Distinct Dasatinib-Induced Mechanisms of Apoptotic Response and Exosome Release in Imatinib-Resistant Human Chronic Myeloid Leukemia Cells

Although dasatinib is effective in most imatinib mesylate (IMT)-resistant chronic myeloid leukemia (CML) patients, the underlying mechanism of its effectiveness in eliminating imatinib-resistant cells is only partially understood. This study investigated the effects of dasatinib on signaling mechanisms driving-resistance in imatinib-resistant CML cell line K562 (K562RIMT). Compared with K562 control cells, exsomal release, the phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt)/ mammalian target of rapamycin (mTOR) signaling and autophagic activity were increased significantly in K562RIMT cells and mTOR-independent beclin-1/Vps34 signaling was shown to be involved in exosomal release in these cells. We found that Notch1 activation-mediated reduction of phosphatase and tensin homolog (PTEN) was responsible for the increased Akt/mTOR activities in K562RIMT cells and treatment with Notch1 γ-secretase inhibitor prevented activation of Akt/mTOR. In addition, suppression of mTOR activity by rapamycin decreased the level of activity of p70S6K, induced upregulation of p53 and caspase 3, and led to increase of apoptosis in K562RIMT cells. Inhibition of autophagy by spautin-1 or beclin-1 knockdown decreased exosomal release, but did not affect apoptosis in K562RIMT cells. In summary, in K562RIMT cells dasatinib promoted apoptosis through downregulation of Akt/mTOR activities, while preventing exosomal release and inhibiting autophagy by downregulating expression of beclin-1 and Vps34. Our findings reveal distinct dasatinib-induced mechanisms of apoptotic response and exosomal release in imatinib-resistant CML cells.


Introduction
Chronic myeloid leukemia (CML) is characterized by the uncontrolled proliferation of myeloid cells [1]. These leukemic cells contain a characteristic t (9:22) translocation resulting in the fusion of the Abelson (ABL) oncogene to the breakpoint cluster region (BCR) gene, and thus express a constitutively activated fusion protein BCR-ABL1, a tyrosine kinase, that is involved in the pathogenesis of CML [1]. The application of imatinib mesylate (IMT), an ATP-competitive selective BCR-ABL tyrosine kinase inhibitor, significantly improved survival in most CML patients [2]. However, not all patients present Figure 1. More exosomes are released from K562R IMT cells. Exosomes were isolated from the cultured media of K562 and K562R IMT cells, respectively. (A) BCA assay shows that the total amount of exosomal proteins from K562R IMT was significant higher than that from K562. Data are shown as mean ± standard deviation (SD). n = 5 replicate experiments; (B) The exosomal proteins from 5 replicate experiments were equally pulled together. Totally, 100 µg each group was used for immunoblot of TGF-β1, Hsc70, and NKG2D as well as other exosomal markers CD63, Tsg101, and CD81. Culture media alone was used as negative control. As compared with K562, increased abundance of exosomal TGF-β1, Hsc70, and NKG2D was detected in K562R IMT cells.

Activity of mTOR and Autophagy Is Increased in K562R IMT Cells
The mammalian target of rapamycin (mTOR), is a key signaling pathway in cell growth and homeostasis, and was shown to be abnormally regulated in tumors [8]. The mTOR is phosphorylated at Ser 2448 via the PI3 kinase/Akt signaling pathway and also autophosphorylated at Ser 2481 [8]. Immunoblot assay showed that the relative abundance of total mTOR protein was significantly (p < 0.05) higher in K562R IMT than K562 cells. Moreover, the level of phosphorylated mTOR at Ser 2448 was increased significantly (p < 0.01) in K562R IMT as compared with K562 cells. Remarkable difference was not detected for phospho-mTOR at Ser 2481 between K562 and K562R IMT cells (Figure 2A).
The mTOR functions in two distinct complexes. Raptor is a major component of mTOR complex1 (mTORC1) that regulates cell growth, survival, and autophagy, while Rictor is specific marker for mTOR complex 2 (mTORC2) that promotes cellular survival by activating Akt, regulates cytoskeletal dynamics via activating PKCα, and that controls ion transport and growth via SGK1 phosphorylation [8]. Here, upregulation of Raptor expression was shown (p < 0.01) in K562R IMT cells in comparison with K562 ( Figure 2B), implying that mTORC1 activity was increased in K562 cells following imatinib resistance development.
The small GTPase Rheb, in its GTP-bound state, is a necessary and potent stimulator of mTORC1 activity [8]. Consistently, the level of GTP-bound Rheb was significantly higher (p < 0.001) in K562R IMT than K562 cells ( Figure 2C). It was reported that mTOR may be a target of ATF5, or activating transcription factor 5 [9]. As compared with K562, the protein level of ATF5 increased More exosomes are released from K562R IMT cells. Exosomes were isolated from the cultured media of K562 and K562R IMT cells, respectively. (A) BCA assay shows that the total amount of exosomal proteins from K562R IMT was significant higher than that from K562. Data are shown as mean˘standard deviation (SD). n = 5 replicate experiments; (B) The exosomal proteins from 5 replicate experiments were equally pulled together. Totally, 100 µg each group was used for immunoblot of TGF-β1, Hsc70, and NKG2D as well as other exosomal markers CD63, Tsg101, and CD81. Culture media alone was used as negative control. As compared with K562, increased abundance of exosomal TGF-β1, Hsc70, and NKG2D was detected in K562R IMT cells.

Activity of mTOR and Autophagy Is Increased in K562R IMT Cells
The mammalian target of rapamycin (mTOR), is a key signaling pathway in cell growth and homeostasis, and was shown to be abnormally regulated in tumors [8]. The mTOR is phosphorylated at Ser 2448 via the PI3 kinase/Akt signaling pathway and also autophosphorylated at Ser 2481 [8]. Immunoblot assay showed that the relative abundance of total mTOR protein was significantly (p < 0.05) higher in K562R IMT than K562 cells. Moreover, the level of phosphorylated mTOR at Ser 2448 was increased significantly (p < 0.01) in K562R IMT as compared with K562 cells. Remarkable difference was not detected for phospho-mTOR at Ser 2481 between K562 and K562R IMT cells (Figure 2A).
The mTOR functions in two distinct complexes. Raptor is a major component of mTOR complex1 (mTORC1) that regulates cell growth, survival, and autophagy, while Rictor is specific marker for mTOR complex 2 (mTORC2) that promotes cellular survival by activating Akt, regulates cytoskeletal dynamics via activating PKCα, and that controls ion transport and growth via SGK1 phosphorylation [8]. Here, upregulation of Raptor expression was shown (p < 0.01) in K562R IMT cells in comparison with K562 ( Figure 2B), implying that mTORC1 activity was increased in K562 cells following imatinib resistance development.
cytosolic LC3-I form. During autophagy, LC3-I is converted to LC3-II through lipidation that allows for LC3 to become associated with autophagic vesicles [18]. The presence of LC3 in autophagosomes and the conversion of LC3-I to LC3-II have been used as indicators of autophagy [8,18]. LC3-II increased significantly (p < 0.01), whereas LC3-I decreased (p < 0.01) in K562R IMT as compared with K562 cells (Figure 2E), indicating that mTOR-independent autophagy pathway is activated in K562R IMT cells. Activities of mTOR and autophagy are enhanced in K562R IMT cells. Total cellular protein and nuclear protein of K562 and K562R IMT cells was extracted by using RIPA lysis buffer and Nuclear Extraction Kit, respectively. (A) Immunoblot of total mTOR and phospho-mTOR at Ser2481 or Ser2448; (B) Immunoblot of two distinct mTOR complex markers Raptor and Rictor; (C) The level of activated Rheb. GTP-bound Rheb was immunoprecipitated by incubating cellular lysates with the specific mouse anti-active Rheb antibody and Protein A/G agarose and detected by using immunoblot with rabbit anti-Rheb antibody. GDP-or GTPγs-treated K562R IMT lysates were used as the negative or positive control, respectively; (D) Immunoblot of the transcription factor ATF5 in nuclear fractions; (E) Immunoblot of different cleaved forms LC3-I and LC3-II of the autophagy marker LC3. Data are shown as mean ± SD. n = 3 independent experiments. N.S.: non significance.

