PROBLEMS OF MEASUREMENT OF MUTATION RATES 67 



Benzer: On what level do you want the specificity? Dr. Demerec 

 asked for specificity at the level of the locus. 



Auerbach: I was looking at this question from the point of view 

 of my discussion tomorrow, and I found only one piece of evidence 

 of mutagen specificity in the sense which Dr. Demerec, I am sure, 

 means: that when you compare several mutagens, you get a pattern 

 which differs between loci in forward mutations. This has been found 

 by Heslot (18). He works with fission yeast, Saccharomyces pombe, 

 and he scored for forward mutations at four different adenine loci. The 

 method of detection is the same that Hershel Roman first used in 

 yeast. He starts with a strain that carries a purple adenine mutation 

 so that the colonies look purple. If the chain of reactions leading to 

 adenine is interrupted at some earlier stage by mutation at one of the 

 other loci, then the purple pigment is not formed, and the colonies 

 are white. It is a very neat method. He used four mutagens; UV, 

 etliyl sulfate, isopropyl methane sulfonate, and ethylene imine. I shall 

 give here his main results. I shall leave the ad-5 locus away because 

 it mutated rarely with all four mutagens. In parentheses after the 

 actual numbers of mutants obtained I shall insert the relative fre- 

 quencies of mutations at the ad-1 and ad-3 loci compared with those 

 at the ad-4 locus; this makes a comparison easier. 



ad-l ad-3 ad-4 



When statistical tests were made they showed that the distribution 

 of mutations after treatment with ethylene imine was significantly 

 different from the distribution after treatment with UV or ethyl 

 sulfate. 



Cotterman: These are frequencies of mutant cells per how many? 

 What is the total cell count? 



Auerbach: I don't remember. I don't think he even said, but he 

 must have taken that into account when calculating the statistical 

 significance. 



Freese: Actually this is not so surprising, because we have the 

 tremendous differences in mutation frequencies if we look at one 

 particular site. If you assume, let's say, 500 or 1000 sites in a gene, or 



