130 MUTATIONS 



the mutation itself, whose response one is testing, may alter the bio- 

 chemical pattern of the cell. 



I might mention the heminless bacteria which, when they are grown 

 without hemin, have no catalase. Naturally, they will react in a dif- 

 ferent way to radiation. One way to overcome this, and the way which 

 is usually used, is to have doubly marked strains and compare the 

 frequency of mutations of two loci on the same background. Then one 

 gets a typical pattern, which would look like this: 



Mutation frequency 

 Mutagen A B 



When such a pattern has been obtained work on it usually stops, 

 and only two interpretations are considered. The one is that this is spec- 

 cificity of the reaction between gene and mutagen; the other, that the 

 mutagen selects spontaneously arisen mutants. The latter can be ruled 

 out by reconstruction experiments, and this is often done. But all other 

 possibilities are usually neglected. 



There is, for instance, a very obvious one. It is the fact that when 

 you use a double strain, the medium for testing mutations is not the 

 same for the two mutants. Let us say we work with a tryptophanless 

 leucineless strain. If one tests for tryptophan reversions one plates on 

 leucine; if one tests for leucine reversions one plates on tryptophan. 

 The influence of this difference in plating medium has, to my knowl- 

 edge, never been tested. A collaborator of mine. Dr. Clarke, is just 

 testing it in fission yeast, Saccharomyces pomhe. He has already ob- 

 served one striking effect of the plating medium. He scores for reverse 

 mutations at two loci, the adenine-1 locus, and a methionine locus. He 

 found that methionine in the medium suppresses mutation at the 

 adenine locus, so that in the double strain, when he plates on methio- 

 nine for adenine reversions, these are very much reduced in number. 

 If this kind of response to the plating medium should be influenced by 

 the previous mutagenic treatment — and I see no reason why this should 

 not be so, especially after chemical treatment — then one would obtain 

 a mutagen specificity which occurs at a very late stage in mutagenesis, 

 when the premutation has to become stabilized or even later when the 

 mutated cell has to make a colony. 



