Alpha-tocopherol attenuates the anti-tumor activity of crizotinib against cells transformed by NPM-ALK

Anaplastic large cell lymphomas (ALCL) are mainly characterized by harboring the fusion protein nucleophosmin-anaplastic lymphoma kinase (NPM-ALK). The ALK inhibitor, crizotinib specifically induced apoptosis in Ba/F3 cells expressing NPM-ALK by inhibiting the activation of NPM-ALK and its downstream molecule, signal transducer and activator of transcription factor 3 (STAT3). We found that α-tocopherol, a major component of vitamin E, attenuated the effects of crizotinib independently of its anti-oxidant properties. Although α-tocopherol suppressed the inhibitory effects of crizotinib on the signaling axis including NPM-ALK and STAT3, it had no influence on the intake of crizotinib into cells. Crizotinib also directly inhibited the kinase activity of NPM-ALK; however, this inhibitory effect was not altered by the co-treatment with α-tocopherol. Whereas the nuclear localization of NPM-ALK was disappeared by the treatment with crizotinib, the co-treatment with α-tocopherol swept the effect of crizotinib and caused the localization of NPM-ALK in nucleus. The administration of α-tocopherol attenuated the anti-tumor activity of crizotinib against NPM-ALK-provoked tumorigenesis in vivo. Furthermore, the α-tocopherol-induced inhibition of crizotinib-caused apoptosis was also observed in NPM-ALK-positive cells derived from ALCL patients, namely, SUDHL-1 and Ki-JK. Collectively, these results not only revealed the novel mechanism underlying crizotinib-induced apoptosis in NPM-ALK-positive cells, but also suggest that the anti-tumor effects of crizotinib are attenuated when it is taken in combination with vitamin E.


Introduction
Anaplastic lymphoma kinase (ALK) is a receptor-type protein tyrosine kinase that belongs to the insulin-receptor superfamily [1]. ALK shows high sequence similarity to receptor tyrosine PLOS

Plasmids
The cDNA encoding NPM-ALK harboring Flag tag on its N terminus was inserted into the MSCV-Puro retroviral vector. The mutagenesis of amino acid residues in NPM-ALK (K210R) was performed using a site-directed mutagenesis kit according to the manufacturer's instructions (

Measurement of intracellular ROS generation
The accumulation of intracellular ROS was detected using 2', 7'-dichlorodihydrofluorescein diacetate (DCFH-DA) (Cayman Chemical, Ann Arbor, MI, USA), which was hydrolyzed by a cellular esterase to 2', 7'-dichlorodihydrofluorescein (DCFH) and then oxidized to the fluorescent compound 2', 7'-dichlorofluorescein (DCF). Cells were incubated with PBS containing DCFH-DA (10 μM) at 37˚C for 1 hr, and then washed with PBS. Cells were treated with crizotinib or pyrrolidinium fullerene in combination with α-tocopherol for 24 hr and washed with PBS. The fluorescence intensity of oxidized DCF was monitored using FACS Calibur with the CELL Quest program as previously described [14].

Cell cycle analysis
Cells were fixed with 70% (v/v) ethanol at -20˚C overnight. They were then centrifuged at 5,000 r.p.m at 4˚C for 2 min and treated with PBS containing 10 μg/ml RNase A (Nacalai Tesque). After the addition of 100 μg/ml propidium iodide (PI) (Wako Pure Chemical Industries, Tokyo, Japan), cell cycle parameters were assessed by a flow cytometric analysis using FACS Calibur as described previously [14].

DNA fragmentation assay
Genomic DNA was prepared for gel electrophoresis as described previously [15]. Electrophoresis was performed on a 1% (w/v) agarose gel in Tris/boric acid buffer. Fragmented DNA was visualized by staining with ethidium bromide after electrophoresis.

