STADLER 



DETECTION OF SPURIOUS GENE 

 MUTATIONS 



The development of criterions for 

 identifying gene mutations of evolu- 

 tionary significance is difficult even in 

 the study of selected genes of the most 

 favorable properties. In past studies, 

 the problem has been given a disarm- 

 ingly simple appearance by various as- 

 sumptions, some of which were un- 

 warranted, and some of which have 

 been invalidated by later discoveries. 



For example, we tend to feel that 

 some of the mutations detected in 

 our experiments must be qualitative 

 changes in the genes concerned, for 

 surely qualitatively altered genes have 

 arisen in the course of evolution. This 

 is mainly responsible for the wide- 

 spread belief that, even though some of 

 the apparent gene mutations identified 

 are demonstrably false, "true" gene 

 mutations must be included in the un- 

 classified residue. 



This belief is fallacious. Granting 

 that qualitatively changed genes must 

 have been evolved by mutation at rates 

 high enough to permit experimental 

 investigation, there is no assurance that 

 the steps in their evolution are repre- 

 sented in the mutants that are found in 

 our mutation experiments. When we 

 set out to identify mutants in a muta- 

 tion experiment, we must confine our- 

 selves to mutations of relatively large 

 effect, large enough to set the mutant 

 beyond the range of varying expres- 

 sion due to environmental and genetic 

 modifiers. If mutant changes occur 

 within the narrower range, we have no 

 way of identifying them. There is no 

 good evidence against the occurrence 

 of such subliminal mutations. The as- 

 sumption of the high constancy of the 

 gene is backed by evidence only con- 

 cerning the rarity of the distinct mu- 

 tations. If convincing evidence were 



253 



adduced tomorrow to show that geno- 

 types breed true only as a statistical re- 

 sult of sampling in each generation in 

 populations of genes genetically fluc- 

 tuating over an imperceptible range, 

 there is nothing in our present knowl- 

 edge that would contradict this con- 

 clusion. 



A study of R alleles of diverse origin 

 showed the common occurrence of 

 minute differences in the level of plant- 

 color expression (22). Such allelic dif- 

 ferences would not be expected if the 

 only source of variation in this gene 

 were mutation of the type that we 

 study in our experiments, but they 

 would be expected as a result of sub- 

 liminal mutation. 



If subliminal mutations occur, it is 

 possible that this type of mutation ac- 

 counts largely or wholly for the evo- 

 lution of new gene forms in nature. 

 Thus it is quite possible that the 

 sharply distinct mutations identified 

 in our experiments may be exclusively 

 the result of extragenic phenomena. 



A second assumption, or group of 

 assumptions, is concerned with the 

 possibility of distinguishing gene mu- 

 tation from gene loss. It was originally 

 supposed that induced recessive "visi- 

 bles" could safely be considered gene 

 mutations, on the assumption that all 

 genes were essential to survival. This 

 was contradicted by various instances 

 of cytologically demonstrable defi- 

 ciencies viable in haploid tissue or in 

 hemizygous individuals, or viable as 

 homozygotes in diploid individuals. 

 Such cases were relatively few, but 

 since both the cytological and the 

 genetic criterions of deficiency ap- 

 proach the limit of their range of ef- 

 fective application as the deficient 

 segment becomes smaller, there is 

 reason to suspect that physical loss may 

 be responsible for observed mutations 

 also in cases in which deficiency can- 



