An activating mutation of GNB1 is associated with resistance to tyrosine kinase inhibitors in ETV6-ABL1-positive leukemia

Abstract

Leukemias harboring the ETV6-ABL1 fusion represent a rare subset of hematological malignancies with unfavorable outcomes. The constitutively active chimeric Etv6-Abl1 tyrosine kinase can be specifically inhibited by tyrosine kinase inhibitors (TKIs). Although TKIs represent an important therapeutic tool, so far, the mechanism underlying the potential TKI resistance in ETV6-ABL1-positive malignancies has not been studied in detail. To address this issue, we established a TKI-resistant ETV6-ABL1-positive leukemic cell line through long-term exposure to imatinib. ETV6-ABL1-dependent mechanisms (including fusion gene/protein mutation, amplification, enhanced expression or phosphorylation) and increased TKI efflux were excluded as potential causes of resistance. We showed that TKI effectively inhibited the Etv6-Abl1 kinase activity in resistant cells, and using short hairpin RNA (shRNA)-mediated silencing, we confirmed that the resistant cells became independent from the ETV6-ABL1 oncogene. Through analysis of the genomic and proteomic profiles of resistant cells, we identified an acquired mutation in the GNB1 gene, K89M, as the most likely cause of the resistance. We showed that cells harboring mutated GNB1 were capable of restoring signaling through the phosphoinositide-3-kinase (PI3K)/Akt/mTOR and mitogen-activated protein kinase (MAPK) pathways, whose activation is inhibited by TKI. This alternative GNB1^K89M-mediated pro-survival signaling rendered ETV6-ABL1-positive leukemic cells resistant to TKI therapy. The mechanism of TKI resistance is independent of the targeted chimeric kinase and thus is potentially relevant not only to ETV6-ABL1-positive leukemias but also to a wider spectrum of malignancies treated by kinase inhibitors.

Figure 1

Sensitivity of TKI-sensitive and TKI-resistant cell lines to selected TKIs. Sensitivity to TKIs determined by MTS assay (a). Viability of TKI-treated cells normalized to untreated cells is shown on y axis. Error bars represent the standard deviation from the mean. Effect of second dose of imatinib (0.5, 5.0 and 15.0 μM) on the viability of TKI-sensitive and TKI-resistant cell lines (b). The percentage of viable (Annexin-V-negative, PI-negative) cells is shown on y axis. Experiments were performed in triplicates (a,b).

Figure 2

TKI-resistant cells gained independence from the ETV6-ABL1 oncogene. Effect of imatinib treatment on phosphorylation of Etv6-Abl1 kinase and its downstream target Crkl was analyzed via western blot (a). TKI-sensitive (S) and TKI-resistant (R) cells were grown in the absence/presence of 0.5, 1.5 or 5.0 μM imatinib for 30 min and 24 h. Level of phosphorylated (p) Etv6-Abl1 and Crkl was determined in nuclear protein and whole cell lysates, respectively. Tbp and B-actin were used as loading controls. Scans were cut and reorganized to show timepoints followed within the experiment in left to right order. Original scans are shown in Supplementary Figure 3. Effect of shRNA-mediated ETV6-ABL1 silencing on proliferation of TKI-sensitive and TKI-resistant cells (b). Dashed and dotted lines indicate proliferation of cells transduced with ETV6-ABL1-targeting and ETV6-targeting shRNAs. Cells transduced with negative control shRNA are represented by solid line. Experiments with two different shRNA represent an independent replicate. Results of additional independent experiment with the same two shRNAs are shown in Supplementary Figure 4.

Figure 3

2-DE based proteomic analysis of TKI-sensitive, TKI-sensitive-LTC and TKI-resistant cell line. A representative 2-DE gel with 975 visualized proteins spots (at the top) with close-ups of the area with a differential GNB1 spot (below). Close-ups of scanned 2-DE gels—black arrow indicates the position of a differential protein spot (identified by mass spectrometry as Gβ), which was detected exclusively in TKI-resistant cells. A representative triplicate of experimental hexaplicate.

Figure 4

Effect of imatinib treatment on the activation of the PI3K/Akt/mTOR and MAPK pathways in TKI-sensitive and TKI-resistant cells. TKI-sensitive and TKI-resistant ETV6-ABL1-positive cells were cultured in the absence/presence of 5.0 μM imatinib for 30 min or 24 h. Levels of total and phosphorylated (p) Erk, p70S6K, S6 and Akt proteins in whole-cell lysates were determined using western blot. Gapdh was used as a loading control. Representative data of three independent experiments. Results of the two additional independent experiments are shown in Supplementary Figure 6.

Figure 5

GNB1^K89M/GNB1^K89E mutations confer multi-TKI resistance in the TKI-sensitive ETV6-ABL1-positive cell line. TKI-sensitive ETV6-ABL1-positive cells were transduced with GNB1^wt/GNB1^K89M/GNB1^K89E and empty vector. Sensitivity of transduced cells to TKIs was determined via MTS assay (a). Tested concentrations of TKIs are shown on x axis. Viability of TKI-treated cells normalized to untreated cells is shown on y axis. Experiments were performed in triplicates. Error bars represent the standard deviation from the mean. Proliferation of transduced cells upon repeated treatment with 1.5 μM imatinib (b). Black arrows indicate imatinib administration. Absolute number of cells is shown on y axis.

Figure 6

Effect of the GNB1^K89M/GNB1^K89E mutations on the activation of the PI3K/Akt/mTOR and MAPK pathways. TKI-sensitive ETV6-ABL1-positive cells were transduced with GNB1^wt/GNB1^K89M/GNB1^K89E and empty vector. Transduced cells were treated with 5.0 μM imatinib for 30 min and 24 h. Levels of total and phosphorylated (p) Erk, p70S6K, S6 and Akt proteins in whole-cell lysates were determined using western blot. Gapdh was used as a loading control. Representative data of three independent experiments. Scans were cut and reorganized to show timepoints followed within the experiment in left to right order. Original scans are shown in Supplementary Figure 8. Results of two additional independent experiments are shown in Supplementary Figures 9 and 10.

Figure 7

Sensitivity to dual PI3K/mTOR inhibitor BEZ235. Sensitivity of TKI-sensitive and TKI-resistant cells (a) and TKI-sensitive cells transduced with GNB1^WT and GNB1^K89M (b) to BEZ235. Sensitivity was determined by MTS assay. Viability of treated cells normalized to untreated cells is shown on y axis. Experiments were performed in triplicates. Error bars represent the standard deviation from the mean.