The protein burden of lac operon products.

A new approach to measuring the slowing of growth due to the manufacture of proteins not needed by a bacterium is presented. An entire single colony of Escherichia coli was used to start a chemostat culture that was then given a selective pressure by the addition of phenylgalactoside (phi-gal). This enriched the population for constitutive mutants that produced beta-galactosidase without induction and could split phi-gal, consume the galactose, and grow faster. When the phi-gal was removed, the constitutives grew slower than the parental strain and were gradually lost. This procedure allows competition experiments to be carried out with minimum effects due to genetic drift. Experiments with both strains having wild-type and mutant permease genes were conducted. With the former the selective disadvantage was initially much greater than expected from the simplest hypothesis that extra unused proteins would slow growth in proportion to their fraction of the total protein synthesis. This phase was followed by a second phase where the selective disadvantage was smaller than predicted by this simple hypothesis. With a very slowly reverting permease negative strain the selective disadvantage, and therefore the protein burden, was found to be much smaller and not statistically different from zero. Thus, while one would expect under carbon and energy limitation in the chemostat the protein burden to be larger than under unlimited conditions, it is so small that even the refined technique used here could not measure it accurately. It is certainly less than the fraction of 'waste' protein synthesis; but it could be between zero and the fraction of the cells' energy and carbon budget spent on manufacture of the proteins of the lac operon.