MUTATIONS IN ESCHERICHIA COLI INDUCED 



BY CHEMICAL AGENTS' 



EVELYN M. WITKIN 



For over two decades geneticists have been in- 

 terested in the possibility of inducing mutations 

 with chemicals. Particularly, it has been hoped that 

 mutagenic compounds might be discovered which, 

 through their specificity of action, would lead to 

 some understanding of the chemical basis of muta- 

 tion, and ultimately of the structure and organiza- 

 tion of the gene. 



The first clearly successful attempt to induce 

 genetic changes chemically was described by Auer- 

 bach and Robson (1944), who produced mutations 

 and chromosomal abberrations in Drosophila by 

 exposing the flies to mustard gas and related com- 

 pounds. More recently, Demerec (1947 and un- 

 pub.) has shown that certain carcinogenic hydro- 

 carbons (1, 2, 5, 6-dibenzanthracene, methylcho- 

 lanthrene, beta naphthylamine, benzpyrene) are also 

 effective in inducing mutations in Drosophila. These 

 agents, like radiations, appear to be entirely non- 

 specific in the sense that the affected loci are dis- 

 tributed at random along the treated chromo- 

 somes. Nitrogen mustard has been shown to induce 

 genetic alterations in Neurospora (Tatum, unpub.; 

 Horowitz, Houlahan, Hungate and Wright, 1946), 

 and in bacteria (Tatum, 1946; Bryson, unpub.), 

 and at least one of the carcinogens, methylcholan- 

 threne, is effective in Neurospora (Tatum, unpub.). 

 These successful results seem to be the opening guns 

 in what Muller called, a few years ago, "the com- 

 ing chemical attack on the nature of the gene" 

 (Muller, 1947). They suggest the need for a system- 

 atic survey to determine the distribution of muta- 

 genic compounds among various chemical groups, 

 and to lay the groundwork for subsequent analysis 

 of their mode of action. This paper will deal with 

 preliminary results obtained in tests of 4 substances, 

 the first of a series to be investigated in an exten- 

 sive survey. 



Two methodological factors are of critical im- 

 portance in an attempt to examine large numbers 

 of compounds for mutagenic activity: the basis upon 

 which the chemicals are selected for test, and the 

 choice of biological material. Concerning the method 

 of selecting chemicals, one sober if somewhat un- 

 imaginative approach is an indiscriminate raid on 

 the nearest chemical shelf, which has the advantage 

 of objectivity and avoidance of the hazards of pre- 

 mature preconceptions. On the other hand, it is 



'This work was done under an American Cancer Society 

 fellowship recommended by the Committee on Growth of the 

 National Research Council. 



far more tempting to extend oneself on the basis 

 of present ideas concerning the possible organiza- 

 tion of genie material, and to select chemicals which 

 might reasonably be expected to affect, or fail to 

 affect, the projected hereditary units. The ap- 

 proach used in these experiments has been to 

 assume that nucleoproteins are somehow centrally 

 involved in the genetic system, and, as a starting 

 point, to investigate chemicals known to have some 

 more or less well-defined chemical or physical effect 

 on nucleoproteins or nucleic acids. It must be em- 

 phasized that this approach has no greater justifica- 

 tion on a priori grounds than many others, and that 

 the basis of selection may prove to be entirely 

 spurious, since none of the chemicals tested thus 

 far is specific in its action on nucleoproteins or 

 nucleic acids. 



The choice of biological material is obviously 

 very important in this type of investigation. The 

 primary requirements are 1) the availability of 

 techniques for treating the organism with chemicals 

 so as to be reasonably certain that they will reach 

 and penetrate the critical sites, and 2) the avail- 

 ability of clear-cut genetic methods for detecting 

 induced mutations. The penetration problem has 

 been the most serious difficulty in the use of Dro- 

 sophila for chemical induction, and although im- 

 proved methods of treatment have been developed, 

 the possibility remains that negative results may 

 be due to the failure of some chemicals to penetrate 

 the germ cells in sufficient concentration. The 

 genetic techniques for detecting induced mutations 

 in Drosophila are unparalleled in many respects, 

 but for purposes of an extensive survey of the mu- 

 tagenic action of chemicals, they are extremely 

 laborious and slow. The problem of penetration is 

 much less serious in microorganisms. Until recently, 

 however, genetic methods analogous to the C1B 

 and similar techniques in Drosophila have not been 

 available for bacteria. At the present time, Escher- 

 ichia coli provides promising material for a survey 

 of the mutagenic activity of chemicals, and for de- 

 tailed analysis of certain aspects of their mode of 

 action. 



Luria and Delbriick (1943) described mutants 

 of strain B of E. coli which are resistant to one or 

 more bacteriophages to which the parent strain is 

 sensitive. These mutants arise spontaneously in cul- 

 tures of the B strain at a rate of about 10 -8 muta- 

 tions per bacterium per generation, and can be de- 

 tected easily by plating out samples of the culture 

 in the presence of an excess of bacteriophage. The 



[256] 



[Reprinted from Cold Spring Harbor Symposia on Quantitative Biology 12 : 256-269, 

 published by the Long Island Biological Assn.] 



IIQ 



