Inhibition of mTOR Kinase Pathway Selectively Sensitizes Acute Myeloid Leukemia Cells to Cytarabine-Induced Apoptosis.

Background: Rapamycin (RAPA) is an inhibitor of mTOR kinase pathway. In vitro low doses of this agent induce cell cycle arrest in G₁ phase, whereas higher concentrations of RAPA exert proapoptotic effects.

Aim: We assessed cytotoxicity of RAPA alone or in combination with cytarabine (cytosine arabinoside, ARA-C) in acute myeloblastic leukemia (AML) cells and in normal lymphocytes obtained from 10 healthy volunteers.

Methods: AML cells (in vitro HL-60 cell line and ex vivo leukemic cells) and phytohemaglutynin (PHA)-stimulated normal lymphocytes were treated for 24 – 48 h with 1 ng/ml RAPA alone or in combination with 50 nM cytarabine (Ara-C). Moreover, cells was pre-incubated with RAPA for 24 h and then Ara-C was added for the next 24 h. Untreated cultures and those treated with RAPA, Ara-C or PHA alone served as respective controls. The proapoptotic effect was assessed by Annexin V assay and presented as a percentage of Annexin-V-positive cells (apoptotic index; AI). Cell cycle was analyzed by DNA distribution in propydium iodide/RN-ase stained cells. Cyclin D3, A and E expression was also measured using flow cytometry.

Results: Median AI induced in HL-60 cells after 24 h treatment with RAPA+Ara-C (30.1%) was significantly higher than induced by RAPA (7.2%) or Ara-C (18.5%) alone (p=0.002 and p=0.03, respectively). The RAPA+Ara-C combination exerted additive effect (combination index 0.87) in that model. Additional 24 hour pretreatment with RAPA further increased apoptosis (median AI 41.5%, vs. 10.9% after 48 h-RAPA alone). In contrast to leukemic cells, pretreatment of normal PHA-stimulated lymphocytes with RAPA caused their G₁ phase cell cycle arrest, with significant decrease in cyclin D3 expression (vs. untreated cells - p<0.001). This resulted in prevention of Ara-C-induced cytotoxicity in healthy lymphocytes, when Ara-C was added for another 24 h. Importantly, that protective effect was reversible when RAPA-treated lymphocytes were rinsed and then cultured in fresh, RAPA-free medium for the next 24 h. In another set of experiments, cells from 12 de novo AML patients were treated with RAPA and Ara-C in above concentrations and time settings. RAPA and Ara-C were administrated to isolated peripheral blood mononuclear cells (PBMC). PBMC were immunophenotyped before and after treatment. Leukemic blasts were marked for individually chosen antigen, most characteristic for leukemic clone in particular patient. Normal CD3+ lymphocytes were also detected. Finally, Annexin V staining was performed. Based on that simultaneous three-color staining the proapoptotic effects of treatment could be measured by flow cytometry in both leukemic blasts and normal CD3+ cells. Thus, we found that pretreatment with RAPA protected majority of CD3+ cells (median of alive cells 85.5%) from Ara-C-induced apoptosis, whereas the leukemic blasts AI was higher than in samples treated with Ara-C. After Ara-C alone CD3+ rate decreased significantly (median 35.1%).

Conclusions: Pretreatment with RAPA enhances cytotoxic effect of Ara-C on leukemic cells, but not on healthy lymphocytes. The phenomenon is probably due to reversible arrest of healthy cells in G₁ phase of cell cycle by low doses of RAPA, what causes their transient resistance to proapoptotic action of cytostatic drugs. In contrast, the same RAPA doses selectively sensitizes leukemic cells to cytostatics. This suggests, that inhibition of mTOR kinase prior to cytostatics administration may result in selective anti-tumor treatment, with protection of normal cells.