Fig 1.
Simplified reaction network of EGF signaling pathway model.
Details of reactions are shown in S1 Fig.
Fig 2.
Validation of EGF signaling pathway model.
The time courses of changes in (A) phosphorylated and (B) nuclear ERK levels at different concentrations of EGF (0–50 ng/mL) are shown. EGF dose response of peak levels at (C) phosphorylated and (D) nuclear ERK were calculated from deterministic simulations. Points and lines represent observed data [14] and simulation results, respectively. The effects of ERK-mediated regulation of NPC on the dose response of (E) phosphorylated and (F) nuclear ERK are shown. The Hill coefficients were obtained by curve fitting of simulation results.
Fig 3.
Effects of intrinsic and extrinsic noise on heterogeneity in nuclear ERK.
Distributions of fold changes in nuclear ERK with (A) intrinsic or (B) extrinsic noise are shown. (C) CV of nuclear ERK in simulation results with intrinsic and extrinsic noise.
Fig 4.
Direct comparison of simulation results with experimental data.
(A) CV of simulated nuclear ERK without/with AME are shown. Points represent experimental data [14]. (B) The residual sum of squares between simulation results with AME and experimental data. (C) Mutual information between EGF concentration and nuclear ERK was calculated from simulation results and experimental data [14]. (D) Distributions of nuclear ERK in simulation results at 25% CV of protein variability and experimental data [14].
Fig 5.
Effects of covariation among initial proteins on heterogeneity in ERK.
(A) An example of the relationship among proteins under initial conditions with and without covariation. Points represent different simulation data. (B) The distributions of steady-state levels of nuclear ERK at 25% CV of initial proteins under conditions of no stimulation with and without covariation. (C) CV of peak levels of nuclear ERK with and without covariation with or without application of AME.
Fig 6.
Effects of protein variability of each species on heterogeneity in nuclear ERK.
CV of nuclear ERK with changes in protein variability of each species at (A) low (0.05 ng/mL) and (B) high (50 ng/mL) concentrations of EGF. Circle size represents CV of nuclear ERK. (C) CV of nuclear ERK at various concentrations of EGF when changing protein variability of each species. The patterns of CV were classified into three types. CV of 25% was used as protein variability of each species. The gray region represents the area from EC10 (0.02 ng/mL) to EC90 (0.12 ng/mL) calculated from Fig 1D.