Figure 6: Analysis of RNAP-dependent intrinsic and extrinsic noise at saturation conditions.

(a) Intrinsic, (b) extrinsic and (c) total noises of the sub-sampled cells depending on RNAP noise using the high-copy-number plasmid (pNL002). We expressed T7 RNAP at a sufficiently high level using pNL002 to saturate the expression levels of CFP and mCherry (Supplementary Fig. 7), and generated sub-sampled cell distributions of RNAP concentrations with pre-determined means and s.d. from the total collection of cells (Supplementary Fig. 8). The dotted lines represent linear guidelines for clarity. Error bars indicate±s.d., n=1,000. (d) The slope (susceptibility to RNAP noise) of the extrinsic noise depending on the T7 RNAP concentration fromFigs 3c and 6b. The red line is fitted to the function of the unoccupied fraction of the promoter by RNAP, a/(1+KR), where R and K denote T7 RNAP concentration and the RNAP affinity for the promoter, respectively (least squares, R²=0.99). Error bars indicate±s.d., n=10. (e) Comparison of mCherry noise induced by various E. coli RNAP promoters with the noise induced by T7 RNAP promoters. The rectangle, triangle and diamond symbols represent the gene expression noise under various E. coli RNAP promoters²⁰. Expression levels in each strain were modulated by applying different IPTG concentrations. When mean mCherry intensity is >6,000 (a.u.), noise becomes saturated at ~0.1 (dashed line), consistent with the findings of Taniguchi et al.⁹ Error bars represent the mean±s.d. of three independent experiments. The red-filled circles represent gene expression noise under the T7 promoter when the noise induced by the RNAP noise is removed, as estimated from the linear fitting shown in Fig. 3c. The blue-filled circle represents gene expression noise under the T7 promoter at saturation-level T7 RNAP concentrations using pNL002.