Discussion 277 



neither they nor Rountree have isolated a penicilhnase-producing 

 staphylococcus from a penicillin-sensitive parent, this is mere speculation. 



Davis: Dr. Barber has emphasized that penicillinase-producing mutants 

 of staphylococci are difTicult to obtain by mutation and selection in vitro, 

 presumably because their resistance to penicillin depends very much on 

 population density and is negligible for single bacterial cells. I gather 

 that this difliculty in selection is one of the reasons for the belief that the 

 appearance of these organisms in a patient is due generally to spread 

 from another patient, rather than to a fresh mutation of sensitive 

 staphylococci. However, I wonder whether the selection of such a 

 mutant in vivo might not be favoured by a factor that is not ordinarily 

 found in vitro; namely, wide fluctuations in drug concentration in the 

 course of time. During a drug-free interval a penicillinase-producing 

 mutant that appeared in the tissues might possibly establish a localized 

 clone that would resist the subsequent invasion of penicillin. 



Lederberg: Dr. Barber, you did show a case where you could demon- 

 strate minute proportions of penicillinase-producing organisms in sensi- 

 tive populations, but what were the quantitative features of this ? What 

 fraction of the input of penicillinase producers were you able to recover? 



Barber: The smallest fraction we could detect was one penicillinase- 

 producing cell mixed with 10^ sensitive cells. This was on gradient 

 plates. 



Prof. Davis's points are very interesting and very important ones. 

 First of all, as regards the transfer of a large enough population from 

 patient to patient, you do not need to have a large number because in 

 the first place, although they are selected out in a patient on penicillin 

 they next pass to a patient who is not on penicillin. They then get 

 established in that patient who has not had penicillin ; that is not difficult. 

 On the question of the wave of penicillin resistance, I am quite certain 

 that something has happened to staphylococci, since the introduction of 

 penicillin; but I think it is something that has taken several years to 

 happen even in any single strain. I don't think there is any selective 

 advantage to these strains except penicillinase production; but this, 

 together with the elimination of other bacteria by antibiotics, is an 

 advantage of no mean order and of course by passage in hospital they 

 may become of increased virulence. On the last point, we cannot exclude 

 the possibility that in experiments with penicillin-sensitive staphylococci 

 we are in fact getting penicillinase-producing mutants which are not 

 appearing under the particular conditions of the test. That, of course, 

 is possible, but the mutation must be a rare one. 



With regard to Szybalski's work, I am not saying that Szybalski did 

 not isolate a penicillinase-producing variant of a clinical type, in a single 

 stage, on a gradient plate. I think the phenomenon may happen, but 

 certainly it is a very rare one, and the evidence given by Szybalski is not 

 conclusive. 



Knox: Is there any evidence on this from experimental animals? 

 Has an epidemic of any sort ever been kept going in animals treated and 

 not treated with penicillin, to see whether they eventually produce 

 penicillinase-producing staphylococci ? 



