Combination of the ABL kinase inhibitor imatinib with the Janus kinase 2 inhibitor TG101348 for targeting residual BCR-ABL-positive cells

Abstract

Background: The ABL kinase inhibitor imatinib is highly effective in treating most, but not all, patients with chronic myeloid leukemia (CML). This is because residual CML cells are generally present in the bone marrow microenvironment and are refractory to imatinib. Hematopoietic cytokine receptor signaling is mediated by Janus kinases (JAKs) and their downstream transcription factor, signal transducer and activator of transcription (STAT). TG101348 (SAR302503) is an oral inhibitor of JAK2.

Methods: We investigated the efficacy of imatinib and TG101348 using the break point cluster region-c-Abelson (BCR-ABL)-positive cell line and primary CML samples wherein leukemia cells were protected by a feeder cell line (HS-5).

Results: Imatinib treatment resulted in partial inhibition of cell growth in HS-5-conditioned medium. Furthermore, combined treatment with imatinib and TG101348 abrogated the protective effects of HS-5-conditioned medium on K562 cells. Phosphorylation of Crk-L, a BCR-ABL substrate, decreased considerably, while apoptosis increased. In addition, the combined treatment of CD34-positive primary samples resulted in considerably increased cytotoxicity, decreased Crk-L phosphorylation, and increased apoptosis. We also investigated TG101348 activity against feeder cells and observed that STAT5 phosphorylation, granulocyte macrophage colony-stimulating factor, and interleukin 6 levels decreased, indicating reduced cytokine production in HS-5 cells treated with TG101348.

Conclusions: These results showed that JAK inhibitors may enhance the cytotoxic effect of imatinib against residual CML cells and that a combined approach may be a powerful strategy against the stroma-associated drug resistance of Philadelphia chromosome-positive cells.

Figure 1

Cytotoxic effects of imatinib in the presence of HS-5 conditioned medium. (A) K562 cells were cultured at a concentration of 8 × 10⁴/mL in the presence of varying concentrations of imatinib in the presence or absence of HS-5 conditioned medium for 72 h. Viable cell numbers were calculated. Results are representative of three separate experiments. (B) K562 cells were treated with 2 μM imatinib alone or in combination with either 1 μM TG101348 or 5 μM AG490 in the presence or absence of HS-5 conditioned medium for 72 h. Viable cell numbers were calculated. Results are representative of three separate experiments. (C) K562 cells were cultured with the indicated concentrations of TG101348 for 72 h. Viable cell numbers were calculated. Results are representative of three separate experiments. (D) K562 cells were treated with TG101348 for 24 h, and total extracts were examined by immunoblotting using anti-phospho ABL, ABL, phospho-STAT5, STAT5, and β-actin Abs. Blots were scanned and optical densities were determined using ImageJ software.

Figure 2

Co-treatment with imatinib and TG101348 increased inhibition of cell growth and induced apoptosis of BCR-ABL-expressing cells in the presence of HS-5 cells. (A) K562 cells were cultured at a concentration of 8 × 10⁴/mL in the presence or absence of HS-5 cells. Cells were then treated with varying concentrations of imatinib alone or in combination with TG101348 for 72 h. Viable cell numbers were calculated. Results are representative of three separate experiments. (B) K562 cells were treated with imatinib alone or in combination with TG101348 for 24 h, and total extracts were examined by immunoblotting using anti-phospho ABL, ABL, phospho-Crk-L, Crk-L, phospho-STAT5, STAT5, phospho-MAPK, Erk-1, cleaved PARP, and β-actin Abs. Blots were scanned and optical densities were determined using ImageJ software. (C) K562 cells were cultured with Z-VAD-fmk (20 μM) and/or TG101348 and imatinib for 24 h. Total extracts were analyzed by immunoblotting using anti-cleaved caspase-3 and cleaved PARP Abs. Actin was used as the loading control. Data are representative of two separate experiments. Blots were scanned and optical densities were determined using ImageJ software.

Figure 3

Effects of imatinib and TG101348 on CD34-positive primary cells. (A) CD34-positive CML cells or healthy donor samples were cultured in the presence of HS-5 cells. Cells were then treated with varying concentrations of imatinib alone or in combination with TG101348 for 72 h. Viable cell numbers were calculated. Results are representative of three separate experiments. (B) CD34-positive primary cells were treated with the indicated concentrations of imatinib alone or in combination with TG101348 for 24 h. Total cell extracts were analyzed by immunoblotting using phospho-specific ABL, STAT5, Crk-L, and cleaved PARP Abs. β-actin was used as the loading control. Results are representative of at least three reproducible experiments. Blots were scanned and optical densities were determined using ImageJ software. (C)In vivo studies were performed as described in Methods. Tumors were treated with or without imatinib and TG101348 5 days per week, and growth was measured. *p < 0.05, compared with control mice.

Figure 4

JAK2 siRNA transfected K562 cells had increased sensitivity to imatinib. (A) K562 cells were transfected with JAK2 siRNA or a control siRNA. Cell lysates were analyzed by immunoblotting 72 h after transfection using Jak2 and β-actin Abs. Blots were scanned and optical densities were determined using the ImageJ system. (B) K562 and siRNA-transfected K562 cells were cultured at a concentration of 8 × 10⁴/mL in the presence or absence of feeder cells. The cells were treated with 1 μM imatinib for 72 h. Viable cell numbers were calculated. Results are representative of three separate experiments. (C) K562 and siRNA-transfected K562 cells were treated with imatinib for 24 h, and total extracts were examined by immunoblotting using anti-phospho ABL, ABL, phospho-Crk-L, Crk-L, phospho-MAPK, Erk-1 and β-actin Abs. Blots were scanned and optical densities were determined using ImageJ software.

Figure 5

Effect of TG101348 on HS-5 cells. (A) HS-5 cells were cultured in the presence of different concentrations of TG101348 for 24 h, and total extracts were examined by immunoblotting using anti-phospho STAT5, STAT5, and β-actin Abs. Blots were scanned and optical densities were determined using ImageJ software. (B) HS-5 cells were treated with TG101348 for 24 h and conditioned medium was collected. The relative levels of 36 different cytokines were analyzed using the Human Cytokine Array Kit. The expression of cytokines was determined using ImageJ software. Results are representative of three separate experiments. *p < 0.05, TG101348 treatment vs. control.