BCR-ABL-positive acute leukaemias often show limited or transient responses to tyrosine kinase (TK) inhibition with imatinib mesylate (IM). These cells show dysregulation of class I PI3K and mTOR signaling and these pathways have been implicated in leukaemia pathogenesis. Previous PI3K inhibitors have shown significant off-target activity and have not been suitable for in vivo use. We have utilised a novel, highly selective dual inhibitor of class I PI3K and mTOR, PI-103, to assess the functional role of these pathways in BCR-ABL positive cell lines and primary acute leukaemia cells.

In IM-sensitive cell lines, including KCL-22, LAMA-84 and AR230, PI-103 (1μM) reduced proliferation assessed by MTS assay to a mean of 41±5% of control at 48 hours. Combination of low concentrations of IM and PI-103 showed a synergistic response in this assay–IM 0.1μM 94±2, PI-103 0.1μM 84±6 and IM+PI-103 65±6%. IM resistant cell lines (Mahon et al, Blood 96:1070), including KCL-22r, AR230-r and BAF/BCR-ABL-r, were also sensitive to PI-103 with MTS levels of 54±15% of control at 48 hours. The combination of PI-103 and IM was again synergistic–IM 1μM 83±9, PI-103 0.1μM 97±3 and IM+PI-103 51±10% of control. This was sustained at higher concentrations of both agents–IM 5μM 56±17, PI-103 0.25μM 67±12 and IM+PI-103 16±2% of control.

In order to examine the effect of PI3K/mTOR inhibition on cell survival, annexin V flow cytometric assays were carried out. In KCL-22 cells, the combination of IM and PI-103 showed marked synergy: after 72 hours, the % of non-viable cells was 25% in PI-103 (0.5μM), 17% in IM (1μM) and 72% in the combination of IM+PI-103. The IM-resistant counterparts KCL-22-r showed a similar trend but with less marked levels of apoptosis (9%, 0%, and 27% for PI-103, IM and PI-103+IM respectively).

To validate the cell line results in primary cells, blasts from 3 cases of BCR-ABL-positive acute leukaemia were incubated with IM, PI-103 or combinations. Apoptosis assays showed that the level of viable cells after 48 hours incubation was 82±9% in IM (1μM), 67±7% in PI-103 (1μM) and 43±7% in IM+PI-103.

To investigate the interaction between IM and PI-103 further, we examined their effects on cell signaling pathways by immunoblotting. Incubation with IM reduced overall phosphotyrosine levels and tyrosine phosphorylation of the CRKL substrate in a dose-dependent manner. There was readily detectable basal phosphorylation of STAT5, MAPK, Akt and the ribosomal S6 protein, a target of mTOR signaling in all cell lines. Increasing concentrations of IM reduced levels of phosphorylated STAT5 (most sensitive), MAPK and Akt (least sensitive). In several cell lines, including KCL22 and AR230, there were residual levels of phospho-Akt (both S473 and T308), the Akt target FOXO3A and phospho-S6 even at high IM concentrations (>10μM). Addition of the Src-kinase inhibitor SU6656 to IM did not affect residual Akt activity and similar results were obtained in serum-free conditions. PI-103 alone did not affect phosphorylation of STAT5 or MAPK but eliminated Akt and FOXO3A phosphorylation whilst leaving a low level of phospho-S6. The combination of IM and PI-103 abrogated this residual S6 phosphorylation, as well as phosphorylation of all other tested pathways.

These results show that BCR-ABL-positive acute leukaemia cells can have residual PI3K and mTOR activity in the presence of high concentrations of IM. The addition of a dual PI3K/mTOR inhibitor to IM eliminates signaling from these pathways and enhances cell killing and inhibition of proliferation. Combinations of TK and PI3K/mTOR inhibitors merit evaluation in clinical studies.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author