Induction of mTOR-Independent Autophagy Increases Exosomes Release in K562R IMT Cells
Rapamycin is widely used as an inhibitor of mTORC1 signaling [18]. K562R IMT cells were treated with rapamycin, and immunoblot assay showed that rapamycin downregulated (p < 0.05) the level of phospho-mTOR at Ser 2448 in a dose-dependent manner ( Figure 3A). Exosomes were isolated from the cultured media of K562R IMT cells, and the total exosomal protein and exosomal TGF-β1 abundance were analyzed, respectively. Our data displayed that rapamycin treatment did Figure 2. Activities of mTOR and autophagy are enhanced in K562R IMT cells. Total cellular protein and nuclear protein of K562 and K562R IMT cells was extracted by using RIPA lysis buffer and Nuclear Extraction Kit, respectively. (A) Immunoblot of total mTOR and phospho-mTOR at Ser2481 or Ser2448; (B) Immunoblot of two distinct mTOR complex markers Raptor and Rictor; (C) The level of activated Rheb. GTP-bound Rheb was immunoprecipitated by incubating cellular lysates with the specific mouse anti-active Rheb antibody and Protein A/G agarose and detected by using immunoblot with rabbit anti-Rheb antibody. GDP-or GTPγs-treated K562R IMT lysates were used as the negative or positive control, respectively; (D) Immunoblot of the transcription factor ATF5 in nuclear fractions; (E) Immunoblot of different cleaved forms LC3-I and LC3-II of the autophagy marker LC3. Data are shown as mean˘SD. n = 3 independent experiments. N.S.: non significance.
The small GTPase Rheb, in its GTP-bound state, is a necessary and potent stimulator of mTORC1 activity [8]. Consistently, the level of GTP-bound Rheb was significantly higher (p < 0.001) in K562R IMT than K562 cells ( Figure 2C). It was reported that mTOR may be a target of ATF5, or activating transcription factor 5 [9]. As compared with K562, the protein level of ATF5 increased significantly (p < 0.05) in K562R IMT cells ( Figure 2D), which may be responsible for the overproduction of the total mTOR protein.
Usually, mTOR plays a crucial role in regulating/inhibiting autophagy [18]. Immediately following synthesis, autophagy Light Chain 3 (LC3) is cleaved at the carboxy terminus and yields the cytosolic LC3-I form. During autophagy, LC3-I is converted to LC3-II through lipidation that allows for LC3 to become associated with autophagic vesicles [18]. The presence of LC3 in autophagosomes and the conversion of LC3-I to LC3-II have been used as indicators of autophagy [8,18]. LC3-II increased significantly (p < 0.01), whereas LC3-I decreased (p < 0.01) in K562R IMT as compared with K562 cells ( Figure 2E), indicating that mTOR-independent autophagy pathway is activated in K562R IMT cells.

Induction of mTOR-Independent Autophagy Increases Exosomes Release in K562R IMT Cells
Rapamycin is widely used as an inhibitor of mTORC1 signaling [18]. K562R IMT cells were treated with rapamycin, and immunoblot assay showed that rapamycin downregulated (p < 0.05) the level of phospho-mTOR at Ser 2448 in a dose-dependent manner ( Figure 3A). Exosomes were isolated from the cultured media of K562R IMT cells, and the total exosomal protein and exosomal TGF-β1 abundance were analyzed, respectively. Our data displayed that rapamycin treatment did not influence the amount of total exosomal protein ( Figure 3B), and also showed no effect on the abundance of exosomal TGF-β1 in K562R IMT cells ( Figure 3C). Moreover, immunoblot assay showed that rapamycin treatment did not affect the conversion of LC3-I to LC3-II in K562R IMT cells ( Figure 3D). Our results suggested that mTORC1 inhibition by rapamycin showed no effects on exosomes release and autophagic activity in K562R IMT cells. not influence the amount of total exosomal protein ( Figure 3B), and also showed no effect on the abundance of exosomal TGF-β1 in K562R IMT cells ( Figure 3C). Moreover, immunoblot assay showed that rapamycin treatment did not affect the conversion of LC3-I to LC3-II in K562R IMT cells ( Figure 3D). Our results suggested that mTORC1 inhibition by rapamycin showed no effects on exosomes release and autophagic activity in K562R IMT cells. (A) Immunoblot assay shows that rapamycin dose-dependently decreased phospho-mTOR level at Ser2448 in K562R IMT cells; (B) Exosomes were isolated from the cultured media of K562R IMT cells, and total exosomal protein was determined using BCA assay. Rapamycin showed no effect on the amount of exosomal proteins in K562R IMT cells; (C) Immunoblot of TGF-β1 and CD63 in the isolated exosomes. The abundance of TGFβ1 was not influenced by rapamycin application; (D) Immunoblot assay shows that rapamycin treatment did not affect the level of LC3-I and LC3-II in K562R IMT cells; (E) As compared with K562, the abundance of beclin-1 and Vps34 increased significantly in K562R IMT cells; (F) Knockdown of beclin-1 was performed in K562R IMT cells by introduction of siRNA targeting to beclin-1 (siBln). The scrambled siRNA was used as control (siCTL). After 24 h, immunoblotting was used to evaluate the level of beclin-1, Vps34, and LC3 expression. In K562R IMT cells, increased LC3-II and Vps34 was prevented significantly by beclin-1 knockdown; (G) K562R IMT cells were treated for 12 h with the autophagy inhibitor spautin-1, and expression level of LC3 was analyzed using immunoblot assay.  (A) Immunoblot assay shows that rapamycin dose-dependently decreased phospho-mTOR level at Ser2448 in K562R IMT cells; (B) Exosomes were isolated from the cultured media of K562R IMT cells, and total exosomal protein was determined using BCA assay. Rapamycin showed no effect on the amount of exosomal proteins in K562R IMT cells; (C) Immunoblot of TGF-β1 and CD63 in the isolated exosomes. The abundance of TGFβ1 was not influenced by rapamycin application; (D) Immunoblot assay shows that rapamycin treatment did not affect the level of LC3-I and LC3-II in K562R IMT cells; (E) As compared with K562, the abundance of beclin-1 and Vps34 increased significantly in K562R IMT cells; (F) Knockdown of beclin-1 was performed in K562R IMT cells by introduction of siRNA targeting to beclin-1 (siBln). The scrambled siRNA was used as control (siCTL). After 24 h, immunoblotting was used to evaluate the level of beclin-1, Vps34, and LC3 expression. In K562R IMT cells, increased LC3-II and Vps34 was prevented significantly by beclin-1 knockdown; (G) K562R IMT cells were treated for 12 h with the autophagy inhibitor spautin-1, and expression level of LC3 was analyzed using immunoblot assay. Increased LC3-II was inhibited significantly by spautin-1 in K562R IMT cells; (H) Immunoblot of TGF-β1 in the isolated exosomes from K562 and K562R IMT cells treated with spautin-1 or siBln. Beclin-1 knockdown or spautin-1 treatment prevented increase of exosomal TGFβ1 in K562R IMT cells. Data are shown as mean˘SD. n = 3 independent experiments. N.S.: non significance.
Vps34, a member of the phosphatidylinositol 3-/4 (PI3/PI4)-kinase family, plays an important role in the regulation of mTOR protein synthesis, and also forms a complex with beclin-1 that promotes autophagy and tumor suppression [19]. Therefore, the protein level of beclin-1 and Vps34 was evaluated by using immunoblot assay in K562R IMT cells. Our data showed that beclin-1 and Vps34 both increased significantly (p < 0.05) in K562 cells following imatinib resistance development ( Figure 3E). To investigate the role of beclin-1 and Vps34 in regulating autophagic activity, silence of beclin-1 was obtained by using the specific siRNA. In K562R IMT cells, knockdown of beclin-1 significantly (p < 0.05) decreased the protein level of Vps34 and LC3-II ( Figure 3F), implying that upregulation of beclin-1 and Vps34 may be responsible for the increased autophagic activity in K562R IMT cells.
Spautin-1, a very specific and potent autophagy inhibitor in mammalian cells, can promote degradation of Vps34 complexes and block the pro-survival autophagy pathway in cancer cells [20]. In K562R IMT cells, the conversion of LC3-I to LC3-II was prevented significantly (p < 0.01) by spautin-1 application in a dose-dependent manner ( Figure 3G). The effect of autophagy inhibition on exosomes release was further explored in K562R IMT cells. The abundance of exosomal TGF-β1 was also decreased significantly (p < 0.05) in K562R IMT cells by either beclin-1 knockdown or spautin-1 treatment ( Figure 3H).

