Model initial conditions and parameters
| Species/Constant | Value | Rationale |
| R for M1 | 500 µm−2 | From fluorescencea |
| R for eP2Y2 | 1 µm−2 | Fig. S7; Fig. 2 A in Dickson et al., 2013 |
| G (G proteins) | 40 µm−2 | To fit concentration–response curve of currenta |
| P (PLC) | 10 µm−2 | To fit concentration–response curve of currenta |
| (free) PIP2 | 5,000 µm−2 | From distribution of PH domainsb |
| Bound PIP2 | 10,000 µm−2 | See Fig. 6 B; generally (fold_PIP2-1) * 5,000 µm−2 |
| PI(4)P | 4,000 µm−2 | To allow doubling of PIP2 by PIP 5-kinase |
| PI | 140,000 µm−2 | As previouslyb |
| IP3 (cytosol) | 0.01 µM | Steady state of basal PLC and IP3ase |
| DAG | 23 µm−2 | Steady state of basal PLC and DAGase |
| Ca2+ (cytosol) | 0.13 µM | Steady state of basal IP3R Ca2+ flux and SERCA (bistable equilibrium, also stable at 0.02 µM) |
| Ca2+ (ER) | 400 µM | Jafri and Keizer, 1995; Duman et al., 2008 |
| h | 0.8 | h3 is the fraction of noninactivated IP3Rs |
| LIBRAvIII | 6 µM | As estimated from fluorescencea |
| IP3 (pipette) | 1–100 µM | Used only to simulate calibration experiment |
| Fura-4F | 1 µM | Excessive buffering in the model above 1 µM |
| C1 domain | 0.5 µM | From fluorescence assuming low expressiona; reproduces 1 nM Oxo-M response in relation to 10 µM Oxo-M response |
| size_pipette | 10,000 µm3 | Reproduce steady-state values of Fig. 6 D in Dickson et al., 2013 |
| Surface (membrane) | 1,500 µm2 | From cell capacitance |
| size_cytosol | 2,500 µm3 | From surface–volume ratio |
| size_ER | 462 µm3 | 18% cytosol, from Jafri and Keizer, 1995, and Duman et al., 2008. |
| fold_PIP2 | 3 | See Fig. 6 B |
| k_4K (rest) | 0.00078 s−1 | KCNQ2/3 current recovery after Oxo-Mb; generally fold_PIP2 * 0.00023 s−1 |
| k_5K (rest) | 0.06 s−1 | KCNQ2/3 current recovery after VSPb; generally fold_PIP2 * 0.02 s−1 |
| k_4K (agonist) & k_5K (agonist) | see Fig. 7 (E and F) c | Concentration dependence informed by Oxo-M dose–response curveb; transition smoothed by an exponential to avoid transients |
| k_4K (recovery) & k_5K (recovery) | see Fig. 7 (E and F) c | KCNQ2/3 current recoveryb; transition smoothed by an exponential to avoid transients |
| k_PLC | 0.6 µm2s−1 | From KCNQ2/3 inhibitionb; fold_PIP2 * 0.2 µm2s−1 |
| k_IP3ase | 0.08 s−1 | Reproduce duration of LIBRAvIII and Fura-4F responses (Figs. 6 and 4 in Dickson et al., 2013); as in Xu et al., 2003 |
| k_DAGase | 0.05 s−1 | To fit C1 decay |
| KIP3 (IP3R) | 0.1 µM | To have maximum activity with 1 nM Oxo-M |
| KCa (IP3R) | 0.2 µM | Bezprozvanny et al., 1991 |
| kCa (IP3R) | 0.2 µM | Bezprozvanny et al., 1991 |
| kP (SERCA) | 1.3 µM | Duman et al., 2008 |
| vP (SERCA) | 0.3 | Height of plateau in Fura-4F response to Oxo-M; see Fig. S8 B in Dickson et al., 2013 |
| k_pipette | 0.03 s−1 | Giving approximately a time constant of 50 s as observed for diffusion of dyes and the onset of Fig. 6 B in Dickson et al., 2013 |
| KD_LIBRAvIII | 0.5 µM | Tanimura et al., 2009 |
| KD_C1 | 0.3 µM | Oxo-M concentration–response of C1/CAAX FRET |
| KD_Fura-4F | 0.77 µM | Invitrogen |
| Species/Constant | Value | Rationale |
| R for M1 | 500 µm−2 | From fluorescencea |
| R for eP2Y2 | 1 µm−2 | Fig. S7; Fig. 2 A in Dickson et al., 2013 |
| G (G proteins) | 40 µm−2 | To fit concentration–response curve of currenta |
| P (PLC) | 10 µm−2 | To fit concentration–response curve of currenta |
| (free) PIP2 | 5,000 µm−2 | From distribution of PH domainsb |
| Bound PIP2 | 10,000 µm−2 | See Fig. 6 B; generally (fold_PIP2-1) * 5,000 µm−2 |
| PI(4)P | 4,000 µm−2 | To allow doubling of PIP2 by PIP 5-kinase |
| PI | 140,000 µm−2 | As previouslyb |
| IP3 (cytosol) | 0.01 µM | Steady state of basal PLC and IP3ase |
| DAG | 23 µm−2 | Steady state of basal PLC and DAGase |
