88 DISCUSSION 



protein; it will still stain with lipid stains, for example. So, whatever 

 the snake venom removes, it does not remove all the hpid. 



Dmnes: I am puzzled by one thing. You gave the nitrogen content 

 of your Triton-soluble lipoprotein as 14-9 per cent, but you have 26 

 per cent extractable lipid, and if that is nitrogen-free it puts the nitrogen 

 content of the protein up to 20 per cent, which is impossible. Is there 

 some nitrogen in the lipid ? 



Kaiidntsch: There is some nitrogen in the lipid, but I have not been 

 able to get a satisfactory analytical value for it. 



Davies: It is just possible that the second component you get on 

 electrophoresis after treatment with snake venom might be lysolecithin. 

 We have had trouble with ultracentrifuging solutions of our material 

 with lysolecithin, because lysolecithin forms micelles in water, and the 

 micelles sediment at about the rate of serum globulin in the ultracentri- 

 fuge, giving an apparent molecular weight of about 150,000. 



Kandutsch: Wouldn't lysolecithin dialyse out? 



Davies: I don't know. I think it might not, because it does form these 

 large micelles and is therefore not distributed in solution in the ordinary 

 sense. 



Mcdawar: Dr. Kandutsch, what is your present feeling about the 

 role of carbohydrate in these antigens, and the inactivation produced 

 by sodium periodate ? 



Katidutsch : The amounts of carbohydrate that we fmd in these pre- 

 parations are very small, and I can't really say that the effect of periodate 

 is on the carbohydrate. Periodate is known to split amino acids such as 

 threonine or serine if they are on the N-terminal end of a protein. We 

 have done some qualitative studies of the amino acid composition of 

 the Triton-soluble lipoprotein, and we have a quantitative study under 

 way and there does seem to be quite a lot of threonine and serine in 

 these molecules. Many of the lipoproteins in serum have serine and 

 threonine end groups, so this is a possibility. It would be very interest- 

 ing if one could simply knock off the N-terminal amino acid and there- 

 by destroy the antigenic activity, thus localizing an antigenic site. 



Davies: Periodate, of course, can attack lipids rather readily; cer- 

 tainly some polyhydroxy lipids are just as susceptible to periodate 

 destruction as sugars. Dr. Marjory Macfarlane described a serologically 

 active lipid wliich contained no sugar, but was destroyed by periodate. 



