438 GENETICS OF SOMATIC CELLS 



of the drug was necessary before the lymphocytic neoplasm P288 developed resistance 

 against amethopterin. Potter 1019 has also obtained data indicating that resistance to 

 azaserine can develop in two different patterns in the plasma-cell neoplasm 70429. 

 There were also differences between the way resistance against DON developed in two 

 different lines of mast-cell neoplasm P815. 



The variations among different neoplasms with regard to the development of 

 resistance against the same drug are reminiscent of the differences in the probability 

 that a given variant phenotype will develop from different tumor lines of closely similar 

 origin and type, 716 and is suggestive of profound differences in the cytogenetics and/or 

 epigenetics of the different lines of tumor cells. The finding that resistance against 

 the same drug can develop in different patterns in the same neoplasm has many parallels 

 in the field of microbial resistance to drugs and is also reminiscent of the finding that 

 the same variant phenotype can be selected in one single step or through intermediate 

 steps from the same neoplasm. 722 



It may now be asked what evidence there is to show that resistance to drugs in 

 neoplastic cells really has a variation-selection basis as postulated. The following ob- 

 servations are relevant. Once established, resistance is usually stable and irreversible, 

 even in the absence of the drug. Resistance is sometimes attained suddenly, but more 

 often progressively, suggesting a series of changes, 1406 and in certain instances discrete, 

 stepwise increases of resistance have been noted. 766 Particular significance has been 

 attributed to the fluctuation test of Luria and Delbriick, 812 modified for this problem 

 by Law, 766 who carried out the experiments with leukemia L1210 in DBA/2 mice, 

 using a transplantation system in vivo. Two different sets of tests were made: one in 

 which leukemic cells were taken from fifteen independent sublines of leukemic cells 

 (originally started from a small, common, cellular pool) and one in which ten repeated 

 tests were made with leukemic cells derived from the same subline. All groups received 

 a standardized inoculum subcutaneously and were given 2.5 mg. of amethopterin per 

 kilogram body weight every other day for 4 doses. At 9 days postinoculation, all 

 mice were sacrificed, and the mean weight of lymphomatous tissue was determined for 

 each group. It was found that this measure showed extreme variability between those 

 groups which represented the independent sublines, while the variability was small and 

 distributed in a uniform probability pattern in the groups inoculated with duplicate 

 samples from the same subline. The differences in the weight of lymphomatous tissue 

 after amethopterin treatment was interpreted as reflecting variations in the number 

 of resistant cells in the different sublines. If this interpretation is correct, highly signi- 

 ficant fluctuations in the number of resistant cells in the independent subline group, 

 started from a common source, would strongly favor the assumption that the appearance 

 of resistant cells is due to random intercellular variation not related to any inducing 

 action of the drug. However, there is one point in this experiment which is not entirely 

 convincing. The fifteen independent sublines were tested by inoculating fifteen differ- 

 ent cellular suspensions prepared artificially in Locke's solution. In contrast, aliquots 

 of the same cellular suspension were used for making repeated tests from a single sub- 



