62 D. A. L, DAVIES 



specificity of its own quite distinct from H-2 specificity. The 

 activity of this protein component, as measured by dilution in 

 agar diffusion, is virtually as strong as that of the "antigen 3" 

 isolated from non-specific tumour fluids, or cells, where it is 

 believed to be associated with the membraneous structures and 

 composes a part of the cell which persists in insoluble form after 

 exhaustive extraction with salt solutions (from i to 15 per cent) 

 and with water (Haughton and Davies, 1962). Attempts to find 

 H-2 specificity in this protein after removal of lipid from the 

 H-2 antigen have not been successful, though the methods used 

 for lipid removal so far are very unsatisfactory. 



The lipid component has not been properly examined but there 

 is every indication that it is a complex of different lipids. The 

 proportion of lipid in successive extractions has varied consider- 

 ably and it seems possible that dissociation may occur in the course 

 of preparation. It is not yet clear how closely bound the protein 

 and lipid moieties really are. 



Information about the immunological determinant parts of the 

 complex has not yet been obtained, but many facts point to the 

 lipid component being involved. The proportion of hpid in- 

 creases with increasing activity on purification, and the protein 

 component has non-H-2 species specificity of its own. H-2 

 specificity is present on the surfaces of cells whereas the specificity 

 of the protein ("antigen 3") is not exposed on the cell surfaces. 

 Loss of activity on freezing and thawing, freeze drying, ultra- 

 sonics, and treatment with sodium dodecyl sulphate, Tween-20 

 (a non-ionic detergent), or deoxycholate, is consistent with lipid 

 specificity and more difficult to associate with protein or carbo- 

 hydrate structures, since the component parts of lipids are not 

 firmly linked as are amino acids or sugars in polypeptides or 

 polysaccharides. Sugars in the lipid might well be susceptible to 

 periodate destruction, indeed lipids are sometimes readily 

 attacked by periodate in the absence of sugar residues. It is felt, 

 therefore, that either the lipid itself carries the specificity, or that 



