42 



IS. D. Davis: Studios on Nutritionally Deficient Bacterial Muta 



this laboratory 1 and by Lederbekg and Zinder 2 . 

 After isolation, the growth requirement of a mutant 

 can be simply determined by distributing tiny drops 

 of solutions of nutrilites on the surface of a heavily 

 seeded pour plate 3 (Beijerinck's "auxanography"). 



Since the publication of this method, a modification 

 has been introduced which substantially improves its 

 efficiency. Instead of inoculating the washed organisms 

 directly into minimal medium containing penicillin, 

 the inocula are placed in a medium lacking nitrogen. 

 After 4 to b hours of incubation, during which dissimil- 

 ation of glucose promotes the exhaustion of stored 

 metabolites, the minimal medium is completed by 

 addition of ammonium sulfate. Penicillin is added, and 

 the procedure is continued in the usual way. The period 

 of dissimilation particularly improves the recovery of 

 certain vitamin-requiring mutants. 



Although the penicillin method provides a marked 

 improvement in the efficiency of isolating mutants, it 

 nevertheless has serious limitations. The population 

 density cannot be indefinitely large, as is possible with 

 the drug-resistant mutants, for at population densities 

 above 10 5 to 10 7 cells per ml, the non-mutants release 

 enough of various metabolites ("syntrophism" ; cf. 4 ) to 

 permit certain mutants to grow slightly and hence be 

 sterilized by penicillin. In addition, even at low popula- 

 tion densities the recovery of mutants is not quantit- 

 ative. Finally, the use of this method for estimating 

 mutation rates is limited by the requirement of a stage 

 of intermediate cultivation between the irradiation 

 and the exposure to penicillin, in order to permit pheno- 

 typic expression of the induced mutation. During this 

 cultivation, the distribution of the population will 

 undoubtedly be distorted. 



Delayed phenotypic expression 



The requirement of a certain amount of growth for 

 the phenotypic expression of an induced mutation is 

 itself an interesting point, and has also been reported 

 for phage-resistance 5 . The explanation, at least with 

 the nutritionally deficient mutants, appears to be that 

 the cell must undergo some growth, possibly several 

 generations, before the premutational products of the 

 mutated gene (enzymes; possible intermediates be- 

 tween genes and enzymes) are exhausted. Only then 

 will the pattern of enzymes in the cell correspond to 

 the new pattern of genes. For this particular mechanism 

 of delayed phenotypic expression we have proposed 

 the term "phenomic delay" 6 , the "phenome" being 



1 i; I). Davis, J. Amer. Chein. Soc. 70, 1267 (1948); Proc. Nat. 

 Acid. Sri. 35, 1 (1949). 



2 J. I.i in Kiuui., and N. J. Zinder, Amer. Cheni. Soc. 70, 1267 



3 J. Lederberg, J. Bact. 52, 503 (1946). G. Pontecorvo, J. 

 Gen. Microbiol. -3, 122 (1949). 



4 J. Lederberg, J. Bart. 52, S03 (1946). 



5 M. Dkmerec, and R. Latarjet, Cold Spring Harbor Symp. 

 Quant, Biol. 11, 38 (1946). 



6 B. I). Davis, Proc. Nat. Acad. Sci. 35, 1 (1949). 



defined as the total non-self-reproducing part of the 

 cell, under the control of the self-reproducing genes. 

 The occurrence of the same phenomenon in mutations 

 in the reverse direction will be discussed later. 



Other possible explanations of the requirement of 

 intermediate cultivation in the penicillin method in- 

 clude segregation of mutant and non-mutant nuclei 

 from a multinucleate cell, and a syntrophic effect of 

 the non-viable irradiated bacteria, which would pro- 

 mote sterilization of mutants by penicillin. Evidence 

 will be published elsewhere that neither of these 

 mechanisms furnishes an adequate explanation, while 

 the phenomic delay accounts for all the available facts. 

 A similar conclusion was reached by Newcombe in a 

 thorough analysis of delayed phenotypic expression of 

 phage resistance 1 . 



Biochemical advantages of bacteria 



Bacteria are less easily studied genetically than 

 molds such as Neurospora, which can be made to 

 multiply sexually or asexually at will. While genetic 

 recombination in bacteria has recently been demon- 

 strated by Tatum and Lederberg 2 , it apparently 

 occurs in only a few strains ami a tiny proportion of the 

 population. For biochemical investigation, however, 

 bacteria appear to have several advantages. Not only 

 can a variety of mutants be isolated relatively quickly, 

 but it is possible to demonstrate very simply, by the 

 syntrophic interaction of adjacent streaks on solid 

 media, the instances in which one mutant accumulates 

 a metabolic intermediate which is utilized as a nutrilite 

 bv another mutant. Metabolite accumulations, when 

 present, are extremely useful in analysing biosynthetic 

 pathways 3 . 



A further advantage of bacteria is the uniformly 

 dispersed growth of certain species in liquid media. 

 which permits simple and precise quantitative experi- 

 ments, using colony counts for low population densities 

 and turbidimetry for high densities. On solid media, 

 the production of uniform colonies has made possible a 

 variety of experiments involving prolonged cultivation. 

 without risk of confusion from back-mutants, which 

 are readily distinguished from the rest of the population. 

 In addition, slight variations in colony size and syntro- 

 phism have made it possible to recognize unexpected 

 phenomena which might easily have gone unnoticed 

 in a mycelial mat. Finally, in relation to chemotherapy, 

 the metabolism of bacteria is of particular interest. 

 Although some of the problems to be described here 

 are still under investigation, it seems desirable to 

 illustrate at this time the types of phenomena that can 

 be revealed by these primitive techniques, especially 

 bv the test for syntrophism. This effect has been long 



1 H. B, Newcombe, Genetics 33, 4 17 (1948). 

 - E. I.. Tatum, and J. Lederberg, J. Bact. 53, 673 (1947). 

 J. Lederberg, Genetics 32, 505 (1947). 



3 N. H. Horowitz, J. Biol. Chem. '«-', 413 11946). 



95 