Loss of PTEN by Notch1 Activation Increases mTOR Activity in K562R IMT Cells
PI3K and Akt is a major pathway that activates mTOR signaling [7,8]. In CML, BCR-ABL1 can promote cell survival by activating the PI3K/Akt pathway [21]. In the current study, we found that expression level of BCR-ABL1 was downregulated significantly in K562R IMT compared to K562 cells ( Figure 4A). Nevertheless, the level of phosphorylated PI3K/p85 Tyr458 and Akt Ser473 increased significantly (p < 0.001) in K562R IMT compared to K562 cells ( Figure 4B). The level of phospho-MAPK Thr180/Tyr182 and phospho-Erk1/2 Thr202/Tyr204 showed no difference between K562 and K562R IMT cells ( Figure 4C), indicating a specific involvement of the PI3K/Akt pathway activation in the development of imatinib resistance. The PI3K/Akt pathway is antagonized by various factors including PTEN, glycogen synthase kinase-3 beta (GSK3β), and homeobox protein 9 (HB9) [22]. Unlike most of the protein tyrosine phosphatases, PTEN preferentially dephosphorylates phosphoinositide substrates, and functions as a tumor suppressor by negatively regulating PI3K/Akt activation [22]. In K562R IMT cells, PTEN expression level was decreased significantly (p < 0.001) ( Figure 4D). It has been shown that Hes-1 could repress PTEN transcription downstream of Ras and Notch1 activation [23]. Notch pathway, a highly conserved cellular signaling system, involves diverse gene regulation mechanisms, and is dysregulated in many cancers [24]. Upon the ligand such as jagged-1 activation, Notch1 intracellular domain (NICD) produced by γ-secretase cleavage is released and transported to the nucleus thus forming a transcriptional activator complex with Rbpj and activating target gene transcriptions [24]. Immunoblot showed an obvious (p < 0.05) increase of both cytosolic and nucleic NICD in K562R IMT cells as compared with K562 ( Figure 4E,F). The specific γ-secretase inhibitor (GSI; RO4929097) was used to block Notch function. In the nuclear fraction from K562R IMT cells, NICD and Hes-1 level was dose-dependently (p < 0.05) decreased by GSI application ( Figure 4F). In addition, GSI significantly (p < 0.01) upregulated the expression level of PTEN in K562R IMT cells ( Figure 4G). The increased phospho-Akt Ser473 and phospho-mTOR Ser2448 level was significantly decreased (p < 0.05) by GSI treatment in K562R IMT cells ( Figure 4H). These results demonstrated that mTOR activation is mediated by the PI3K/Akt pathway through Notch-induced inhibition of PTEN in K562R IMT cells.

Figure 4.
Loss of PTEN by Notch1 activation increases mTOR activity in K562R IMT cells. Total cellular protein and nuclear protein was extracted by using RIPA lysis buffer and Nuclear Extraction Kit, respectively. Immunoblot assay was then performed. (A) Compared to K562, The BCR-ABL fusion protein level decreased significantly in K562R IMT cells; (B) Abundance of phospho-PI3K/p85 Tyr458 and phospho-Akt Ser473 was obviously higher in K562R IMT than K562 cells; (C) The level of phospho-Erk1/2 Thr202/Tyr204 and phospho-p38MAPK Thr180/Tyr182 showed no difference between K562 and K562R IMT cells; (D) Expression level of PTEN inK562R IMT cells was lower than that in K562; (E) Increased Notch 1 intracellular domain (NICD) was detected in K562R IMT cells; (F) K562R IMT cells were treated for 12 h with Notch γ-secretase inhibitor (GSI, RO4929097). Abundance of NICD and transcription factor Hes-1 was evaluated in nuclear fractions. GSI treatment dose-dependently prevented increase of NICD and Hes-1 in K562R IMT cells; (G) Effect of GSI (10 µM) on PTEN expression level, showing that GSI increased PTEN level in K562R IMT cells; (H) Effect of GSI (10 µM) on phospho-Akt Ser473 and phospho-mTOR Ser2448 levels. In K562R IMT cells, increased phospho-Akt Ser473 and phospho-mTOR Ser2448 was inhibited by GSI application. Data are shown as mean ± SD. n = 3 independent experiments. N.S.: non significance.