Immunoblot analysis
Cells were washed with PBS and lysed in NP-40 lysis buffer (50 mM Tris-HCl pH 7.4, 10% glycerol, 50 mM NaCl, 0.5% sodium deoxycholate, 0.5% NP-40, 20 mM NaF, and 0.2 mM Na 3 VO 4 ) supplemented with protease inhibitors. Cell lysates were cleared by centrifugation at 15,000 r.p.m at 4˚C for 15 min and proteins were then denatured with Laemmli buffer. For preparation of cytosolic fraction and nuclear fraction, cells were disrupted in Buffer A (10 mM Hepes-KOH (pH7.8), 10 mM KCl, 0.1 mM EDTA (pH8.0), 0.1% NP-40) and were centrifuged at 5,000 r.p.m for 15 min at 4˚C. The supernatant was collected and used as cytosolic fraction. Isolated nuclei were lysed in Nonidet P-40 lysis buffer and homogenized using the ultrasonic homogenizer VP-50 (TAITEC, Japan), and then centrifuged at 15,000 r.p.m at 4˚C for 15 min in order to remove debris. Cytosolic fraction and nuclear fraction were denatured with Laemmli's sample buffer. Denatured proteins were resolved by SDS-PAGE and transferred onto PVDF membranes (Millipore, Billerica, MA). Membranes were probed using the designated antibodies and visualized with the ECL detection system (GE Healthcare UK., Ltd.). The intensity of each band was quantified by Image-J software. The phosphorylation levels of STAT3 and STAT5 were normalized with the expression levels of STAT3 and STAT5. To show the relative amounts of NPM-ALK in cytosol and nucleus, the band intensity of NPM-ALK were normalized with band intensities of β-tublin and Lamin B, respectively.

Measurement of crizotinib uptake by LC-MS
Ba/F3 cells expressing NPM-ALK (9×10 4 cells/200 μL) were cultured with crizotinib (0.5 μM) in combination with α-tocopherol (25 μM) for 60 or 120 min and washed with ice-cold PBS, and then lysed with 125 μL of 50% acetonitrile. Samples were centrifuged at 5,000 r.p.m at 4˚C for 2 min and the supernatant was analyzed by an Agilent 6120 quadrupole mass spectrometer equipped with an electrospray interface attached to an Agilent 1200 series (Agilent Technologies, Santa Clara, CA, USA). Chromatographic separations were performed on Agilent ZOR-BAX Eclipse Plus C18 column (4.6 × 100 mm, 3.5 μm) at a flow rate 0.5 mL/min. The mobile phases used were water with 0.1% formic acid (A) and acetonitrile with 0.1% formic acid (B). At time zero the flow consisted of 90% of mobile phase A and 10% mobile phase B. One minute after injection, the proportion of B was linearly increased to 100% over 4 min followed by keeping constant at 100%B for 5 min. From this point, the mobile phase was set to initial conditions (90%A and 10%B) and the column was equilibrated for 5 min prior to the next injection. The electrospray ionization probe was set at 350˚C, with the nebulizing gas pressure and electrospray gas flow set at 20 psi. and 13 L/min, respectively. Detection of crizotinib was carried out by selected ion monitoring of [M+H] + ions at m/z 450 in the positive ion mode.
In vitro kinase assay NPM-ALK autophosphorylation was measured by modified methods as described previously [16]. Cells were lysed in NP-40 lysis buffer supplemented with protease inhibitors and NPM-ALK was immunoprecipitated using the anti-Flag antibody and protein G-sepharose

Transplantation of tumor cells into nude mice
Six-week-old female BALB/c nude mice were subcutaneously injected with transduced Ba/F3 cells (2×10 7 cells). In order to investigate oncogenic potentials in vivo, tumor, liver, and lymph node weights were analyzed 16 days after transplantation. Regarding the administration of crizotinib and α-tocopherol, nude mice were randomized into three groups; vehicle, crizotinib, and crizotinib plus α-tocopherol. Crizotinib (1 mg) and/or α-tocopherol (4 mg) were dissolved in 200 μL olive oil, and orally administered to nude mice for 7 consecutive days after transplantation. Mice were then sacrificed by an overdose of isoflurane. All experimental protocols were approved by the Animal Usage Committee of Keio University (Approval number, 15029-(0)). The methods were carried out in accordance with the approved guidelines.

