In situ biochemical demonstration that P-glycoprotein is a drug efflux pump with broad specificity

Abstract

While P-glycoprotein (Pgp) is the most studied protein involved in resistance to anti-cancer drugs, its mechanism of action is still under debate. Studies of Pgp have used cell lines selected with chemotherapeutics which may have developed many mechanisms of resistance. To eliminate the confounding effects of drug selection on understanding the action of Pgp, we studied cells transiently transfected with a Pgp-green fluorescent protein (GFP) fusion protein. This method generated a mixed population of unselected cells with a wide range of Pgp-GFP expression levels and allowed simultaneous measurements of Pgp level and drug accumulation in living cells. The results showed that Pgp-GFP expression was inversely related to the accumulation of chemotherapeutic drugs. The reduction in drug concentration was reversed by agents that block multiple drug resistance (MDR) and by the UIC2 anti-Pgp antibody. Quantitative analysis revealed an inverse linear relationship between the fluorescence of Pgp-GFP and MDR dyes. This suggests that Pgp levels alone limit drug accumulation by active efflux; cooperativity between enzyme, substrate, or inhibitor molecules is not required. Additionally, Pgp-GFP expression did not change cellular pH. Our study demonstrates the value of using GFP fusion proteins for quantitative biochemistry in living cells.

Figure 1

Effect of PgpGFP expression on the accumulation of fluorescent dyes. (A) PgpGFP-transfected HeLa cells were incubated for 30 min with 1 μM daunorubicin without (top) or with (bottom) 40 μg/ml verapamil. GFP and daunorubicin fluorescence were imaged using a confocal microscope. The fluorescence profile along the line drawn in the merge image was quantified (right). (B) Cells were incubated for 30 min with 200 nM TMRE and 200 nM Ho342, and the fluorescence images of GFP, TMRE, and Ho342 were acquired using an epifluorescence microscope. (C) Cells were incubated for 30 min with 5 μM Fura Red AM and GFP and Fura Red fluorescence were imaged on a confocal microscope. Bars, 20 μm.

Figure 2

Effect of PgpGFP expression on vincristine-mediated microtubule depolymerization. PgpGFP-transfected HeLa cells were incubated with 2 μM or 80 nM vincristine for 30 min and stained with a Cy3-conjugated anti-β-tubulin antibody. Arrows indicate PgpGFP expressing cells and arrowheads indicate nonexpressing cells. Bars, 20 μm.

Figure 3

FACS^® analysis of TMRE accumulation as a function of PgpGFP expression. (A) Histogram of GFP fluorescence of HeLa cells transfected with PgpGFP. Approximately 50% of the cells were nonexpressing showing a single peak of ∼5 × 10⁰ units. Transfected cells showed a >100-fold range of GFP fluorescence. (B) Dot plot of transfected cells incubated for 30 min with 50 nM TMRE. The solid diagonal line of slope −1 was fit to the data. To estimate the average TMRE fluorescence exhibited by cells with GFP fluorescence of 10³ units, a dashed line was drawn from the center of cell density at 10³ down and across to the x and y axes. At all levels of PgpGFP expression, there were some cells that showed TMRE fluorescence of 10² on the y-axis (marked as a horizontal oval). Coincubation with propidium iodide selectively increased the fluorescence of these cells, indicating that they were most likely dead (data not shown). (C–G) Dot plot of cells coincubated with the indicated verapamil concentration. The solid diagonal line was reproduced from B for reference. The average TMRE fluorescence exhibited by cells with GFP fluorescence of 10³ units was estimated as in B. (H) The ratio of drug in/drug out was modeled is described in text. Plots of D_in/D_out as a function of number of Pgp per cell are shown for five different catalytic constants of 10, 1, 0.1, 0.01, and 0.001 drug molecules per pump per second. (I) Plot of average TMRE fluorescence of cells exhibiting GFP fluorescence of 10³ as a function of verapamil concentration from the dash lines drawn in B–G. Open circle indicates no verapamil control (not 1 μm).

Figure 4

Effect of PgpGFP expression on cellular pH. PgpGFP-transfected HeLa cells were loaded with SNARF-1 by a 30-min incubation with 1 μM SNARF-1 AM. (A–F) SNARF-1 was calibrated by measuring the fluorescence cells incubated in calibration solution at pH 6.75, 7.25, and 7.75 (shown), and 7.0, 7.5, and 8.0 (not shown). The GFP image identifies PgpGFP expressing cells (left). The ratio of the SNARF-1 base/acid fluorescence are shown as a pseudocolor intensity image (right). The scale on the right of the lookup color bar represents the average SNARF-1 ratio calibrated at the indicated pH values. (G–H) The GFP and SNARF-1 fluorescent ratio images of cells in media. Bar, 20 μM.