Suppression of mTOR, Not Autophagic Activity, Increases Apoptosis in K562R IMT Cells
The mTOR signaling promotes cellular survival possibly by inhibition of apoptosis through downstream signaling p70S6K-induced p53 blockage [25]. The effect of rapamycin, GSI and spautin-1 on cellular apoptosis was explored in K562R IMT cells by using flow cytometry. Compared to spautin-1, apoptosis was induced significantly (p < 0.001) by rapamycin and GSI treatment ( Figure 5A). As compared with K562, the phosphorylated p70S6K level at Thr389 increased significantly (p < 0.001) in K562R IMT cells, which were significantly inhibited (p < 0.001) by mTOR inhibitor rapamycin and Notch1 inhibitor GSI, but not by autophagy inhibitor spautin-1 ( Figure 5B). The level of apoptotic proteins phospho-p53 Ser15 , Bax, and the activated caspase 3 was significantly Total cellular protein and nuclear protein was extracted by using RIPA lysis buffer and Nuclear Extraction Kit, respectively. Immunoblot assay was then performed. (A) Compared to K562, The BCR-ABL fusion protein level decreased significantly in K562R IMT cells; (B) Abundance of phospho-PI3K/p85 Tyr458 and phospho-Akt Ser473 was obviously higher in K562R IMT than K562 cells; (C) The level of phospho-Erk1/2 Thr202/Tyr204 and phospho-p38MAPK Thr180/Tyr182 showed no difference between K562 and K562R IMT cells; (D) Expression level of PTEN inK562R IMT cells was lower than that in K562; (E) Increased Notch 1 intracellular domain (NICD) was detected in K562R IMT cells; (F) K562R IMT cells were treated for 12 h with Notch γ-secretase inhibitor (GSI, RO4929097). Abundance of NICD and transcription factor Hes-1 was evaluated in nuclear fractions. GSI treatment dose-dependently prevented increase of NICD and Hes-1 in K562R IMT cells; (G) Effect of GSI (10 µM) on PTEN expression level, showing that GSI increased PTEN level in K562R IMT cells; (H) Effect of GSI (10 µM) on phospho-Akt Ser473 and phospho-mTOR Ser2448 levels. In K562R IMT cells, increased phospho-Akt Ser473 and phospho-mTOR Ser2448 was inhibited by GSI application. Data are shown as mean˘SD. n = 3 independent experiments. N.S.: non significance.

Suppression of mTOR, Not Autophagic Activity, Increases Apoptosis in K562R IMT Cells
The mTOR signaling promotes cellular survival possibly by inhibition of apoptosis through downstream signaling p70S6K-induced p53 blockage [25]. The effect of rapamycin, GSI and spautin-1 on cellular apoptosis was explored in K562R IMT cells by using flow cytometry. Compared to spautin-1, apoptosis was induced significantly (p < 0.001) by rapamycin and GSI treatment ( Figure 5A). As compared with K562, the phosphorylated p70S6K level at Thr389 increased significantly (p < 0.001) in K562R IMT cells, which were significantly inhibited (p < 0.001) by mTOR inhibitor rapamycin and Notch1 inhibitor GSI, but not by autophagy inhibitor spautin-1 ( Figure 5B). The level of apoptotic proteins phospho-p53 Ser15 , Bax, and the activated caspase 3 was significantly (p < 0.05) lower in K562R IMT than that in K562 cells. As compared with spautin-1, GSI and rapamycin significantly (p < 0.05) increased the level of phospho-p53 Ser15 , Bax, and the activated caspase 3 in K562R IMT cells ( Figure 5C). Bcl-2 level showed no difference between K562 and K562R IMT cells. Spautin-1 treatment slightly decreased Bcl-2 level in K562R IMT cells ( Figure 5C). Our results implied that mTOR, not autophagy pathway, plays a predominant role in regulation of apoptotic signaling in K562R IMT cells. (p < 0.05) lower in K562R IMT than that in K562 cells. As compared with spautin-1, GSI and rapamycin significantly (p < 0.05) increased the level of phospho-p53 Ser15 , Bax, and the activated caspase 3 in K562R IMT cells ( Figure 5C). Bcl-2 level showed no difference between K562 and K562R IMT cells. Spautin-1 treatment slightly decreased Bcl-2 level in K562R IMT cells ( Figure 5C). Our results implied that mTOR, not autophagy pathway, plays a predominant role in regulation of apoptotic signaling in K562R IMT cells. Apoptosis assay showed that compared to control, rapamycin and GSI significantly induced apoptosis in the K562R IMT cell; (B) Abundance of phospho-p70S6K Thr389 was analyzed using immunoblot assay, showing that increased phospho-p70S6K in K562R IMT cell was prevented by rapamycin and GSI; (C) Immunobot assay was performed for expression of pro-apoptotic proteins phospho-p53 Ser15 , Bax and active caspase 3 as well as anti-apoptotic protein Bcl-2. In K562R IMT cells, reduction of phospho-p53 Ser15 , Bax and activated caspase 3 was inhibited by rapamycin and GSI. Bcl-2 level showed no difference between K562 and K562R IMT cells. Spautin-1 treatment only decreased Bcl-2 level in K562R IMT cells; (D) Effect of rapamycin, GSI and spautin-1 on cell cycle proteins Cyclin D1 and p21 was evaluated using immunoblot assay. In K562R IMT cells, cyclin D1 increased while p21 decreased, both of which were prevented by rapamycin and GSI; (E) Cellular proliferative ability was assessed using MTT assay in K562R IMT cells, showing that the percentage of cellular proliferation was decreased by rapamycin and GSI. Data are shown as mean ± SD. n = 3 or 4 independent experiments. Apoptosis assay showed that compared to control, rapamycin and GSI significantly induced apoptosis in the K562R IMT cell; (B) Abundance of phospho-p70S6K Thr389 was analyzed using immunoblot assay, showing that increased phospho-p70S6K in K562R IMT cell was prevented by rapamycin and GSI; (C) Immunobot assay was performed for expression of pro-apoptotic proteins phospho-p53 Ser15 , Bax and active caspase 3 as well as anti-apoptotic protein Bcl-2. In K562R IMT cells, reduction of phospho-p53 Ser15 , Bax and activated caspase 3 was inhibited by rapamycin and GSI. Bcl-2 level showed no difference between K562 and K562R IMT cells. Spautin-1 treatment only decreased Bcl-2 level in K562R IMT cells; (D) Effect of rapamycin, GSI and spautin-1 on cell cycle proteins Cyclin D1 and p21 was evaluated using immunoblot assay. In K562R IMT cells, cyclin D1 increased while p21 decreased, both of which were prevented by rapamycin and GSI; (E) Cellular proliferative ability was assessed using MTT assay in K562R IMT cells, showing that the percentage of cellular proliferation was decreased by rapamycin and GSI. Data are shown as mean˘SD. n = 3 or 4 independent experiments.
Additionally, the role of mTOR and autophagy pathway on cellular proliferation was also investigated. Cyclin D1, a protein required for progression through the G1 phase of the cell cycle, has been found to be overproduced in some cancer cells [26]. In comparison, p21, a potent cyclin-dependent kinase inhibitor, functions as a regulator of cell cycle progression at G1 and S phase [26]. As compared with K562, overproduced cyclin D1 and decreased p21 was significantly (p < 0.05) detected in K562R IMT cells, which was prevented (p < 0.01) by rapamycin and GSI, but not by spautin-1 ( Figure 5D). MTT cellular proliferation assay showed that rapamycin and GSI significantly (p < 0.05) decreased the percentage of proliferative K562R IMT cells ( Figure 5E).

Dasatinib Enhances Apoptosis by Preventing mTOR Activation via Targeting Akt Pathway in K562R IMT Cells
It has been shown that the second-generation tyrosine kinase inhibitor dasatinib functions well in the CML patients with imatinib resistance [4]. However, the underlying mechanism has notbeen fully clarified. In dasatinib-treated K562R IMT cells, the percentage of apoptotic cell increased significantly (p < 0.01) in a dose-dependent manner ( Figure 6A). Inhibition of mTOR signaling with rapamycin, not autophagy inhibition by spautin-1, significantly (p < 0.05) enhanced the efficacy of dasatinib on induction of cellular apoptosis in K562R IMT cells ( Figure 6A). Consistently, dasatinib treatment significantly (p < 0.05) increased the level of the activated caspase 3, which was enhanced (p < 0.01) by rapamycin, not by spautin-1 in K562R IMT cells ( Figure 6B). Furthermore, dasatinib dose-dependently prevented increase of phospho-p70S6K Thr389 in K562R IMT cells ( Figure 6C). The potential target of dasatinib on cellular apoptosis was further investigated. In K562R IMT cells, dasatinib significantly (p < 0.05) decreased the level of phospho-Akt Ser473 and phospho-mTOR Ser2448 in a dose-dependent manner ( Figure 6D), but only showed an obvious influence on the level of nuclear NICD and total PTEN at high concentrations (25 nM) ( Figure 6E). Our results indicated that dasatinib may promote apoptosis by preventing mTOR activation predominantly through downregulation of Akt activation in K562R IMT cells.

Dasatinib Decreases Exosome Release by Inhibiting Autophagy Activation in K562R IMT Cells
Our data has shown that exosome release was increased in K562R IMT cells. The effect of dasatinib on exosome release was thus explored in the current study. Total exosomal protein was dose-dependently (p < 0.05) decreased by dasatinib treatment in K562R IMT cells, which was enhanced (p < 0.05) by spautin-1 administration, not by rapamycin ( Figure 7A). The abundance of TGF-β1 in the isolated exosomes from the media of K562R IMT cells was also significantly (p < 0.05) decreased by dasatinib treatment, and spautin-1 application, not rapamycin, showed a combined or synergistic effect ( Figure 7B). Immunoblot assay showed that beclin-1 and LC3-II were also (p < 0.05) decreased by dasatinib, which was (p < 0.05) enhanced by spautin-1, not rapamycin ( Figure 7C,D). Our results demonstrated that dasatinib decreases exosome release by inhibiting autophagy pathway activation through downregulation of beclin-1 in K562R IMT cells. Additionally, the role of mTOR and autophagy pathway on cellular proliferation was also investigated. Cyclin D1, a protein required for progression through the G1 phase of the cell cycle, has been found to be overproduced in some cancer cells [26]. In comparison, p21, a potent cyclin-dependent kinase inhibitor, functions as a regulator of cell cycle progression at G1 and S phase [26]. As compared with K562, overproduced cyclin D1 and decreased p21 was significantly (p < 0.05) detected in K562R IMT cells, which was prevented (p < 0.01) by rapamycin and GSI, but not by spautin-1 ( Figure 5D). MTT cellular proliferation assay showed that rapamycin and GSI significantly (p < 0.05) decreased the percentage of proliferative K562R IMT cells ( Figure 5E).

Dasatinib Enhances Apoptosis by Preventing mTOR Activation via Targeting Akt Pathway in K562R IMT Cells
It has been shown that the second-generation tyrosine kinase inhibitor dasatinib functions well in the CML patients with imatinib resistance [4]. However, the underlying mechanism has notbeen fully clarified. In dasatinib-treated K562R IMT cells, the percentage of apoptotic cell increased significantly (p < 0.01) in a dose-dependent manner ( Figure 6A). Inhibition of mTOR signaling with rapamycin, not autophagy inhibition by spautin-1, significantly (p < 0.05) enhanced the efficacy of dasatinib on induction of cellular apoptosis in K562R IMT cells ( Figure 6A). Consistently, dasatinib treatment significantly (p < 0.05) increased the level of the activated caspase 3, which was enhanced (p < 0.01) by rapamycin, not by spautin-1 in K562R IMT cells ( Figure 6B). Furthermore, dasatinib dose-dependently prevented increase of phospho-p70S6K Thr389 in K562R IMT cells ( Figure 6C). The potential target of dasatinib on cellular apoptosis was further investigated. In K562R IMT cells, dasatinib significantly (p < 0.05) decreased the level of phospho-Akt Ser473 and phospho-mTOR Ser2448 in a dose-dependent manner ( Figure 6D), but only showed an obvious influence on the level of nuclear NICD and total PTEN at high concentrations (25 nM) ( Figure 6E). Our results indicated that dasatinib may promote apoptosis by preventing mTOR activation predominantly through downregulation of Akt activation in K562R IMT cells.

Dasatinib Decreases Exosome Release by Inhibiting Autophagy Activation in K562R IMT Cells
Our data has shown that exosome release was increased in K562R IMT cells. The effect of dasatinib on exosome release was thus explored in the current study. Total exosomal protein was dose-dependently (p < 0.05) decreased by dasatinib treatment in K562R IMT cells, which was enhanced (p < 0.05) by spautin-1 administration, not by rapamycin ( Figure 7A). The abundance of TGF-β1 in the isolated exosomes from the media of K562R IMT cells was also significantly (p < 0.05) decreased by dasatinib treatment, and spautin-1 application, not rapamycin, showed a combined or synergistic effect ( Figure 7B). Immunoblot assay showed that beclin-1 and LC3-II were also (p < 0.05) decreased by dasatinib, which was (p < 0.05) enhanced by spautin-1, not rapamycin ( Figure 7C,D). Our results demonstrated that dasatinib decreases exosome release by inhibiting autophagy pathway activation through downregulation of beclin-1 in K562R IMT cells.        Exosomes were isolated, and total exosomal protein was determined using BCA assay. Compared to K562, the amount of exosomal proteins increased significantly in K562R IMT cells, which was prevented by dasatinib in a dose-dependent manner and spautin-1 showed a synergistic effect; (B) Immunoblot of TGF-β1 in the isolated exosomes. Dasatinib dose-dependently decreased the abundance of TGF-β1 in K562R IMT cells, and spautin-1 showed a synergistic effect; (C) The level of beclin-1 was analyzed using immunoblot assay. In K562R IMT cells, dasatinib dose-dependently decreased the level of beclin-1, and spautin-1 showed a synergistic effect; (D) The levels of LC3-I and LC3-II were analyzed using immunoblot assay. Increased LC3-II level was dose-dependently prevented by dasatinib, and spautin-1 showed a synergistic effect. Data are shown as mean ± SD. n = 3 independent experiments.

Discussion
Overactivation of the PI3K/Akt signaling pathway induced by upregulation of the BCR-ABL level is a major factor for the development of imatinib resistance in CML [3,4,21]. In addition, frequencies of mutation in BCR-ABL kinase domain appear to increase and induce resistance to tyrosine kinase inhibitors as CML progresses [27]. It was reported that sequencing of the BCR-ABL kinase domain did not find mutations both in K562 and in K562R IMT cells [28,29]. As Lee et al. [30] described, we also found that the BCR-ABL level decreased significantly in K562R IMT cells ( Figure 4A). Therefore, overactivation of PI3K/Akt in these cells is BCR-ABL-independent ( Figures 4B and 6D). The mTOR pathway is a major target of the PI3K and Akt signaling [8]. There are two distinct complexes of mTOR, mTORC1 and mTORC2, which are indicated by the specific markers Raptor and Rictor, and contain mTOR phosphorylated predominantly on Ser2448 and Ser2481, respectively [8,31]. GTP-bound Rheb, a small GTPase, is a potent stimulator of the mTORC1 activity responsible for activation of mTOR on Ser2448 [8]. In this study, increased phospho-mTOR Ser2448 and Raptor as well as increased active form of Rheb ( Figure 2C) were significantly detected in K562R IMT cells (Figure 2A,B), indicating that the mTORC1 is activated in Figure 7. Dasatinib decreases exosome release by inhibiting autophagy activity in K562R IMT cells. K562R IMT cells were treated for 24 h with dasatinib (Dtn) at different concentrations. Alternatively, K562R IMT cells were pretreated for 12 h with rapamycin (5 nM) or spautin-1 (2 µM), and then treated for 24 h with dasatinib (5 nM). (A) Exosomes were isolated, and total exosomal protein was determined using BCA assay. Compared to K562, the amount of exosomal proteins increased significantly in K562R IMT cells, which was prevented by dasatinib in a dose-dependent manner and spautin-1 showed a synergistic effect; (B) Immunoblot of TGF-β1 in the isolated exosomes. Dasatinib dose-dependently decreased the abundance of TGF-β1 in K562R IMT cells, and spautin-1 showed a synergistic effect; (C) The level of beclin-1 was analyzed using immunoblot assay. In K562R IMT cells, dasatinib dose-dependently decreased the level of beclin-1, and spautin-1 showed a synergistic effect; (D) The levels of LC3-I and LC3-II were analyzed using immunoblot assay. Increased LC3-II level was dose-dependently prevented by dasatinib, and spautin-1 showed a synergistic effect. Data are shown as mean˘SD. n = 3 independent experiments.

