eIF4B is a convergent target and critical effector of oncogenic Pim and PI3K/Akt/mTOR signaling pathways in Abl transformants

Abstract

Activation of eIF4B correlates with Abl-mediated cellular transformation, but the precise mechanisms are largely unknown. Here we show that eIF4B is a convergent substrate of JAK/STAT/Pim and PI3K/Akt/mTOR pathways in Abl transformants. Both pathways phosphorylated eIF4B in Abl-transformed cells, and such redundant regulation was responsible for the limited effect of single inhibitor on Abl oncogenicity. Persistent inhibition of one signaling pathway induced the activation of the other pathway and thereby restored the phosphorylation levels of eIF4B. Simultaneous inhibition of the two pathways impaired eIF4B phosphorylation more effectively, and synergistically induced apoptosis in Abl transformed cells and inhibited the growth of engrafted tumors in nude mice. Similarly, the survival of Abl transformants exhibited a higher sensitivity to the pharmacological inhibition, when combined with the shRNA-based silence of the other pathway. Interestingly, such synergy was dependent on the phosphorylation status of eIF4B on Ser422, as overexpression of eIF4B phosphomimetic mutant S422E in the transformants greatly attenuated the synergistic effects of these inhibitors on Abl oncogenicity. In contrast, eIF4B knockdown sensitized Abl transformants to undergo apoptosis induced by the combined blockage. Collectively, the results indicate that eIF4B integrates the signals from Pim and PI3K/Akt/mTOR pathways in Abl-expressing leukemic cells, and is a promising therapeutic target for such cancers.

Keywords: AKT; Abl oncogene; Pim kinase; eIF4B; tumorigenesis.

Figure 1. eIF4B phosphorylation is regulated by PI3K/Akt/mTOR pathway in Abl transformants

A. Bcr-Abl transformed cells (K562) were treated with LY294002 (10 μM) for indicated time. Whole cell lysates were prepared and examined for eIF4B Ser422 or Ser406 phosphorylation levels by Western blotting. B. v-Abl transformed cells (NS2 and W44) were treated with LY294002 (5 μM) for indicated time. The eIF4B phosphorylation was analyzed as described in A. C. eIF4B Ser422 phosphorylation levels in Supplementary Figure 1B and 1C were quantitated by densitometry and normalized to total protein levels. The eIF4B S422 phosphorylation levels of Abl transformants expressing sh-luciferase were set to 100%. Plotted are results from three independent experiments. Error bars represent SEM, n = 3 (**P < 0.01). D. and E. Bcr-Abl⁺ cells (D) or v-Abl⁺ cells (E) were treated with 4 μM (D) or 1.5 μM (E) Akti-1/2 for indicated time. Analysis of eIF4B phosphorylation were performed as described in A. F and G. K562 (F) or W44 (G) cells expressing luciferase-specific shRNA or Akt1-specific shRNAs were analyzed by Western blotting with indicated antibodies. H. eIF4B Ser422 phosphorylation levels in Supplementary Figure 1H and 1I were quantitated by densitometry and normalized to total protein levels. The levels of eIF4B S422 phosphorylation were set to 100% at 0 hour. Plotted are results from three independent experiments. Error bars represent SEM, n = 3 (*P < 0.05, **P < 0.01). I. eIF4B-WT or S422A mutant was co-transfected with empty vector, Akt1-WT, or Myr-Akt1 in 293T cells. Total proteins were extracted and analyzed for eIF4B S422 phosphorylation by Western blotting. J. K562 cells overexpressing Akt1-WT, Akt1-E17K, or control were treated with or without rapamycin (10 μM) for 4h and analyzed as described in Figure 1I with indicated antibodies.

Figure 2. Long time inhibition of one signaling activates the other pathway and restores eIF4B Ser422 phosphorylation

A. K562 cells were treated with 5 μM SMI-4a for indicated time. Examination of eIF4B and Akt phosphorylation was performed with indicated antibodies. B. NS2 and W44 cells were treated with 2 μM SMI-4a for indicated time. The phosphorylation of eIF4B Ser422 and Akt was analyzed as described in A. C. and D. eIF4B phosphorylation levels in A and B were quantitated by densitometry and normalized to total protein levels. The levels of eIF4B S422 phosphorylation are 100% at 0 h. Plotted are results from three independent experiments. Error bars represent SEM, n = 3 (*P < 0.05, **P < 0.01). E. and F. K562, NS2, or W44 cells were treated with 5 μM or 2 μM LY294002 for indicated time. The cell lysates were analyzed by Western blotting with indicated antibodies. G. and H. eIF4B phosphorylation levels in E and F were quantitated as described in C. I. experiments were performed as described in E. At 40 h, K562 cells treated with LY294002 were collected, washed with PBS and incubated with mixture of 5 μM LY294002 and 5 μM SMI-4a or with mixture of 5 μM LY294002 and vehicle for 8 h. Cells were analyzed for phosphorylation of eIF4B Ser422 by Western blotting.

Figure 3. Combined inhibition of Pim and PI3K/Akt/mTOR signaling impairs eIF4B phosphorylation more effectively than suppressing one pathway alone

A. K562 cells were treated with 12.5 μM SMI-4a or 10 μM LY294002 alone, or in combination for indicated time. Cell lysates were prepared and immunoblotted with indicated antibodies. B. and C. K562 cells were treated with 12.5 μM SMI-4a or 4 μM Akti-1/2 alone, or in combination (B); or 12.5 μM SMI-4a or 5 μM rapamycin alone, or in combination (C) for indicated time. Experiments were performed as described in A. D–F. NS2 cells were treated with/without 6.25 μM SMI-4a ± 5 μM LY294002 (D), 1.5 μM Akti-1/2 (E), or 2 μM rapamycin (F), for indicated time. Cell lysates were examined for eIF4B phosphorylation as described in A. G–I. experiments on W44 cells were performed as described in D. W44 cells were treated with/without 6.25 μM SMI-4a ± 5 μM LY294002 (G), 1.5 μM Akti-1/2 (H), or 2 μM rapamycin (I), for indicated time. Cell lysates were examined for levels of phospho-eIF4B (Ser422).

Figure 4. Combined inhibition of Pim and PI3K/Akt/mTOR pathways induces synergistic apoptosis

A. K562 cells were treated with SMI-4a or LY294002 alone, or in combination at indicated concentrations for 48 h. Cells were harvested for Annexin V/PI staining of apoptotic cells. Values represent means ± SEM, n = 3 (**P < 0.01). B. The apoptosis of K562 cells in A were analyzed by Chou and Talalay method. Fractional effect denotes decrease of cell viability relative to control for apoptosis assay. Combination index (CI) value < 1 represents synergistic combination. C. experiments were performed as described in A. K562 cells were treated with SMI-4a or Akti-1/2 alone, or in combination at indicated concentrations for 48 h. Then cells were harvested for apoptosis analysis. D. the synergism of cell apoptosis in C was determined as described in B. E–H. NS2 cells were treated with SMI-4a or LY294002 alone, or in combination (E); with SMI-4a or Akti-1/2 alone, or in combination (G), at indicated concentrations for 24 h; cells were analyzed for apoptosis as described in A and CI values were calculated as described in B (F and H).

Figure 5. Silencing one pathway confers Abl transformants more sensitive to inhibition of the other pathway

A. K562 cells stably expressing luciferase-specific shRNA or PDK1-specific shRNAs were treated with 15 μM SMI-4a for indicated time and immunoblotted for eIF4B phosphorylation. B. K562 cells in A were treated by indicated concentrations of SMI-4a for 48 h. Cells were harvested and apoptosis was displayed as percent viability using Annexin V staining followed by FACS. Values represent means ± SEM, n = 3 (*P < 0.05, **P < 0.01). C–F. K562 cells (C) or W44 cells (E) ectopically expressing shRNA targeting luciferase or Akt1 were treated with 15 μM SMI-4a (K562) or 8 μM SMI-4a (W44) in a time course, and analyzed by Western blotting as described in A. Cells were treated with indicated concentrations of SMI-4a for 48 h (K562) or 24 h (W44), and apoptosis of K562 (D) or W44 (F) was evaluated by FACS. G. and H. S6K1 knockdown or control K562 cells were incubated with 15 μM SMI-4a for indicated time and analyzed as described in A and B. I–L. K562 cells stably expressing luciferase-specific shRNA or Pim-1-specific shRNAs were treated with 10 μM Akti-1/2 (I) or 10 μM rapamyicn (K) for indicated time, and immunoblotted for eIF4B phosphorylation. Cells were incubated with Akti-1/2 for 48h (J) or rapamyicn for 36h (L) at indicated concentrations. Cell apoptosis was examined as described in B.

