Posted by: Dan | May 26, 2006

Molecular evolution in a bottle

This one on an interesting paper in the May 19th issue of Molecular Cell, examining the link between adaptation within populations and the biophysical origins of protein evolution.

Couñago et al. employed a weak link approach in this study, which involves a single essential chromosomal gene being replaced by a homolog that is poorly adapted to the growth conditions of the host organism. The gene in question: for adenylate kinase (AK) in the moderate thermophile Geobacillus stearothermophilus, which was replaced with that of the mesophile Bacillus subtilis, resulting in a strongly temperature-dependent phenotype, where AK is required for proper energy metabolism.

This evolution of this phenotype was studied at the level of a single gene in a large population using in vivo selection ina turbidostat (a fermentor that maintains constant cell density) to monitor populations of bacteria as they evolved to greater fitness. The experiment was conducted over the period of a month, with incremental increases in temperature from 55 to 70 degrees C., and daily samples to examine population changes in AK’s amino acid sequence.

What they found was that a single mutation, Q199R, conferred greater fitness in the initial days of the experiment, between 55 and 61deg C., and that all subsequent mutations above 62-63deg C. were found to be within this Q199R background.

Couñago et al.‘s discussion examines this finding in broader perspective:

Although only those mutations that confer substantive gains in fitness will allow a particular mutant to appear in numbers large enough to be observed or take over the experiment to achieve fixation, it is equally true that the fitness landscape is circumscribed by a mutational landscape as well. The population dynamics therefore reflect both the fitness and mutational landscape that can be traversed as selection proceeds. The early dominance of Q199R ensures that future, more fit mutants will originate in the Q199R background, as there are few wild-type cells to be acted upon by natural selection by 57deg C. Fixation of Q199R is an example of how a particular history of selection influences future generations, a phenomenon Gould referred to as historical contingency in the evolution of more complex life (Gould, 1989). This phenomenon is readily supported by the observation that the single mutation Q16L makes AK substantially more stable than any of the other mutations, with the exception of Q199R/Q16L, but was not observed in the populations. Although it is evident that mutation Q16L could be made by the bacteria, the selection regime favored Q199R and dictated that future mutations would probably arise from it rather than wild-type cells.

And so, we have a well-articulated instance of descent with modification at the molecular level.


  • In vivo molecular evolution reveals biophysical origins of organismal fitness. Couñago R, Chen S, Shamoo Y. Mol Cell. 2006 May 19; 22(4):441-9. Pubmed
  • Wonderful Life: The Burgess Shale and the Nature of History. Gould SJ. New York: Norton; 1989. Amazon


  1. […] About a month ago I went to a talk by Princeton’s Bob Austin titled “Evolution on a Chip,” and it appears that the paper is out in this past week’s PNAS: Bacterial metapopulations in nanofabricated landscapes. The driving force behind the project, he said, was that while bacterial populations are exceptional for studying evolution and popgen in real time, chemostats are bad examples of evolution in action. Afterall, chemostats keep bacterial growth in log phase, with nutrients in excess – and as we all know, biology in the real world deals with competition for limited resources. […]


%d bloggers like this: