ABSTRACT
A new method is described for measuring the thickness of thin spherical birefringent membranes. It consists of measuring a curve of retardation against radial distance at the edge of the membrane, and comparing this curve with other curves calculated from theory for membranes of known thickness. Diffraction is taken into account.
This method shows that the wet thickness of the human red cell ghost in glycerol is about A good agreement with the experimental results would be given by a model membrane consisting of a weakly birefringent layer 0·5μ thick surrounded by a strongly birefringent layer 40 A. thick. It is suggested that the thick layer is a 2 % protein gel, and that the thin layer is a bimolecular layer of lipids.
The birefringence indicates that there is radial molecular and tangential micellar orientation in the protein gel. This can be explained by an arrangement of the protein chains in looped bundles.
On the basis of these results a new model is put forward for the structure of the red cell membrane, and some of its implications are discussed.
In a membrane with a radial optical axis, like the ghosts, it is a matter of convention whether the sign of birefringence is referred to the tangent or the radius (positive with respect to the radius = negative with respect to the tangent, and vice versa). The sign is given with respect to the tangent throughout this paper since it stresses the similarity between membranes and fibres or long plates. Extended protein chains would be positive in all these cases.
In the calculation in this paper, line diffraction patterns were used rather than point patterns, since this simplifies the computation. This is discussed in Appendix 1.
Compensation increases contrast, and enables the sign of retardation in the image to be determined.
Radial orientation, producing negative birefringence, being assumed for both layers.
According to Ponder (private communication), these figures are probably underestimates rather than overestimates. Many of the measurements have been made with washed ghosts, and it seems likely that the more a ghost is washed the thinner it becomes. The true thicknesses, however, are unlikely to exceed twice the value above.
The positive form birefringence of ghosts is shown in a striking way in photographs of ghosts prepared for the electron microscope by Williams (1953). These are freeze-dried in a special way so that they maintain their spherical shape when dry in air. They show the strong positive form birefringence which would be expected if air replaced water as the intermicellar medium.
This is disputed by Bessis & Bricks (1950) who find that ghosts under the electron microscope are thicker at the edge.
Randall & Friedlander (1950) from EM photographs of ram sperm give diameters of 0·50−0·53 p for the main piece, and 0·15−0·20μ for the tail end or axial filament. If a bull sperm is similar to a ram sperm, the main piece would seem to be too broad for a line source. It does, however, appear to act as a line source, because it gives a linear diffraction pattern which is the same as that produced by the thinner tail end. We must therefore presume that only a part of the main piece is acting as the major source of birefringence. This may well be the axial filament, but, even if the whole tail is birefringent, most of the birefringence would appear to come from a strip narrower than its diameter, since the tail is circular in cross-section (possibly semicircular when dried on to a surface).
