110 MUTATIONS 



slightest doubt of this. Some one-armed chemicals are better mutagens 

 than the corresponding two-armed ones (67). 



For chromosome breakage this has hardly been tested until recently. 

 It seemed possible that breakage of two strands might require cross- 

 linkage, but there is no evidence for it. For instance, beta propio- 

 lactone and ethylene imine, both one-armed, are good chromosome 

 breakers. However, in view of the findings of the cancer workers and 

 of the connection between carcinostatic action and chromosome break- 

 age, it seemed worth while to test this question properly. 



Last year I had a co-worker, Professor Nakao, who did experiments 

 to test whether bifunctional alkylating chemicals are relatively more 

 effective than monofunctional ones in causing chromosome breakage. 

 By relatively I mean relative to their ability to produce what we 

 think are mainly gene mutations. Nakao used the two related com- 

 pounds ethylene oxide and diepoxybutane for producing lethals and 

 translocations in Drosophila. We were not interested in the absolute 

 frequencies, which were definitely lower for ethylene oxide, probably 

 because it is a very unstable compound. We were interested in the 

 relative frequencies of translocations to lethals, for the former require 

 chromosome breakage while most of the latter do not. The ratio 

 of the two effects turned out to be exactly the same for the two 

 compounds, so there was no evidence for an effect of cross-linkage 

 on chromosome breakage. Nakao intends to test another pair of com- 

 pounds in the same way. 



Lederberg: Is that correlation really very good for carcinostasis or — 



Auerbach: Professor Haddow of the Chester-Beatty Institute 

 assured me of this. 



Goldstein: I want to know whether you can use nitrous acid to solve 

 this question of what exactly is going on, in the following way: If you 

 start out by assuming your proposal begins with a GC, which, in the 

 presence of this alkylating agent winds up as gap-C, you don't know 

 what is pulled into the gap, so we can draw four possibilities. 



One is that G is drawn into the gap, in which case we have no 

 permanent mutation. Another is that C is drawn into the gap and we 

 get CG. Or T is drawn into the gap and we then get TA. Or A is 

 drawn into the gap and we then get AT. 



You now treat the mutant with nitrous acid. Then CG should go to 

 TA, which is not a reversion. TA should go to CG, which is not a 

 reversion. Only AT should go to GC, which is a reversion. So one 

 would conclude that if you can get reversions with nitrous acid, this 

 must be what has happened. Is that reasonable or is it not? 



