“The single greatest experiment in the history of biology”

“The single greatest experiment in the history of biology.”  Quite an accolade.  Yet this is how Richard Lenski (no stranger to seminal investigations himself) described work carried out in 1943 by Salvador Luria and Max Delbrück.

Picture yourself in 1943.  Evolution has now been established as the explanation for the seemingly-designed nature of life, but many questions still remain.  One of the greatest in both its simplicity and importance is this: does genetic variation occur because of the action of selection, or is it present before selection acts?  This is so fundamental to our understanding of evolution as to seem obvious; however, before 1943 the issue was still very much up in the air.

The structure of DNA was not to be discovered for another 10 years, somewhat precluding the analysis of protein structure or DNA sequencing.  Luria and Delbrück therefore used a far more elegant method to infer whether bacteria had mutated before or after selection. They first calculated the expected theoretical variability in growth when multiple colonies were exposed to viruses under either hypothesis (mutation-endowed resistance or acquired resistance).  They then experimentally measured growth of multiple cultures of E. coli when exposed to viral challenge, and compared their experimental findings to the theoretical expectations under each hypothesis.  If the experimental measures matched the expectations of one of these hypotheses, the question would be answered.

To carry out their experiment they established hundreds of cultures from a single bacterial cell each, and grew each culture up for a set time.  They then assessed their viral resistance by plating them onto virally-infected agar, and observing how many colonies formed (and therefore how many bacteria from the original culture were resistant).

Their theoretical calculations threw up some interesting findings (trust me here).  They modelled the frequency distribution (where a category is plotted on the x axis, and the frequency of individuals in that category is on the y) of the number of resistant bacteria in multiple individual cultures expected under each hypothesis.  They found that under the mutation hypothesis, there will be a high variance between different cultures, with some cultures with high numbers of resistant bacteria, causing this distribution to have a long tail at the right hand end.  Both of these expectations stem from the many generations in which resistant mutants can arise: if a mutant arises during the first bacterial generation, half of the bacteria in that culture will be resistant to viral infection; if a mutant arises during the last generation, there will be only one resistant bacterium.  In contrast, under the acquired hypothesis, each bacterium has the same probability of becoming resistant to the virus upon being exposed to it, and each becomes immune or not at the same moment (when they are plated onto the viral agar); therefore the frequency distribution under acquired immunity will have very short tails, and low variance between cultures.

So they knew their expectations.  All that remained now was to expose E. coli cultures to viruses, plot the frequency distribution of resistant bacteria, and measure the variance between cultures.

Their findings emphatically supported the mutation hypothesis.  The frequency distributions of the different cultures formed a curve with a long right-hand tail, and revealed a large number of cultures with over nine resistant bacteria, as predicted by a resistant mutant arising early in the bacterial pedigree of a culture.  The variance between different cultures was also massively higher than their theoretical predictions under the acquired immunity hypothesis, and even higher than they predicted under the mutation hypothesis.

These results are simply inexplicable under the acquired immune hypothesis; the probability of a culture having over nine resistant bacteria is astronomically low, and the repeated observation of such cultures is nigh-on impossible.  However, they are perfectly explained by the mutation hypothesis.

Luria and Delbrück’s work also helped to settle a more philosophical question: is evolution a guided process?  By demonstrating that evolutionary change depends upon mutations that are randomly generated, they banished any element of guidance or divine, benevolent intervention from the evolutionary process.  When selection pressure is applied, time is up: the genetic variation has to be there already.




Luria, S.E. and Delbrück, M. (1943) Mutations of bacteria from virus sensitivity to virus resistance.  Genetics 28: 491-511

Lenski, R.E. (2011) Evolution in action: a 50,000-generation salute to Charles Darwin.  Microbe 6:30-33

The Salvador E. Luria Papers.  National Library of Medicine: Profiles in Science, accessed from


Luria-delbruck diagram.svg, Wikipedia, accessed from



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s