Figure 1.
Growth rate as a function of eGFP expression in different growth media.
A. Growth rate plotted as a function of eGFP mass fraction of dry cell weight (d.c.w.) for promoter variants grown in M9 (blue open squares), M9-CA (green open triangles), and LB (red open circles). Dashed lines are best linear fits of each dataset. B. Relative growth rate plotted as a function of eGFP mass fraction of d.c.w. for promoter variants. The growth rate is normalized to the maximum growth rate in the absence of gene expression. The dashed olive lines represent the range of values from theoretical predictions. Error bars represent one standard deviation of at least two independent measurements.
Figure 2.
Growth rate dependence on amidase expression in different E. coli strains.
A. Sample Western blot of pJK_proB_amiE expressed in E. coli TUNER. T – total cell lysate, I – insoluble fraction of the cell lysate, S – soluble fraction of cell lysate. B. Quantification of insoluble amiE relative to total amiE expressed for plasmid pJK_proB_amiE expressed in TUNER and in MG1655rph+. The percentage of insoluble amiE was determined by Western blots, while the absolute amount of soluble amiE was determined by enzyme assay of total cell lysates. C. Relative growth rate plotted as a function of soluble (open symbols) or total (insoluble plus soluble; closed symbols) amiE mass fraction of d.c.w. for promoter variants grown in M9-CA for E. coli TUNER (orange circles) or MG1655rph+ (blue diamonds). Error bars represent one standard deviation of at least three independent measurements. The solid lines represent linear best fits through the combined datasets. The dashed olive lines represent the range of values from theoretical predictions.
Figure 3.
Experimental validation of predicted promoter strengths.
For each promoter variant driving eGFP expression, the bundled promoter activity α (A), αo (C), and eGFP mass fraction (B,D) of cells grown in M9-CA are plotted against values of cells grown in M9 (blue open squares) or LB (red open circles). Dashed lines indicate the predictions from the promoter activity model in the absence (A,B) or presence (C,D) of a correction term for promoter activity under weaker growth rates. Note the strong deviations from prediction in the absence of the correction term when cells are grown in M9. Error bars are one standard deviation from at least two independent measurements.
Figure 4.
The protein fraction of dry cell weight (ΦU, triangles) and relative growth rate (μ/μmax, squares) necessary to support a flux of 10 mmol product per gDCW per h is plotted against the number of pathway enzymes involved in a secondary metabolic pathway.
Each enzyme is assumed to have a turnover number of 2.6 s−1 (the median value for secondary pathway enzymes) and have a molecular weight of 40 kDa. Because saturating kinetics and reaction irreversibility are assumed, this is strictly a lower bound on the fitness cost. Lines are guides for the eye.