rarely more than one or two possible ways of drawing the waves, and the remaining points of uncertainty can probably be cleared up by improved accuracy of measurement. However, even if the combination of signs along each of the seven layer lines were known, this would still leave us with 27 alternative combinations of signs. It would obviously be helpful to have other forms of haemoglobin, with the molecules suitably oriented in the unit cell, which would enable us to bridge the gaps between the layer lines. This method of approach has become practicable since we discovered that the haemoglobins of three different animal species appear to be closely similar, at least in the gross features of their molecular structure. Patterson projections along the polypeptide chain direction of horse methaemoglobin, reduced human haemoglobin, and oxyhaemoglobin of ox show similar distributions of intramolecular vector peaks, and the distribution of peaks around the origin in tetragonal human oxyhaemoglobin can be reproduced by super­position of two identical Pattersons of horse methaemoglobin, one turned by 90° relative to the other.