A Mechanism for the Development of Resistance to Streptomycin and Penicillin.

Demerec (1945) and Luria (1946), in studying the resistance of staphylococci to penicillin, found that for a given inoculum (approximately 300,000,000 bacteria) there was a variation in resistance of approximately tenfold for all of their strains; e.g., if the inoculum required 0.1 unit of penicillin for complete inhibition, many of the bacteria in the inoculum would be inhibited by as little as 0.01 unit. The titer of 0.1 unit represented the resistance of a relatively small number of bacteria in the inoculum. Demerec indicated that the development of penicillin resistance resulted from the selection of the most resistant bacteria in concentrations of penicillin not completely inhibiting growth. By continued subcultures in higher penicillin concentrations resistance gradually developed through a series of small increments. We found that, in the case of streptomycin (Klein and Kimmelman, 1946) and penicillin, if one examined only several million bacteria, there was a similar relatively small range of variation in resistance. However, by examining very large numbers of bacteria not previouslyexposed to streptomycin, namely, several billion, it was possible to isolate from all of six strains of the shigellae studied (standard inocula of which were inhibited by 3 to 7 units of streptomycin) variants resistant to more than 1,000 units of streptomycin. The presence of these highly resistant variants (approximately one resistant bacterium to one billion susceptible cells) was indicated to be the critical factor for the very rapid development of streptomycin resistance. Clinically and in vitro (Graessle and Frost, 1946) bacteria become resistant to penicillin at a far slower rate than to streptomycin. If the few highly resistant variants are the factors determining the very rapid development of streptomycin resistance, then it follows that highly resistant variants should regularly be found present in strains showing a rapid rate of development in their streptomycin resistance. Highly resistant penicillin variants should be absent or far less frequent in strains showing a slow rate in their development of penicillin resistance. In the present work we have tested several billion bacteria in each member of a group of strains for the presence of highly resistant variants against penicillin and streptomycin. We have found that while highly resistant variants against streptomycin were present in all the strains tested, no highly resistant variants were found against penicillin. These results were reflected by the rates at which the strains became resistant in vitro to the two chemotherapeutic agents.