Different Levels of Catabolite Repression Optimize Growth in Stable and Variable Environments
(A) Example of a growth curve showing the biphasic growth associated with the switch from one carbon source to another (diauxic shift) of a strain (YS4) growing in the presence of LG supplemented with galactose. The figure shows a marked decrease in growth rate (lag phase) during the switch from glucose to maltose. MaxR is the maximal growth rate (or maximal fitness) attained at the beginning of the experiment when glucose levels are high, and correspond closely with growth rate measurements made of cells growing in very dilute conditions (not shown). GMR is a measure of average fitness throughout the experiment, calculated as the average growth rate between two preset cell densities that represent the beginning of measurable growth and the onset of stationary phase (see Materials and Methods for details). (B) Same data as panel (A), where the instantaneous growth rates of the culture are plotted as a function of population size. This representation shows more clearly how growth decelerates during the lag phase, leading to the often large difference between MaxR and GMR (black dotted line). (C and D) Growth pattern of a reference strain (S288c) with a pronounced lag phase growing either HG conditions (3% glucose, green) or 0.5% glucose, either alone (red) or supplemented with galactose (blue), maltose (purple). The growth rate in 3% glucose is relatively stable, whereas growth rates in the other media are more variable, with a temporary decrease typical of the lag phase when cells shift their metabolism from glucose to another carbon source. (E and F) Similar to (B) and (C) but with a strain UWOPS83-787.3 that shows almost equal fitness in different media. Note that the lag phase is barely detectable, and that growth only slows down at the end of the experiment, probably because of the depletion of nutrients and the accumulation of ethanol and other toxic metabolites. (G) Live-cell microscopy of yeast populations shifting between glucose and other carbon sources allows measurement of the lag phase of individual cells. Each curve represents the cumulative distribution histogram of single-cell lag phases of 1 of 18 different yeast strains. Each trace represents the fraction of a population of one given strain that has escaped the lag phase after a transfer from glucose to maltose as measured by budding events (Materials and Methods). The histograms reveal large differences in lag duration between strains, as well as variation in lag duration between individual cells within populations. One strain was omitted from this analysis because fewer than 1 in 150 cells resumed growth after transition to maltose. (H) Correlation between the average single-cell lags from (1 g) and population-level fitness variability (i.e., the variability of the GMR across different growth media). The vertical axis shows the average duration of a strain's lag phase (as measured by single-cell live microscopy), and error bars on this axis correspond to the lower and upper quartiles. The size of each data point is proportional to the fraction of cells that were observed to resume growth after transition to maltose. The horizontal axis represents the ratio of a strain's fitness in media requiring diauxic shift (LG, LG + galactose, and LG + maltose), relative to its fitness in stable HG conditions. Error bars on this axis are the standard deviations of 1,000 repeated calculations of the statistic obtained by random sampling of one biological replicate from each condition (n = 2–6 per strain in each condition). See also main text, Dataset S1, and Figure S1.