MUTAGENESIS 121 



Demerec: Could you lose several nucleotides? 



Zarnenhof: You probably could. 



Demerec: Because that is where the short deletions come in. 



Auerbach: But the long deletions come in, too. 



Demerec: No, the long ones take about 75 of the identified sites. 



Auerbach: There was a somewhat similar story by Yanofsky (82) 

 in a strain of E. coli where he almost alw^ays got deletions embracing 

 two loci in a particular chromosome region, and this property was 

 transmitted in crosses. Yanofsky does not, I believe, mention this pos- 

 sibility, but could this not also be a region with a duplication and the 

 tendency to form a loop? 



Atwood: What were the criteria for deciding that part of the in- 

 cluded region in the deletions is inert? 



Demerec: The long deletions cover all the sites that have been identi- 

 fied in the right end of the cysC locus, and it seems probable that the 

 deletions terminate in the region beyond the locus. I have a feeling, 

 but unfortunately no evidence, that nonsense (heterochromatic) regions 

 frequently intervene between loci that do not belong to a sequence. 

 Meager support for this belief has been obtained from our studies of 

 proline mutants (51). If the region to the right of cysC is heterochro- 

 matic, and if a portion of it or of any other heterochromatic region is 

 duplicated in the left part of cysC, one could expect synapsis between 

 the duplicate sections to be just as frequent as if the duplicated seg- 

 ment were functional. 



Atwood: If you went over your shorter deletions, you might find 

 one or more of them that covered only one of the presumed duplicated 

 regions, and if that is the reason for its apparent inertness, then, 

 using that stock with one part deleted, you would find point muta- 

 tions in the remaining part. That is one way of getting it. Those could 

 be surveyed to see if they get new point mutations. What I mean is 

 you could survey them to see whether there are any of them which, 

 when treated again, now yield a new class of point mutations not 

 yielded by the original stock, because the redundancy has been taken 

 out by the deletion. 



Benzer: But these are mutants, and you can't see any new point 

 mutations. 



Demerec: There is no way of selecting them. 



Atwood: Oh, you have already selected them? You can't do it 

 by- 



Benzer: The double mutants will look just the same as the single. 



Atwood: You don't delete the same way? 



