The EGFR-Inhibitor Erlotinib Induces Differentiation, Cell Cycle Arrest and Apoptosis in EGFR-Negative Cells of MDS and AML.

Background: The EGFR-inhibitor erlotinib was designed to antagonize the deregulated EGFR-activity in solid tumors. After showing that the EGFR-inhibitor erlotinib induces apoptosis in EGFR-negative ex vivo cells of patients with MDS and AML while sparing normal CD34-positive progenitor cells (ASH 2006, abst n°856), we here update these results and identify crucial molecules conveying this off-target effect.

Methods: Ex vivo cells from patients with MDS and AML and myeloid cell lines (P39, KG-1, HL-60) were incubated with 10μM erlotinib, and their potential to differentiate (CD11b and Giemsa staining), to arrest the cell cycle (PI-staining), and to undergo apoptosis (AnnexinV/PI-staining) were tested. Mechanisms of action of erlotinib were studied by immunoflourescence, immunoblotting and siRNAs. To verify the anti-neoplastic effect in vivo, SCID mice were inoculated intraperitoneally with KG-1 cells and erlotinib administered(100mg/kg/day orally,5 days/week, starting on day 7).

Results: Erlotinib overcame the leukemia-associated differentiation block in P39, HL-60 and CD34-positive patient cells as demonstrated by an increased surface expression of CD11b and the induction of morphological differentiation. This effect depended on PDGFRb- and Src-mediated signalling, as shown by the inhibition of erlotinib-induced differentiation upon siRNA knock-down of these molecules. Noteworthy, erlotinib was able to overcome the differentiation block in malignant myeloblasts, which are resistant towards its apoptosis-inducing capacity. Furthermore, erlotinib arrested malignant myeloblasts in the G1 phase of the cell cycle after as early as 24h of incubation. Concomitantly, G1/S cyclins E and D1, as well as phosphorylation of the retinoblastoma protein (serines 807/811, 780 and 795) were reduced. Determining the pathways underlying erlotinib’s ability to induce apoptosis we found that erlotinib disrupted JAK-STAT signaling as demonstrated by an abrogation of constitutive JAK2 (tyrosine 1007/1008) and STAT5 (tyrosine 694) phosphorylation in KG-1 cells. In apoptosis-sensitive KG-1 cells, abrogation of JAK2 expression by siRNA alone was sufficient to diminish activation of STAT-5 and to concomitantly induce apoptosis. Of note, combination of JAK2 knock-down and erlotinib did not cause more apoptosis than erlotinib alone supporting that erlotinib induces apoptosis at least in part by inhibiting JAK2. Assessing more closely the apoptosis-inducing capacity of erlotinib on CD34+ bone marrow cells from AML and MDS patients demonstrates that apoptosis induction (increase of at least 15%) is more often observed in overt AML (7/10 cases) than in high-risk (3/7 cases) and low-risk MDS (2/7 cases). In addition, AML and high-risk MDS samples with a normal karyotype exhibited a higher erlotinib response ex vivo than those with abnormal karyotype. Finally, in vivo efficacy of erlotinib was observed in the SCID mouse model inoculated intraperitoneally with KG-1 cells, where it significantly decreased leukemia development.

Conclusion: we here provide in vitro, ex vivo and in vivo evidence for a potential therapeutic interest of erlotinib in MDS and AML and delineate important mechanisms underlying its off-target effects in EGFR-negative blast cells.