Figure 1.
Scheme of glycolysis model (B) and its link to the segregated cell growth model (A) established previously [23].
Green boxes represent metabolite pools that were quantified experimentally while white ones were not measured. Enzymes are shown as ellipses with blue background if the maximum enzyme activity was measured in vitro and with white background otherwise. Reactions and their directions are shown as arrows. Dashed arrows represent allosteric regulation of enzymes by metabolites. The activity of the enzymes relative to the highest activity in glycolysis (see legend) is attached to the corresponding reactions/enzymes and expressed by colored bars (blue: cell growth at 24 h of Cult1; green: cell maintenance at 100 h of Cult1; orange: limitation at 6 min). Absolute flux rates (unit: mmol L−1 min−1) are given next to the bars. GLCx extracellular glucose; GLC glucose; G6P glucose 6-phosphate; UGLC uridyl diphosphate glucose; R5P ribose 5-phosphate; PPP pentose phosphate pathway; F6P fructose 6-phosphate; F16BP fructose 1,6-bisphosphate; 3PG 3-phosphoglyceric acid; PEP phosphoenol pyruvate; HK hexokinase; UT UTP-glucose-1-phosphate uridylyltransferase; G6PDH glucose 6-phosphate dehydrogenase; GPI glucose phosphate isomerase; ALD aldolase; ENO enolase; PK pyruvate kinase.
Figure 2.
Metabolites pools of glycolysis during adherent MDCK cell cultivation.
Glucose 6-phosphate (A–C), fructose 6-phosphate (D–F), fructose 1,6-bisphosphate (G–I), 3-phosphoglyceric acid (J–L) and phosphoenolpyruvate (M–O) concentrations in three independent MDCK cell cultivations (Cult1 ▵, Cult2 □, Cult3 ○) in 6-well plates and GMEM-Z. Data and error bars represent mean and standard deviation of three wells. Dashed lines are the limit of quantification (LOQ; data below LOQ marked in grey). Lines represent the respective simulation result based on the experiment-specific parameters in Table 1 and parameters in Table 2. The intermediate growth phase (95%–5% proliferating cells) is indicated as grey bar for the respective cultivation.
Table 1.
Initial conditions for the structured model comprising metabolic status, growth status and culture conditions for the simulated experiment.
Figure 3.
Estimated fluxes for energy and precursors production during adherent MDCK cell cultivation.
Net flux into pentose phosphate pathway (PPP) relative to glucose transport flux (A, see section “Exchange of glycolytic metabolites through other reactions”), net flux into glycogenesis relative to glucose transport flux (B), glucose transport flux (C), and ATP production rate (D) are simulated for the three cultivations (Cult1 – 3) and shown in the color code of Fig. 2.
Figure 4.
The response of intracellular metabolite pools to perturbation experiments.
Glucose 6-phosphate (A–C), fructose 6-phosphate (D–F), fructose 1,6-bisphosphate (G–I), 3-phosphoglyceric acid (J–L) and phosphoenolpyruvate (M–O) concentrations of three independent perturbation experiments with MDCK cells in 6-well plates. Cells originating from a cultivation experiment (see Table 1) were deprived of extracellular nutrients by removal of medium and addition of phosphate buffered saline, shown in the first column (Lim1, A,D,G,J,M) and second column (Lim2, B,E,H,K,N). After a 2 h limitation, PBS was exchanged by fresh medium (Pulse, C,F,I,L,O). Data (○) and error bars represent mean and standard deviation of three wells, respectively. Dashed lines are the limit of quantification (LOQ; data below LOQ marked in grey). Lines represent the respective simulation result based on the experiment-specific parameters in Table 1 and parameters in Table 2.
Figure 5.
Intracellular metabolite pools of pentose phosphate pathway and glycogenesis during adherent MDCK cell cultivation.
Ribose 5-phosphate (A–C) and uridyl diphosphate glucose (D–F) concentrations of three independent MDCK cell cultivations (Cult1 ▵, Cult2 □, Cult3 ○) in 6-well plates and GMEM-Z. Data and error bars represent mean and standard deviation of three wells, respectively. Dashed lines are the limit of quantification (LOQ; data below LOQ marked in grey). Lines represent the respective simulation result based on the experiment-specific parameters in Table 1 and parameters in Table 2. The intermediate growth phase (95%–5% proliferating cells) is indicated as grey bar for the respective cultivation.
Figure 6.
The response of metabolite pools of pentose phosphate pathway and glycogenesis to perturbation experiment.
Ribose 5-phosphate (A–C) and uridyl diphosphate glucose (D–F) concentrations in three independent perturbation experiments with MDCK cells in 6-well plates. Cells originating from a cultivation experiment (see Table 1) were deprived of extracellular nutrients by removal of medium and addition of phosphate buffered saline, shown in the first column (Lim1, A,D) and second column (Lim2, B,E). After 2 h of incubation, PBS was exchanged by fresh medium (Pulse, C,F). Data (○) and error bars represent mean and standard deviation of three wells, respectively. Dashed lines are the limit of quantification (LOQ; data below LOQ marked in grey). Lines represent the respective simulation result based on experiment-specific parameters in Table 1 and parameters in Table 2.
Figure 7.
Impact of in silico GLUT modulation on (A) ATP and (B) pentose phosphate pathway (PPP) production rates.
At total of 1900 model parameterizations (0.025–0.975 quantile of 2000 model parameterizations) were assessed for the GLUT modulation and were derived from the optimal result of each bootstrap run, which was also the basis for estimation of the parameter confidence intervals of Table 2. The colored bars on the right hand show the respective production rate; the black vertical line represents the original GLUT activity of cells of Cult1 at 24 h.
Figure 8.
Prediction of glycolytic metabolite pools during cultivation of adherent MDCK cells in DMEM with 2.5 mmol L−1 extracellular glucose.
Glucose 6-phosphate (A), fructose 6-phosphate (B), fructose 1,6-bisphosphate (C), 3-phosphoglyceric acid (D) and phosphoenolpyruvate (E) concentrations during MDCK cell cultivations in 6-well plates and DMEM medium with 2.5 mmol L−1 extracellular glucose. Data (◊) and error bars represent mean and standard deviation of three wells. Dashed lines are the limit of quantification (LOQ; data below LOQ marked in grey). Lines represent the model prediction based on the modifications of the cell growth model described in the supporting information 4 and the parameters in Table 1 and Table 2. The intermediate growth phase (95%–5% proliferating cells) is indicated as grey bar.
Table 2.
Measured and estimated parameters of adherent MDCK cell glycolysis used to simultaneously capture all experiments of this study with confidence intervals between 0.025-quantile and 0.975-quantile
.