GENETICS OF SOMATIC CELLS 439 



line. Different suspensions of cells may differ greatly from each other with regard to 

 their physiologic quality and with regard to the fraction of the population capable of 

 continued multiplication. This may in turn result in variations in the amount of 

 lymphomatous tissue found in different groups 9 days after inoculation. In general, 

 the fluctuation test always presupposes that all environmental and preparative condi- 

 tions are equal for the lines carried independently and for the duplicate samples of the 

 single line. Since this does not seem to have been the case, it remains uncertain to 

 what extent the fluctuation found reflects the actual variation in the number of pre- 

 formed resistant cells. 



Of great interest are the recent experiments in vitro of Szybalski, 1307 also based on 

 a modified fluctuation test, and dealing with azaguanine resistance of cells in vitro. 

 The strain, designated D98S and originally derived from a single-cell isolate of an 

 established tissue-culture line originating from human bone marrow carried in tissue 

 culture, formed well-defined colonies of uniform size, composed of epithelial-type cells 

 and firmly attached to glass. Since such colonies contained approximately equal 

 numbers of cells, they were regarded as analogous to the series of separate test-tube 

 cultures used in Luria and Delbriick's fluctuation test. A standard inoculum, con- 

 taining approximately 100 cells, was plated in drug-free medium. At various intervals 

 after inoculation, four plates were selected at random. In two of these, the cells were 

 fixed and stained while the remaining two were exposed to various concentrations of 

 azaguanine. The two first plates were used for determining the average colony 

 count per plate, S (reflecting the inoculum size and the plating efficiency), and the aver- 

 age number of cells per colony, N. The other two plates were incubated for an 

 additional period of 10-12 days, with replacement of the azaguanine-containing 

 medium every 3-4 days. The number of colonies arising from resistant mutants (R) 

 were scored after fixation and staining. The expression (S — R)/S was employed 

 to calculate the proportion (P ) of the primary colonies that did not contain resistant 

 mutant cells after the initial incubation and were therefore sloughing off the glass 

 after the addition of the selective medium. On the basis of these data, the average 

 mutation rate to azaguanine resistance was calculated from the equation of New- 

 combe 942 as 4.9 + 1.8 x 10 4 per cell per division cycle. This figure was valid for 

 azaguanine levels between 4 and 8 y per ml., whereas there was an apparent twofold 

 increase of the mutation rate when the level of drug was lowered to 2 y per ml. The 

 variants resistant to azaguanine were inhibited by 8-azaguanosine, but mutants resistant 

 to this compound could be selected in an additional step. With the same method, 

 the mutation rate from azaguanosine sensitivity to azaguanosine resistance was deter- 

 mined as 1 .2 x 1 " 6 per cell per division cycle. No direct mutations from azaguanine 

 sensitivity to azaguanosine resistance could be detected in populations as large as 1 ~ 7 

 cells. 



This approach appears to be very promising and applicable to many other similar 

 problems. With regard to the argument of preadaptive mutations, that is, appearance 

 of mutants in the absence of the drug, it is highly suggestive, although the final proof 