| Ca2+ (cytosol) | 0.13 µM | Steady state of basal IP3R Ca2+ flux and SERCA (bistable equilibrium, also stable at 0.02 µM) |
| Ca2+ (ER) | 400 µM | Jafri and Keizer, 1995; Duman et al., 2008 |
| h | 0.8 | h3 is the fraction of noninactivated IP3Rs |
| LIBRAvIII | 6 µM | As estimated from fluorescencea |
| IP3 (pipette) | 1–100 µM | Used only to simulate calibration experiment |
| Fura-4F | 1 µM | Excessive buffering in the model above 1 µM |
| C1 domain | 0.5 µM | From fluorescence assuming low expressiona; reproduces 1 nM Oxo-M response in relation to 10 µM Oxo-M response |
| size_pipette | 10,000 µm3 | Reproduce steady-state values of Fig. 6 D in Dickson et al., 2013 |
| Surface (membrane) | 1,500 µm2 | From cell capacitance |
| size_cytosol | 2,500 µm3 | From surface–volume ratio |
| size_ER | 462 µm3 | 18% cytosol, from Jafri and Keizer, 1995, and Duman et al., 2008. |
| fold_PIP2 | 3 | See Fig. 6 B |
| k_4K (rest) | 0.00078 s−1 | KCNQ2/3 current recovery after Oxo-Mb; generally fold_PIP2 * 0.00023 s−1 |
| k_5K (rest) | 0.06 s−1 | KCNQ2/3 current recovery after VSPb; generally fold_PIP2 * 0.02 s−1 |
| k_4K (agonist) & k_5K (agonist) | see Fig. 7 (E and F) c | Concentration dependence informed by Oxo-M dose–response curveb; transition smoothed by an exponential to avoid transients |
| k_4K (recovery) & k_5K (recovery) | see Fig. 7 (E and F) c | KCNQ2/3 current recoveryb; transition smoothed by an exponential to avoid transients |
| k_PLC | 0.6 µm2s−1 | From KCNQ2/3 inhibitionb; fold_PIP2 * 0.2 µm2s−1 |
| k_IP3ase | 0.08 s−1 | Reproduce duration of LIBRAvIII and Fura-4F responses (Figs. 6 and 4 in Dickson et al., 2013); as in Xu et al., 2003 |
| k_DAGase | 0.05 s−1 | To fit C1 decay |
| KIP3 (IP3R) | 0.1 µM | To have maximum activity with 1 nM Oxo-M |
| KCa (IP3R) | 0.2 µM | Bezprozvanny et al., 1991 |
| kCa (IP3R) | 0.2 µM | Bezprozvanny et al., 1991 |
| kP (SERCA) | 1.3 µM | Duman et al., 2008 |
| vP (SERCA) | 0.3 | Height of plateau in Fura-4F response to Oxo-M; see Fig. S8 B in Dickson et al., 2013 |
| k_pipette | 0.03 s−1 | Giving approximately a time constant of 50 s as observed for diffusion of dyes and the onset of Fig. 6 B in Dickson et al., 2013 |
| KD_LIBRAvIII | 0.5 µM | Tanimura et al., 2009 |
| KD_C1 | 0.3 µM | Oxo-M concentration–response of C1/CAAX FRET |
| KD_Fura-4F | 0.77 µM | Invitrogen |
Falkenburger et al., 2010a.
Falkenburger et al., 2010b.
The steady-state Oxo-M concentration dependence of PIP2 synthesis was inferred from the concentration dependence of KCNQ2/3 inhibition (Fig. 9 C in Falkenburger et al., 2010b). Specifically, k_4K values that reproduced the Oxo-M concentration dependence of KCNQ2/3 inhibition (markers in Fig. 7 E) were fitted by a sigmoid and the concentration dependence of k_5K computed with the same midpoint and slope. The extent of PI 4-kinase and PIP 5-kinase acceleration was fine-tuned to reproduce the >90% depletion of PIP2 and 80% depletion of PIP measured biochemically for 10 µM Oxo-M (i.e., 7.5-fold acceleration of the PI 4-kinase 10-fold acceleration of the PIP 5-kinase). In addition, the onset of the acceleration of PIP2 synthesis was smoothed by an exponential to avoid positive PIP2 transients resulting from too fast acceleration (<1-s time constant); the recovery of PIP2 synthesis to resting values was smoothed by an exponential (5-s time constant) to avoid negative PIP2 transients upon Oxo-M wash (Fig. 7 F; see Fig. S9 in Dickson et al., 2013). Thus, k_4K (agonist) = k_4K_rest + stim_4K * (1 − e−t/τ_onset) with stim_4K = 0.0078/(1 + e4.86-278*agonist) and k_5K (agonist) = k_5K_rest + stim_5K * (1 − e−t/τ_onset) with stim_5K = 0.2 /(1 + e4.86-278*agonist); τ_onset = 1 s. k_4K_recovery = k_4K_rest + stim_4K * e(end−t)/τ_recovery, and k_5K_recovery = k_5K_rest + stim_5K * e(end−t)/τ_recovery with τ_recovery = 5 s.