Escherichia coli with a Tunable Point Mutation Rate for Evolution Experiments

Abstract

The mutation rate and mutations' effects on fitness are crucial to evolution. Mutation rates are under selection due to linkage between mutation rate modifiers and mutations' effects on fitness. The linkage between a higher mutation rate and more beneficial mutations selects for higher mutation rates, while the linkage between a higher mutation rate and more deleterious mutations selects for lower mutation rates. The net direction of selection on mutations rates depends on the fitness landscape, and a great deal of work has elucidated the fitness landscapes of mutations. However, tests of the effect of varying a mutation rate on evolution in a single organism in a single environment have been difficult. This has been studied using strains of antimutators and mutators, but these strains may differ in additional ways and typically do not allow for continuous variation of the mutation rate. To help investigate the effects of the mutation rate on evolution, we have genetically engineered a strain of Escherichia coli with a point mutation rate that can be smoothly varied over two orders of magnitude. We did this by engineering a strain with inducible control of the mismatch repair proteins MutH and MutL. We used this strain in an approximately 350 generation evolution experiment with controlled variation of the mutation rate. We confirmed the construct and the mutation rate were stable over this time. Sequencing evolved strains revealed a higher number of single nucleotide polymorphisms at higher mutations rates, likely due to either the beneficial effects of these mutations or their linkage to beneficial mutations.

Keywords: Experimental Evolution; Mismatch Repair.

Figure 1

Diagram of the mCherry-mutH construct. mCherry is translationally fused to the N-terminus of mutH. Both mCherry-mutH and tetR are expressed from $P_{LTetO1}$ promoters, meaning their expression is normally suppressed but can be induced by anhydrotetracycline. The regulation of tetR expression by TetR protein makes the response of the system to anhydrotetracycline less sensitive (Nevozhay et al. 2009)

Figure 2

a Mean mCherry-MutH expression as a function of the concentration of the inducer of mCherry-MutH expression anhydrotetracycline (aTc). The green line is the expression of mCherry-MutH from the wildtype locus which is not sensitive to anhydrotetracycline (aTc). The red line is the level of fluorescence measured from a control strain (ME121) which expresses no mCherry. The light blue points are the mean expression measured by fluorescent imaging. The dark blue curve is a Hill function plus shift fit to that data $y=\frac{A}{\left(1+{\left(\frac{k_a}{x}\right)}^n\right)}+C$. The parameters were $A=.22$, $k_a=42$, $n=1.6$, and $C=1.6\times {10}^{-3}.$ b The mutation rate to rifampicin resistance per cell division as a function of the aTc concentration. The red line is the mutation rate of ME121 which is defective for mismatch repair. The green line is the mutation rate of ME120 which has fully functioning mismatch repair. The light blue points are data from NS001 at different levels of aTc induction of mutH expression. mutL induction was saturated with 2000 μM of IPTG. The dark blue curve is a Hill function plus shift fit to that data $y=\frac{A}{\left(1+{\left(\frac{k_a}{x}\right)}^n\right)}+C$. The parameters were $A=-8.6\times {10}^{-8}$, $k_a=.57$, $n=2.9$, and $C=8.8\times {10}^{-8}.$ c The mutation rate to rifampicin resistance per cell division vs. the expression of mCherry-MutH. The orange data are MutH expression and mutation rates of NS001 grown with varying concentrations of aTc. The green data are the expression of MutH from the wildtype locus and the mutation rate of ME120. The blue curve uses the mean expression of mCherry-MutH from the hill curve fit of expression vs. aTc as x and the mutation rate from the hill curve fit of mutation rate vs. aTc as y.

Figure 3

First row - optical density over time for (A) A single well on the first day of the experiment, (B) All wells on the first day, and (C) All wells on the last day of the experiment. Second row - first derivative of optical density over time plotted vs. the optical density for (D) the same well as in A, (E) the same wells as in B, and (F) the same wells as in C.

Figure 4

Evolution of the growth rate in early exponential over time. Slashes have been placed on the x-axis where the experiment was frozen and resumed later. Blue circles - Low mutation rate; orange triangles - LoMid mutation rate; green squares - Mid mutation rate; red diamonds - HiMid mutation rate; purple stars - High mutation rate. a Mean growth rate in exponential phase by mutation rate condition over time. b Coefficient of variation of the the growth rate in exponential phase by mutation rate condition over time.

Figure 5

Evolution of the time before leaving exponential phase growth a The mean time before the optical density exceeded that of exponential phase growth (OD600 = 0.16) by mutation rate condition. Blue circles - Low mutation rate; orange triangles - LoMid mutation rate; green squares - Mid mutation rate; red diamonds - HiMid mutation rate; purple stars - High mutation rate. b The coefficient of variation of the time before the optical desnity exceeded that of exponential phase growth by mutation rate condition. Blue circles - Low mutation rate; orange triangles - LoMid mutation rate; green squares - Mid mutation rate; red diamonds - HiMid mutation rate; purple stars - High mutation rate.

Figure 6

Evolution of saturating optical density over time. Slashes have been placed on the x-axis where the experiment was frozen and resumed later. a - mean saturating optical density by mutation rate condition over time. Blue circles - Low mutation rate; orange triangles - LoMid mutation rate; green squares - Mid mutation rate; red diamonds - HiMid mutation rate; purple stars - High mutation rate. b - Coefficients of variation of saturating optical density by mutation rate condition over time. Blue circles - Low mutation rate; orange triangles - LoMid mutation rate; green squares - Mid mutation rate; red diamonds - HiMid mutation rate; purple stars - High mutation rate. c - Saturating optical density over time for replicates in the high mutation rate condition. d - Saturating optical density over time for replicates in the low mutation rate condition.

Figure 7

The mean number of cumulative SNPs for each mutation rate condition. Shading is an estimate of the standard error of the mean for each mutation rate condition. Blue - Low mutation rate; orange - LoMid mutation rate; green - Mid mutation rate; red - HiMid mutation rate; purple - High mutation rate. a is on day 24. b is on day 41.