Functional consequence of MDR1 expression on imatinib intracellular concentrations

In their recent article, Mahon et al¹  demonstrate that MDR1 gene overexpression can confer resistance to imatinib mesylate (Gleevec; Novartis, Basel, Switzerland) in leukemia cell lines. Several other cellular mechanisms of resistance to imatinib have also been identified,^2-4  but the possible importance of drug-transporter proteins has been only recently appreciated with the demonstration that imatinib was a substrate of P-glycoprotein (Pgp), the gene product of MDR1.⁵  The generally accepted action of MDR1 is to reduce intracellular drug accumulation through Pgp-mediated efflux, thus hampering the achievement of effective drug levels at the target site. Mahon et al¹  show that MDR1 gene overexpression results in resistance to imatinib, but they did not assess the functional consequence of MDR1 expression on imatinib intracellular availability. As a complement to their work, we report here the first direct evidence of the marked impact of Pgp expression on imatinib intracellular concentrations.

Homologous MDR1⁺ and MDR1^- LLC-PK1 cells (ie, transfected porcine kidney epithelial cells, a recognized model for assessing drug efflux transporter activity⁶ ) were obtained from Schinkel et al.⁷  Cells were incubated for 18 hours at 37°C, in growth medium containing 10% fetal calf serum in the presence of imatinib at the clinically relevant concentrations of 500, 1500, and 3000 ng/mL. After incubation, the excess supernatant medium was discarded, and adherent cell cultures were washed 3 times with ice-cold phosphate buffer before extraction with methanol/water 60:40 on a planar shaker by gentle agitation at 100 rpm for 3 hours. The methanolic extracts were collected, centrifuged, evaporated to dryness under nitrogen, and reconstituted in methanol/water 60:40 prior to analysis using a validated high-performance liquid chromatography ultraviolet method. Imatinib concentrations were normalized on total protein measurements (bicinchonic acid assay), which were used as a surrogate of cell counts. Incubation experiments were done in duplicate. The results in Table 1 highlight the striking differences (up to 24-fold) in intracellular concentration of imatinib according to the cells' ability to express Pgp.

Our experiment provides direct evidence of the influence of Pgp on imatinib intracellular availability. Imatinib is very efficiently expelled from MDR1⁺ cells at all tested concentrations. Intracellular concentrations and transmembrane ratios are significantly affected by genotype (P < .0001) and applied concentration (P < .0001, 2-way ANOVA). In MDR1⁺ cells, the intracellular/extracellular concentration ratio remains constant, indicating that imatinib Pgp-mediated efflux is never saturated at all tested concentrations. By contrast, the higher ratio observed at low imatinib level in MDR1^- cells increases further at higher levels, suggesting the existence of subsidiary, less efficient and saturable transport mechanisms.

This experiment formally confirms that imatinib interaction with Pgp⁵  has important functional consequences and that imatinib efflux from cancer cells by the drug transporter Pgp should be seriously considered among the mechanisms of imatinib resistance, supporting the observations of Mahon et al.¹ 

The clinical activity of imatinib may thus be significantly hampered in Pgp-expressing cells; this seems to be particularly the case of chronic myeloid leukemia (CML) cells in blastic phase.^8,9  These findings should stimulate further research to evaluate the addition of Pgp inhibitors to imatinib for the treatment of CML blastic crisis.

We are indebted to Alfred H. Schinkel, The Netherlands Cancer Institute, for providing access to cell lines. Hugues Henry, Laboratory of Clinical Chemistry, University Hospital, Lausanne, is acknowledged for helpful technical support. Imatinib mesylate is a kind gift of Novartis (Switzerland).