Erlotinib Antagonizes Efflux Via ABC Transporters and Decreases P-Gp Cell Surface Expression by Inhibiting SRC Kinase and mTOR Pathways in Acute Myeloid Leukemia (AML)

Erlotinib (Erlo) was originally developed as an epidermal growth factor receptor inhibitor, yet it also exerts antileukemic “off-target” effects, in vitro and in vivo in MDS and AML (Boehrer et al., Blood, 2008). In a preliminary pre-clinical study, we observed that Erlo increased chemosensitivity to current AML drugs in different AML cell lines and in ex vivo AML patient cells (n=3) (ASH 2010, 2163). Those first results suggested an implication of ABC-transporters in the potentiation of apoptosis. Here, we bring direct evidence for Erlo's ability to hinder efflux pumps and to decrease their expression on AML cells.

Drug efflux via ABC-transporters (substrate: mitoxantrone-MTZ or doxorubicin-Dox), and specific efflux via P-gp (substrates: DioC₂₃ and Rho-123), MRP (s: Calcein and CDCFDA) and BCRP (s: Hoechst 33342) were quantified by FACS following incubation with 10mM Erlo. Intracellular VP-16) content was quantified by Rapid Resolution Liquid Chromatography (RRLC). Biochemical inhibitors of the respective ABC-transporters (CSA (1μM), verapamil (Vera-10μM), MK571 (10μM), KO143 (500nM) served as positive controls. To assess chemosensitivity, 10mM Erlo was combined to AraC (100nM), Dox (100nM), or VP-16 (1mM) and apoptosis over-time (24, 48, 72h) quantified by DioC₃(6)/PI staining. Assessment of sensitivity to the drug combinations listed above were carried out in KG-1 cells, and its more immature variant KG-1a and in ex vivo CD34+ marrow cells from AML patients (AML post MDS n=5, de novo AML n=5). P-gp's ATPase activity was quantified with the luminescence-based Pgp-Gloä Assay System. Surface expression of P-gp was determined by FACS analysis and total protein expression of MRP, BCRP and P-gp by immunoblot analysis. Functional relevance of signaling pathways was tested using the SRC inhibitor PP2 (10μM) and the mTOR inhibitor Rapamicin (10nM).

We found that I) Erlo inhibited efflux via P-gp, MRP and BCRP as demonstrated by increased intracellular retention of DioC₂³/Rho-123, Calcein/CDCFDA and Hoechst 33342, respectively, andby its ability to retain MTX (300nM) and Dox (200nM) intracellularly II) Inhibition of drug efflux was higher in KG-1 than in KG-1a cellss, in agreement with a lower expression of P-gp and BCRP on KG-1a as compared to KG-1 cells; III) Quantification of VP-16 by RRLC after incubation with or without Erlo showed the ability of Erlo to increase intracellular VP-16 contents by approximately 60%; IV) Erlo increased ATPase activity in a dose-dependant manner, supporting the notion that Erlo is a competitive inhibitor of P-gp; IV) Erlo combined to VP-16 induced synergistic effects on apoptosis in KG-1 cells, and to a lesser extent in KG-1a (48h KG-1: Erlo 20%, VP-16 38%, Erlo+VP16 78%, KG-1a 48h: Erlo 10%, VP-16: 12%, Erlo+VP16: 35%); V) 48h of incubation with Erlo reduced cell surface expression of P-gp in KG-1 cells by 50%, whereas total P-gp protein expression remained unchanged, suggesting that Erlo interferes exclusively with the protein form expressed on the cell surface, VI) Decrease of P-gp cell surface expression was recapitulated upon incubation with PP2 (10μM) or Rapamicin (10nM); VII) the combination of Erlo+VP-16 in 10 AML-patient samples induced synergistic effects on apoptosis in 5 of them and additive effects in 3 of them.

We here confirm that Erlo increases sensitivity towards chemotherapeutic agents subjected to drug efflux via ABC-transporters and delineate the molecular pathways conveying these effects.

Fenaux:Celgene: Honoraria, Research Funding.