PROBLEMS OF MEASUREMENT OF MUTATION RATES 27 



not be used with red cells. The amount bound by red cells is less than 

 that of tissue cells, so you haven't as much fluorescence to see around 

 a red cell. But such techniques have been worked out. I can't remember 

 the reference, but now it should be possible to use fluorescent anti- 

 body, and that is a much better way than these indirect ways that I 

 have just given, provided you are looking for positive cells in a nega- 

 tive background. 



Neel: This is work by Wolf Zuelzer and Flossie Cohen in Detroit (8) . 



Cotterman: Were they using a fluorescein-labeled anti-A or anti-B 

 or fluorescein-labeled anti-antibody, or both? 



Neel: I can't recall. 



Cotterman: That last possibility I mentioned is one to keep in mind; 

 if your tagged fluorescent antibody does not give sufficient fluorescence, 

 you can, perhaps, double up on this by using an anti-antibody and 

 have both of them labeled with your fluorescent dye. 



Neel: Would an antibody work after an anti-antibody had doubled 

 up on it? 



Cotterman: Yes. This technique has been used in other fields. 



AtiDOod: One might suppose from the lack of independence of the 

 two alleles that mutation may be playing a minor role, and that many 

 exceptional cells are phenocopies that represent simply a failure to 

 form the antigen, even though the stem cell progenitor was of normal 

 genotype. 



Another type of evidence on the question of how the cells originate 

 is even more confusing than what I have just shown. It is the effect of 

 radiation on the process. We studied two polycythemics who were 

 going to be treated with P^-, with the idea of killing off part of their 

 bone marrow so they would no longer form too many red cells. We 

 got blood from them just before the P^- injection and followed their 

 exceptional cell levels at intervals afterward. Before going further, I 

 must refer to the A subgroups. 



The A phenotype has certain variants, Ai, Ao and weaker sub- 

 groups. But the principal distinction here is that if you take ordinary 

 human isoanti-A and absorb it with certain A's, you find that there 

 will be activity left against certain other A's. On the other hand, if 

 you take those other A's and absorb the serum no activity is left, so 

 we can subdivide A's according to whether they remove all activity 

 from isoantibody or only part of it. 



If they remove all the activity, they are called Ai; if they remove 

 part of it, they are called Ao, or some other subgroup of A that is 