Discussion
Overactivation of the PI3K/Akt signaling pathway induced by upregulation of the BCR-ABL level is a major factor for the development of imatinib resistance in CML [3,4,21]. In addition, frequencies of mutation in BCR-ABL kinase domain appear to increase and induce resistance to tyrosine kinase inhibitors as CML progresses [27]. It was reported that sequencing of the BCR-ABL kinase domain did not find mutations both in K562 and in K562R IMT cells [28,29]. As Lee et al. [30] described, we also found that the BCR-ABL level decreased significantly in K562R IMT cells ( Figure 4A). Therefore, overactivation of PI3K/Akt in these cells is BCR-ABL-independent ( Figures 4B and 6D). The mTOR pathway is a major target of the PI3K and Akt signaling [8]. There are two distinct complexes of mTOR, mTORC1 and mTORC2, which are indicated by the specific markers Raptor and Rictor, and contain mTOR phosphorylated predominantly on Ser2448 and Ser2481, respectively [8,31]. GTP-bound Rheb, a small GTPase, is a potent stimulator of the mTORC1 activity responsible for activation of mTOR on Ser2448 [8]. In this study, increased phospho-mTOR Ser2448 and Raptor as well as increased active form of Rheb ( Figure 2C) were significantly detected in K562R IMT cells (Figure 2A,B), indicating that the mTORC1 is activated in imatinib-resistant CML cells. Additionally, our results suggested that upregulation of ATF5 ( Figure 2D) may lead to an increase of total mTOR protein (Figure 2A) since ATF5 is a key transcription factor of mTOR. The mTOR signaling promotes cellular survival possibly by inhibition of apoptosis through downstream signaling p70S6K-induced p53 blockage [8]. In our study, increased phospho-p70S6K Thr389 but decreased apoptotic proteins including phospho-p53 Ser15 , Bax and active caspase 3 were detected in K562R IMT cells, compared to K562 cells ( Figure 5B,C).
Generally, autophagy pathway is negatively regulated by mTOR signaling [18]. Rapamycin inhibits mTOR pathway by directly binding to mTORC1 complex [32]. Our data demonstrated that the increased autophagy activity is mTOR-independent in imatinib-resistant CML cells ( Figures 2E and 3B-D). It is well known that autophagy activity can also be promoted by Vps34 by forming a complex with beclin-1 [20]. In this study, compared to K562, beclin-1 and Vps34 both increased significantly in K562R IMT cells ( Figure 3E). Consistently, infection of leukemia cells including K562 cells with adenovirus overexpressing beclin-1 enhanced autophagic activity [33]. Mounting evidence discloses a close relationship between the autophagy pathway and the biogenesis and secretion of exosomes [10]. It has been well known that exosomes have some specialized functions, either a beneficial or a detrimental impact on neighboring cells [34]. Exosome release has been reported in the human CML cell line K562 and LAMA84 cells [11,12,16,35,36]. In this study, we also found that K562 cells release or secrete many more exosomes when developing resistance to imatinib ( Figure 1A,B). In K562 cells, it has been reported that spautin-1, a specific autophagy inhibitor, significantly blocked imatinib-induced autophagic activation by downregulating beclin-1 [13]. Our results also showed that spautin-1 treatment or beclin-1 knockdown prevented increase of autophagic activity and thus exosomal TGF-β1 release ( Figure 3F-H). Therefore, our findings further demonstrated that the mTOR-independent beclin-1/Vps34 signaling may be responsible for induction of autophagy and exosomal release in K562R IMT cells. Nevertheless, the molecular components and function of exosomes need be further examined in imatinib-resistant CML cells.
It has been well known that PTEN preferentially dephosphorylates phosphoinositide substrates, and functions as a tumor suppressor by negatively regulating the Akt signaling pathway [22]. We found that the PTEN protein level was much lower in K562R IMT cells than that in K562 ( Figure 4D). Dahia et al. also described that the abundance of PTEN both at mRNA and protein level was very low in three myeloid cell lines including K562, KU812, and U937, and that sequencing analysis showed no mutation of PTEN in these cell lines [37]. As the downstream target of Ras and Notch1 pathway, Hes-1 could repress PTEN transcription [23]. Notch pathway necessary for diverse gene regulation is dysregulated in many cancers [23,24]. Upon the ligand such as jagged-1 induced activation, notch intracellular domain (NICD) is produced through cleavage of γ-secretase, and then transported to the nucleus and forms a transcriptional activator complex with Rbpj [24]. An obvious increase of both cytosolic and nucleic NICD as well as nuclear Hes-1 was detected in K562R IMT cells as compared with K562 ( Figure 4E,F). RO4929097, the specific Notch γ-secretase inhibitor (GSI), effectively blocks Notch signaling activation [24]. In K562R IMT cells, our results from GSI inhibition assay demonstrated that Notch1 activation-mediated downregulation of PTEN protein level may be responsible for Akt activation and mTOR activity induction ( Figure 4F-H).
Furthermore, the effects of Notch1, mTOR and autophagy pathways on apoptosis induction were investigated by using their specific inhibitors in this study. Our data suggested that rapamycin and GSI treatment, not spautin-1 induced apoptosis by reducing phospho-p70S6K Thr389 ( Figure 5B) and increasing apoptotic proteins in K562R IMT cells ( Figure 5C). In K562 cells, it was reported that spautin-1 enhanced imatinib-induced cell apoptosis by inactivating PI3K/Akt and activating its downstream protein GSK3β, leading to downregulation of the anti-apoptotic proteins Mcl-1 and Bcl-2, the downstream effectors of autophagic signaling [13]. In the present study, spautin-1 led to a slight reduction of Bcl-2 in K562R IMT cells ( Figure 5C), implying that inhibition of autophagic activity may play a minor role in induction of apoptosis in imatinib-resistant K562 cells. Additionally, our results also showed that rapamycin and GSI significantly decreased the proliferative ability of K562R IMT cells ( Figure 5D,E). Hence, our results implied that mTOR may play a predominant role in regulation of apoptotic and proliferative signaling in K562R IMT cells. Nevertheless, it was reported that the activated Notch signaling by overexpression of NICD in K562 cells mildly but significantly inhibited cell proliferation and reduced the ability of colony formation, suggesting that the Notch signaling may function as a tumor inhibitor in human CML cells [38]. Therefore, further studies are needed to illustrate the function and mechanisms of Notch pathway in human CML, especially in imatinib resistant CML cells.
Rapamycin, not spautin-1, significantly enhanced the efficacy of dasatinib on induction of cellular apoptosis in K562R IMT cells ( Figure 6A,B). In K562R IMT cells, dasatinib dose-dependently decreased phosphorylation of Akt Ser473 , mTOR Ser2448 and p70S6K Thr389 , but only showed a slight effect on NICD and PTEN at high concentrations ( Figure 6C,D). Our results suggested that the Akt/mTOR/p70S6K/caspase 3 pathway may be involved in dasatinib-mediated growth suppression of K562R IMT cells. Nevertheless, it should be further verified by overexpressing constitutively active p70S6K in K562R IMT cells. Similarly, it was reported that the downregulation of the PI3K/Akt/mTORC1 signaling cascades may be a crucial mediator in the inhibition of proliferation and induction of apoptosis by resveratrol in K562 cells [39]. In addition, it was found that activation of p38MAPK signaling pathway is essential for the anti-leukemic effects of dasatinib [40], but we did not find obvious difference of phospho-p38MAPK Thr180/Tyr182 and phospho-Erk1/2 Thr202/Tyr204 between K562 and K562R IMT cells ( Figure 4C). In several imatinib-resistant CML cell lines such as K562R, LAMA84R, and KCL22R, simvastatin, one of the most pharmacologically potent inhibitors of HMG-CoA reductase, was found to have a synergistic killing effect by induction of apoptosis and cell cycle arrest by inhibiting tyrosine phosphorylation and activating STAT5 and STAT3 [41]. Nilotinib is more potent than imatinib in inhibiting BCR-ABL tyrosine kinase activity and proliferation of BCR-ABL-overexpressing cells [42]. In K562 cells resistant to nilotinib, it was found that dasatinib induces apoptosis potentially by inhibiting Lyn kinase activity, downregulating cyclin D1 and upregulating p21 [43]. Therefore, the mechanism by which CML cells develop resistance following treatment with imatinib, and the mechanism of the killing effects of the novel anti-CML drugs, should be further investigated.
Total exosome release was reduced by imatinib and dasatinib in K562 cells [12]. In this study, we found that dasatinib decreased beclin-1 and LC3-II expression as well as exosomal release in a dose-dependent manner in K562R IMT cells, which were enhanced by spautin-1, not by rapamycin (Figure 7). Our results implied that dasatinib decreased exosome release by inhibiting autophagy activity by downregulating beclin-1 in K562R IMT cells. Additionally, it was reported that Rab11 and VAMP3 are required for the fusion between multivesicular bodies (MVBs) with autophagosomes to allow the maturation of the autophagosome, while VAMP7 and ATPase NSF, a protein required for SNAREs disassembly, participate in the fusion between MVBs with the plasma membrane to release exosomes into the extracellular medium [44].
Altogether, our study indicated a distinct role of mTOR signaling and autophagic pathway in imatinib-resistant K562 cells, being responsible for inhibition of apoptosis and induction of exosome release, respectively. Moreover, our data demonstrated that dasatinib promotes cellular apoptosis through downregulation of Akt/mTOR activity, and prevents exosome release by inhibiting beclin-1/Vps34-depedndent autophagic activity ( Figure 8). Notably, we only used K562R IMT cells in the current study, and a variety of imatinib-resistant CML cell lines need be examined to understand well the mechanisms of the killing effects of dasatinib in CML patients. autophagy inhibitor; Rapamycin: mTOR inhibitor; si-Beclin-1: knockdown of beclin-1 by siRNA. Red circle: phosphorylation status.

Cell Culture and Treatment
The CML cell line K562 (ATCC, Manassas, VA, USA) was cultured at 37 °C in DMEM containing 10% FBS (Gibco, Waltham, MA, USA) and 100 U/mL of Penicillin/Streptomycin. To make imatinib-resistant CML cell line, K562 was continuously exposed to higher concentrations of imatinib mesylate (IMT; Sigma-Aldrich, St. Louis, MO, USA) in stepwise increase of 100 nM from 0.1 to 1 µM after 7 days of culture [43]. Finally, the viable imatinib-resistant cells (K562R IMT ) were maintained in culture media containing 1 µM of imatinib. To evaluate the effects of dasatinib on cellular apoptosis and exosomes release, K562R IMT cells were treated with the indicated dosage of dasatinib (Sigma-Aldrich, St. Louis, MO, USA) for 24 h in the absence of imatinib. As Okabe et al. [45] reported, our preliminary results demonstrated that removal of imatinib for 24 hours showed no effects on cleaved caspase 3 and Akt activation at Ser473 in imatinib-resistant K562 cells. In imatinib-resistant K562 cells, BCR-ABL level is decreased. Reduction of PTEN via Notch/Hes-1 signaling leads to the PI3K/Akt/mTOR pathway activation, which is responsible for upregulation of p-p70S6K Thr389 and thus inhibition of apoptosis. GSI and rapamycin, not spautin-1, results in induction of apoptosis. In addition, activation of autophagy through mTOR-independent beclin-1/Vps34 signaling enhances exosome release since beclin-1 knockdown and spautin-1 treatment, not rapamycin, prevents exosome release. Dasatinib induces p-p53 Ser15 /active caspase 3 expression and thus promotes apoptosis by inhibiting the mTOR/p70S6K Thr389 activity via downregulation of Akt Ser473 activation, and decreases exosomal release by inhibiting autophagy activity via downregulation of beclin-1 and Vps34. GSI: Notch1 γ-secretase inhibitor; Spautin-1: autophagy inhibitor; Rapamycin: mTOR inhibitor; si-Beclin-1: knockdown of beclin-1 by siRNA. Red circle: phosphorylation status.