Figure 6. Combined inhibition of Pim and PI3K/Akt/mTOR pathways suppresses the growth of K562 tumor engrafted in nude mice in a synergistic manner

A. nude mice carrying K562 xenografts were treated with SMI-4a (50 mg/kg) alone, Akti-1/2 (80 mg/kg) alone, their combination, or vehicle control every other day. Tumors were excised at 8 h after the last treatment. Shown are representatives from six independent experiments. B. plotted are relative volumes of tumors in A. The average volume of vehicle control was set as 100%. Error bars, SEM, n = 24 (**P < 0.01). C. eIF4B Ser422 phosphorylation in representative tumors in A was examined by immunoblotting. D. levels of phosphorylated eIF4B in C were quantitated by densitometry, and normalized to total eIF4B levels. The levels of Ser422 phosphorylation of Mock group were 100%. Plotted are results from three independent experiments. Error bars represent SEM, n = 3 (**P < 0.01). E. nude mice bearing GFP-expressing K562 xenografts were treated with SMI-4a (50 mg/kg) alone, rapamycin (10 mg/kg) alone, combined treatment, or vehicle control every other day. Tumors were excised at 8 hours after the last treatment. Shown are representatives from five independent experiments. F. shown are relative volumes of tumors in E. The average volume of vehicle control was set as 100%. Error bars, SEM, n = 20 (**P < 0.01). G. over a 21-day period after xenograft inoculation, tumors under indicated treatments were measured by bioluminescent imaging. Shown are representative images from at least three independent experiments with similar results. H. tumors in E were examined by immunoblotting with indicated antibodies. I. levels of phosphorylated eIF4B in H were quantitated by densitometry, and normalized to total eIF4B levels as described in D.

Figure 7. Altered phosphorylation of eIF4B on Ser422 is responsible for synergic inhibition of cell survival by combined blockage of Pim and PI3K/Akt/mTOR pathways

A. interference efficiency of shRNA-based knockdown of eIF4B in K562 cells was determined by Western blotting. shRNA targeting luciferase (sh-luc) served as a control. B. protein levels of eIF4B in A were quantitated by densitometry and normalized to β-actin. Plotted are the average levels from three independent experiments. The error bars represent the SEM (**P < 0.01). C. K562 cells in A were treated with SMI-4a alone, rapamycin alone, or their combination at indicated concentrations. After 36 h, the cells were assessed for apoptotic fraction. Values represent means ± SEM, n = 3 (*P < 0.05, **P < 0.01). D. K562 cells ectopically expressing empty vector, eIF4B-WT or eIF4B-S422E mutant were examined by Western blotting. E. cells in D were treated with 12.5 μM SMI-4a alone, 10 μM Akti-1/2 alone, or their combination. After 48 h, cells were harvested for Annexin V staining as described in C. F. cells in D were treated with 12.5 μM SMI-4a alone, 5 μM rapamycin alone, or their combination for 48 h. Apoptosis was examined as described in C. G. K562 cells expressing vector, eIF4B-WT or eIF4B-S422E mutant were transplanted into nude mice. At the 11^th day after inoculation, mice bearing xenografts were treated with mock or combination of 50 mg/kg SMI-4a and 10 mg/kg rapamycin each other day. At the 29^th day mice were sacrificed. Shown are representative tumors under mock or drug administration. H. the relative tumor volumes were displayed as described in Figure 6B. The average volume of tumor (empty vector) with mock treatment was set as 100%. Error bars, SEM, n = 18 (*P < 0.05). I. tumors in G were immunoblotted with indicated antibodies. Shown are representative results from three independent experiments.