€8 M. DEMEREC [VOL. 56 



hand, the variation in number of resistant colonies among platings of samples 

 taken from one culture was very small; and the probability that this variation 

 was due to chance is 26 per 100 trials. Very similar results were obtained in 

 experiments using S. aureus and penicillin (Demerec, 1945a). 



These results, then, favor the assumption that resistance to certain concentra- 

 tions of penicillin or streptomycin originates through mutation, and that resistant 

 bacteria may be found in any large population, the proportion depending on the 

 mutation rate. 



Oakberg and Luria (1947) reached identical conclusions after experimenting 

 with S. aureus and sodium sulfathiazole. This suggests that mutations may be 

 generally responsible for the origin of resistance that is transmitted to the off- 

 spring of the individuals that acquire it. 



RESISTANCE STEPS 



A very interesting feature of bacterial resistance to antibiotics is the stepwise 

 increase in degree of resistance that can be brought about by selection. This 

 feature is particularly well expressed in penicillin resistance. Figure 3 reproduces 

 curves from an earlier paper (Demerec, 1945a) showing the effect of selection on 

 the increase in resistance of S. aureus to penicillin. The first is the survival curve 

 of the stock culture. At a concentration of 0.15 units per ml there were no sur- 

 vivors, but at a concentration of 0.12 units about 4 per 10 8 bacteria lived. First- 

 step resistant strains were isolated from stock culture bacteria surviving sub- 

 lethal concentrations. The second curve of figure 3 is a typical survival curve 

 of such first-step resistant strains. Some individuals of these strains survived 

 concentrations up to about 0.2 units. When first-step resistant strains were 

 grown on sublethal concentrations of penicillin, second-step resistant strains 

 were isolated from the survivors. A typical second-step survival curve is shown 

 third on figure 3. Third- and fourth-step resistant strains were obtained in 

 similar manner. 



It is of interest to note that the building up of resistance is more rapid with each 

 selection step. Thus, with our strain of S. aureus, a concentration of 0.15 units 

 was sufficient to eliminate all bacteria of the original strain, but a concentration 

 of about 0.2 units was required to eliminate all bacteria of the first-step resistant 

 strain, and concentrations of about 0.4 units for the second-step, 1 unit for the 

 third-step, and 7 units for the fourth-step. The fifth-step strain was for all 

 practical purposes completely resistant to penicillin, since it was not affected by 

 a concentration of 250 units per ml. With each step the increase in resistance 

 appeared to be exponential. 



Whereas the building up of resistance to penicillin followed a definite pattern, 

 resistance to streptomycin showed a considerable degree of variability. Among 

 first-step resistant strains — that is, among strains isolated from colonies of the 

 original strain that survived sublethal doses of streptomycin — there were some 

 that were only slightly more resistant than the original strain, some that were 

 almost completely resistant, and some that fell between these two extremes 

 (figure 4). It has been found that the variability in degree of resistance among 



84 