Cell Culture and Treatment
The CML cell line K562 (ATCC, Manassas, VA, USA) was cultured at 37˝C in DMEM containing 10% FBS (Gibco, Waltham, MA, USA) and 100 U/mL of Penicillin/Streptomycin. To make imatinib-resistant CML cell line, K562 was continuously exposed to higher concentrations of imatinib mesylate (IMT; Sigma-Aldrich, St. Louis, MO, USA) in stepwise increase of 100 nM from 0.1 to 1 µM after 7 days of culture [43]. Finally, the viable imatinib-resistant cells (K562R IMT ) were maintained in culture media containing 1 µM of imatinib. To evaluate the effects of dasatinib on cellular apoptosis and exosomes release, K562R IMT cells were treated with the indicated dosage of dasatinib (Sigma-Aldrich, St. Louis, MO, USA) for 24 h in the absence of imatinib. As Okabe et al. [45] reported, our preliminary results demonstrated that removal of imatinib for 24 hours showed no effects on cleaved caspase 3 and Akt activation at Ser473 in imatinib-resistant K562 cells.

Exosome Isolation
Culture media were ultra-centrifuged at 100,000ˆg for 16 h at 4˝C to remove exosomes that may be possibly present in media. K562 or K562R IMT cells were cultured in a 150-cm 2 flask for 24 h in ultra-centrifuged media, and then the cultured media were collected for exosome purification as described previously [12]. Briefly, the media were centrifuged progressively at 300ˆg for 10 min, 2000ˆg for 10 min, and then filtered through a 0.22 µm filter. Effluent was ultra-centrifuged at 100,000ˆg for 2 h. The exosomal pellets from 5 replicate experiments were resuspended in PBS, and then the content of exosomal protein was determined using BCA assay (Thermo Scientific, Waltham, MA, USA).

Western Blot
Total cellular protein was extracted with RIPA buffer containing freshly added protease and phosphotase inhibitor cocktail (Roche, Indianapolis, IN, USA). Nuclear protein was extracted using the Nuclear Extraction Kit (Abcam). Protein concentration was determined using BCA assay. Equal amounts of protein were separated on a 7.5% or 12% SDS gel and electrophoretically transferred to a nitrocellulose membrane. Membranes were blocked with 5% non-fat milk or BSA in Tris-buffered saline containing 0.05% Tween-20 (TTBS) for 1 h. The indicated primary antibody was incubated overnight at 4˝C. After washing three times with TTBS, membranes were incubated with HRP-conjugated goat anti-rabbit or mouse antibody (1:10,000; Thermo Scientific) for 1 h, and then developed with the enhanced chemiluminescence reagent (Thermo Scientific). The specific band was quantified using Image J software (NIH).

Rheb Activity Detection
The active Rheb-GTP level was measured with Rheb Activation Assay Kit (Abcam). Briefly, K562 or K562R IMT cells were lysed with 1ˆAssay/Lysis Buffer supplemented with protease inhibitors (Roche). The anti-active Rheb mouse monoclonal antibody was incubated with 250 µg of cellular lysates (adjusted volume to 500 µL with 1ˆAssay/Lysis Buffer) at 4˝C overnight. The bound activated Rheb-GTP was then pulled down by incubating for 1 h with 25 µL of Protein A/G agarose at 4˝C. The agarose beads were washed 5 times by centrifugation (5000ˆg, 1 min) with 500 µL of 1ˆAssay/Lysis Buffer. The beads were then resuspended in 25 µL of 2ˆLamelli loading buffer and boiled for 5 min. The precipitated active Rheb-GTP was detected by Western blot using rabbit anti-Rheb polyclonal antibody. Notably, the same amount of lysates from K562R IMT cells were treated with 0.5 M EDTA (final 20 mM) and 100ˆGDP (final 1 mM; negative control) or 100ˆGTPγs (final 1 mM; positive control) for 30 min with agitation at room temperature, and then the reactions were stopped by putting the tubes on ice and adding 1 M MgCl 2 (final 60 mM).

Cellular Proliferation Assay
Cell proliferation was measured using MTT Cell Proliferation Assay Kit according to the manufacture's protocol (BioVision, Rockland, MA, USA).

Apoptosis Detection Assay
The cleaved caspase 3 was evaluated in live cells by using the APO ACTIVE 3™ Kit (Cell Technology, Fremont, CA, USA) as previously described [47]. Briefly, 1ˆ10 6 cells were fixed by incubation in 500 µL of 1ˆfixative solution at room temperature for 15 min. After washing twice with PBS, the cells were resuspended in 1 mL of 1% Saponin/PBS and 100 µL of cells were pipetted out into a 2-mL tube. After adding 10 µL of the 1ˆrabbit anti active caspase 3 antibody, they were incubated for 1 h at room temperature. After washing twice with 1% saponin/PBS, 10 µL of the 1ˆFITC labeled goat anti rabbit IgG was added and then the contents was incubated for 1 h at room temperature. After washing once with 1% Saponin/PBS and 2% BSA/PBS respectively, the cells were resuspended in 500 µL of 2% BSA/PBS. The caspase 3 positive cells were counted with flow cytometry (FACScan, BD, San Jose, CA, USA). Additionally, the level of the activated caspase 3 was also evaluated by using Western blotting with the specific anti-activated caspase 3 antibody.

Statistics
Data are shown as mean˘SD. Statistical analysis was performed with Prism 4 (GraphPad Software, La Jolla, CA, USA) using unpaired test or ONE-WAY ANOVA followed by Tukey's multiple comparisons test. p ď 0.05 was regarded as significant differences.
Author Contributions: Juan Liu and Yujing Zhang performed all the experiments, analyzed the data, and drafted the manuscript. Jinghua Wang, Lianqiao Li, Xi Chen, and Xinyu Gao carried out the partial study design and the manuscript revision. Yanming Xue, Xiaomin Zhang and Yao Liu participated in partial experiments and data analysis. Juan Liu and Aichun Liu conceived and designed the experiments, and performed the whole revision process.

Conflicts of Interest:
The authors declare no conflict of interest.