Engineered ribosomal RNA operon copy-number variants of E. coli reveal the evolutionary trade-offs shaping rRNA operon number

Abstract

Ribosomal RNA (rrn) operons, characteristically present in several copies in bacterial genomes (7 in E. coli), play a central role in cellular physiology. We investigated the factors determining the optimal number of rrn operons in E. coli by constructing isogenic variants with 5-10 operons. We found that the total RNA and protein content, as well as the size of the cells reflected the number of rrn operons. While growth parameters showed only minor differences, competition experiments revealed a clear pattern: 7-8 copies were optimal under conditions of fluctuating, occasionally rich nutrient influx and lower numbers were favored in stable, nutrient-limited environments. We found that the advantages of quick adjustment to nutrient availability, rapid growth and economic regulation of ribosome number all contribute to the selection of the optimal rrn operon number. Our results suggest that the wt rrn operon number of E. coli reflects the natural, 'feast and famine' life-style of the bacterium, however, different copy numbers might be beneficial under different environmental conditions. Understanding the impact of the copy number of rrn operons on the fitness of the cell is an important step towards the creation of functional and robust genomes, the ultimate goal of synthetic biology.

Figure 1.

Genomic position and transcriptional direction of natural (empty arrowheads) and newly inserted (black arrowheads) rrn operons.

Figure 2.

Construction of MG1655 strains with extra rrnB operons. (A) The intermediate construct harboring a new copy of rrnB with a downstream marker gene (for the construction steps see Supplementary Figure S1). (B) Preparation of a landing pad at an arbitrary genomic site of MG1655, insertion of the intermediate rrnB construct into the landing pad, and scarless removal of the antibiotic resistance marker gene. Blue color indicates genomic sequence copied from the rrnB region (blue arrows represent rrn promoters, and hairpins represent T1, T2 terminators). Red arrows indicate λ Red-mediated recombineering steps. Boxes indicate homology arms participating in the recombination events (genomic coordinates in Materials and Methods). Black arrows represent I-SceI sites.

Figure 3.

RNA and protein content of the cells with various rrn operon numbers, compared to the values of wt (MG1655) cells (100%). The RNA (A) and protein (B) content is normalized to DNA content. RNA:protein ratios are shown in (C). Strains are labeled by their rrn operon number and the site of the operon modification. Center lines show the medians; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range; and outliers are represented by circles. At least five measurements per strain were performed (compared to the wt in each experiment). Asterisks indicate significant difference compared to the wt (*P < 0.05, paired samples t-test).

Figure 4.

(A) Microscopic images of selected strains. (B) The size (forward scatter, FSC) of the modified cells was measured by flow cytometry and compared to the wt in each measurement. Center lines show the medians of relative size; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range; and outliers are represented by circles. Results are based on seven independent experiments (20 000 cells per experiment). Asterisks denote significant difference compared to the wt (*P < 0.05, paired samples t-test).

Figure 5.

Growth parameters of the strains grown in rich (LB) medium. Doubling time (A) and lag time (B) of cells are shown (at least 5 biological replicates per strain with 40 technical replicates each). Center lines show the medians; box limits indicate the 25th and 75th percentiles; whiskers extend 1.5 times the interquartile range; and outliers are represented by circles. Asterisks show significant difference compared to the wt (*P < 0.05, **P < 0.001; one-way ANOVA with post hoc Dunnett's test).

Figure 6.

Selection coefficients of the strains with various rrn copy numbers during pairwise competitions with wt cells in batch cultures (A and C) and in a chemostat (B). The medium used in the experiments was LB (A), M9-glucose (B), or as indicated (C). Positive values mean that the winner of the competition is the modified strain. Dots represent results of individual experiments, averages are shown by lines. Asterisks mark significant difference compared to the wt (*P < 0.05, **P < 0.001; one-sample t-test against zero). (D) Progress of competition between the wt strain and rrn6(ΔD) in batch cultures in rich medium (LB), resolved at different growth phases. The % change of wt in a single growth cycle is shown. Samples were taken immediately after re-inoculation (0 time point), at the end of lag phase (2 h) and at the beginning of stationary phase (11 h). Values are averages of six independent 3-day experiments; error bars show standard errors of the means.