Luria and Delbrück’s Fluctuation Test
The randomness of reproduction is well illustrated by the classic experiment of Salvador Luria and Max Delbrück, carried out in 1943 (for some history, see Fox 1998). Until the 1940s, many microbiologists believed that adaptations were acquired in direct response to a change in environment; for example, resistance to infection by bacteriophage was held to be induced by exposure to phage. Luria called microbiology the “last bastion of Lamarckism” (Judson 1995, Chapter 1). Seeing the fluctuating returns from a slot machine suggested to Luria an experiment that would demonstrate whether resistance was due to mutations that arose at random within bacterial cultures, before they were exposed to phage. The experiment was simple. Bacterial populations were grown up from small initial numbers to high density; samples were then taken from these and grown on agar plates in the presence of phage. If several samples were taken from the same population, the numbers of resistant colonies that were seen varied rather little (Fig. WN15.1, red lines). In contrast, samples taken from different populations showed wide variations in numbers of resistant colonies. This variation reflects the random proportions of resistant bacteria within each population before exposure to phage. If a resistance mutation happened to occur early in the growth of the population and reproduced successfully, it would become abundant; on the other hand, mutations might occur late or might be lost from the population by accident. The wide variation in numbers of resistant bacteria present in the replicate populations matched Luria and Delbrück’s predictions based on a simple model of random reproduction.
Luria and Delbrück’s experiment has had renewed interest, in relation to evidence that mutation rates in bacteria increase in response to stress. It has been argued that such an increase might be adaptive and might even be directed toward those genes in which mutations could enhance survival. However, more detailed experiments indicate that the mutations are not targeted in any adaptive way but instead occur without any connection to their potential benefits. For reviews, see Rosenberg (2001), Brisson (2003), Foster (2004), and Roth et al. (2006). See also Web Notes.