GENETIC IMPLICATIONS OF MUTATIONS IN S. TYPHIMURIUM 279 



sharp in the antigenic analyses of Salmonella. The tree-like relationship 

 which they suggest was our chief stimulus to a study of bacterial genetics in 

 this organism. 



Preliminary tests of all of the auxotrophic mutants made by Miss Dorothy 

 Farley show that they are unchanged antigenically. Not only the specific 

 antigens, but the agglutination titers are the same as the original strains. 

 This has been confirmed by reciprocal absorption tests, as well as by precipi- 

 tation, and inhibition of agglutination using supernatants from boiled cul- 

 tures. Thus it appears that the loss of ability to synthesize a particular amino 

 acid in no way alters the antigenic configuration. Apparently if proteins are 

 formed at all they take on the antigenic configuration of the cytoplasm al- 

 ready there. The auxotrophic mutants and the antigenic patterns fall into 

 two quite independent systems so far as present evidence goes. This seems 

 to be true also for variations in or loss of virulence. The relation of the auxo- 

 trophic mutants to virulence for mice is being studied in detail by Gowen 

 and his associates and will be reported separately, but so far at least it ap- 

 pears that there is no relation between virulence and the biochemical re- 

 quirements of the strain. 



It was originally and is still our hope to be able to induce antigenic 

 variants by radiation, but so far such attempts have given negative results. 

 We have inoculated radiated suspensions into one end of U tubes of semi- 

 solid agar containing low concentrations of O serum from a rabbit im- 

 munized against the specific strain, and the organisms grow through the 

 medium. When agar containing specific H serum is used, however, the 

 organisms grow only at the site of inoculation. If antigenic mutants had 

 occurred we would expect that the homologous serum would act as a screen 

 to block off the original and let the mutants through, just as the penicillin 

 does for the auxotrophs. The result simply means that we have not found any 

 antigenic mutants following radiation. Perhaps we should not expect any. 

 Antigenic mutants have been induced in several bacteria by other meth- 

 ods, especially by McCarty (1946) in the pneumococcus, by Bruner and 

 Edwards (1947) in Salmonella, and by Boivin (1947) in E. coU. The pneu- 

 mococcus method is not applicable to Salmonella, and the Boivin method in- 

 volving exposure of the organism to autolysates of rough variants of other 

 strains gives negative results. Tests using similar culture filtrates have 

 been unsuccessful in altering the antigenic constitution of our organism. On 

 the other hand, Miss Farley has made use of the Edwards technique of grow- 

 ing an auxotrophic mutant in a semi-solid medium containing homologous O 

 serum previously absorbed with a related organism which lacked one of the 

 major antigens, XII (and in another case lacked V but carried an additional 

 antigen XXVII). By this method two successful transformations of type 

 have been secured out of several tried. Both of these transformations were 

 performed on an auxotrophic mutant (519-PlO) requiring histidine. 



