TABLE 4

P-gp–overexpressing cells, P-gp efflux-active concentration, and kinetic parameters for the qualified inhibitors

Cell	Laboratory	Inhibitor	T(0)^{^a}	k_B^{^a}	k_rQ^{^b}	K_QB^{^c}	%CV^{^c}	K_i^{^d}	IC₅₀^/K_i^{^e}
Cell	Laboratory	Inhibitor	M	s⁻¹	s⁻¹	M⁻¹	%CV^{^c}	μM	IC₅₀^/K_i^{^e}
MDCK	2	Carvedilol	1E−03	30	1E+04	1E+05	0.17	0.3	24
	2	Diltiazem	′′	′′	7E+04	1E+04	0.32	2.0	27
	2	Isradipine	′′	′′	6E+04	7E+04	0.42	2.0	21
	2	Mibefradil	′′	′′	8E+03	7E+03	0.15	0.2	32
	2	Nicardipine	′′	′′	7E+03	2E+04	0.28	0.2	13
	2	Quinidine	′′	′′	1E+04	2E+04	0.23	0.3	29
	2	Ranolazine	′′	′′	5E+04	5E+03	0.30	1.0	48
	2	Verapamil	′′	′′	2E+04	3E+03	0.22	0.6	21
	7	Carvedilol	2E−03	27	1E+04	1E+05	0.50	0.3	28
	7	Nicardipine	′′	′′	7E+03	5E+04	0.25	0.2	28
	7	Ranolazine	′′	′′	5E+04	3E+03	0.44	1.0	80
	7	Verapamil	′′	′′	2E+04	3E+03	0.48	0.6	56
Caco-2	6	Carvedilol	3E−04	30	1E+04	2E+05	0.29	0.2	4
	6	Diltiazem	′′	′′	7E+04	2E+04	0.22	1.0	7
	6	Isradipine	′′	′′	6E+04	1E+04	0.12	1.0	7
	6	Nicardipine	′′	′′	7E+03	5E+04	0.26	0.1	9
	6	Quinidine	′′	′′	1E+04	1E+04	0.58	0.2	14
	6	Ranolazine	′′	′′	5E+04	1E+04	0.29	0.8	12
	6	Verapamil	′′	′′	2E+04	2E+04	0.22	0.3	5
	11	Carvedilol	4E−04	15	1E+04	2E+05	0.57	0.2	8
	11	Diltiazem	′′	′′	7E+04	1E+05	0.31	1.0	5
	11	Isradipine	′′	′′	6E+04	1E+05	0.40	1.0	3
	11	Nicardipine	′′	′′	7E+03	5E+04	0.23	0.1	15
	11	Quinidine	′′	′′	1E+04	1E+04	0.31	0.2	14
	11	Ranolazine	′′	′′	5E+04	1E+05	0.29	0.8	18
LLC-PK	2	Mibefradil	1E−03	5	8E+03	5E+05	0.17	0.2	20
	2	Quinidine	′′	′′	1E+04	5E+04	0.40	0.3	55
	2	Ranolazine	′′	′′	5E+04	2E+04	0.16	1.0	39
	2	Verapamil	′′	′′	2E+04	1E+04	0.22	0.6	15

↵^a T(0), the P-gp efflux-active concentration (moles P-gp per liter of apical membrane), and k_B, the basolateral plasma membrane uptake transporter clearance for digoxin (s⁻¹), are inhibitor-independent, i.e., they are fitted simultaneously for all qualified inhibitors for each laboratory.
↵^b k_rQ is the consensus dissociation constant of inhibitor from P-gp from Table 3, and K_QB is the affinity constant of the inhibitor binding to the BT.
↵^c %CV is the coefficient of variation for the fit versus the data expressed as a percentage.
↵^d K_i = (k_rQ/k₁) × K_QPC × 10⁶ (micromolar) is the system-independent dissociation constant of the inhibitor measured relative to the aqueous concentration of the inhibitor in the cytosol. Most of the inhibitors we used do not have a partition coefficient, K_QPC, measured by the technique we used (Tran et al., 2005; Lumen et al., 2013). So we used the value of 350 for all inhibitors, which is the value we measured for quinidine binding to 0.1 μm liposomes composed of a phosphatidylethanolamine/phosphatidylserine/cholesterol (1:1:1) mole ratio (Lumen et al., 2013). This roughly mimics the cytosolic face of the plasma membrane (Lumen et al., 2013). Verapamil had a measured partition coefficient of 650, which would give a K_i roughly half as large as that shown in this table (Lumen et al., 2013). None of the other inhibitors in this work have known partition coefficients measured using this uniform system. k_1Q has been measured for MDCKII-hMDR1-NKI cells for several P-gp substrates, including quinidine and verapamil, and was found to be well fitted as 1 e⁸ M^–1s⁻¹ (Agnani et al., 2011; Lumen et al., 2013). The same value has been assumed for the LLC-PK1-hMDR1-NKI cells. However, for Caco-2 cells, Meng et al. (2017a) found that k_1Q was about 1.7-fold larger. This means that the K_i for an inhibitor with the Caco-2 cells would be 1.7-fold smaller than with MDCKII-hMDR1-NKI cells.
↵^e IC₅₀/K_i is the IC₅₀ divided by the dissociation constant of the inhibitor to P-gp, both in micromolar units, so it is dimensionless.