Statistical analysis
Data are expressed as average ± SD for in vitro and ± SEM for in vivo experiments. Statistical analyses were conducted using SPSS Statistics software (Version 23 for Macintosh, IBM Inc). A one-or two-way analysis of variance (ANOVA) followed by Tukey's test was used to evaluate differences between more than three groups. Differences were considered to be significant for values of P<0.05.

Crizotinib specifically induced apoptosis in Ba/F3 cells transformed by NPM-ALK
Similar to NPM-ALK, the oncogenic tyrosine kinase, TEL-JAK2 is generated from reciprocal chromosomal translocations and causes lymphoma and leukemia [17][18][19]. In order to examine the specificity of the drug efficacy of crizotinib against tumor-related tyrosine kinases, murine hematopoietic Ba/F3 cells were infected with an empty retrovirus (-) and retroviruses harboring NPM-ALK and TEL-JAK2, respectively. NPM-ALK induced the phosphorylation of STAT3, whereas TEL-JAK2 induced the phosphorylation of STAT5 (Fig 1A). NPM-ALK and TEL-JAK2 both induced the cytokine-independent proliferation of Ba/F3 cells (Fig 1B). We then investigated the effects of the ALK inhibitor, crizotinib on the viability of Ba/F3 cells transformed by NPM-ALK and TEL-JAK2. As shown in Fig 1C, crizotinib specifically decreased the viability of Ba/F3 cells expressing NPM-ALK in a dose-dependent manner, while Ba/F3 cells expressing TEL-JAK2 appeared to exhibit resistance against the treatment with crizotinib. On the other hand, mitomycin C (MMC), a DNA crosslinking anti-cancer drug, significantly decreased the viabilities of Ba/F3 cells expressing NPM-ALK and TEL-JAK2 (Fig 1C, right). The treatment of Ba/F3 cells expressing NPM-ALK with crizotinib resulted in the accumulation of cells in the sub-G1 phase, which is well established as a characteristic of apoptotic cell death; however, this was not observed in cells expressing TEL-JAK2. In contrast, MMC significantly increased the percentages of both cells in the sub-G1 phase ( Fig 1D). To confirm that crizotinib induces the apoptotic cell death, we investigated whether internucleosomal DNA fragmentation, a biochemical characteristic of apoptosis, could be detected in the crizotinib-treated cells. As shown in Fig 1E, the treatment with crizotinib clearly induced the ladder pattern of DNA internucleosomal fragmentation in Ba/F3 cells expressing NPM-ALK but not Ba/F3 cells expressing TEL-JAK2. On the other hand, MMC induced DNA internucleosomal fragmentation in both cells ( Fig 1E). The treatment with crizotinib inhibited STAT3 phosphorylation in a dose-dependent manner in Ba/F3 cells expressing NPM-ALK, but had no effect on STAT5 phosphorylation in Ba/F3 cells expressing TEL-JAK2 ( Fig 1F). These results suggest that crizotinib specifically induced apoptosis in cells transformed by NPM-ALK.

α-Tocopherol rescued crizotinib-induced apoptosis in cells transformed by NPM-ALK in a ROS-independent manner
A previous study reported that crizotinib increases intracellular ROS levels in human alveolar rhabdomyosarcoma and embryonal rhabdomyosarcoma cells [13]. Therefore, we evaluated the generation of intracellular ROS by measuring oxidized DCFH-DA (2 0 ,7 0 -dichlorodihydrofluorescein-DA), and tested whether crizotinib induces apoptosis in Ba/F3 cells expressing NPM-ALK through the generation of ROS. In a previous study, pyrrolidinium fullerene caused apoptotic cell death via the generation of ROS in HL60 cells [14]; therefore, we utilized it as a positive control for ROS generation. As shown in Fig 2A, while the treatment with pyrrolidinium fullerene drastically enhanced the generation of ROS in Ba/F3 cells expressing NPM-ALK, the treatment with crizotinib slightly enhanced the generation of ROS. In addition, the co-treatment with two kinds of anti-oxidants, α-tocopherol and edaravone significantly suppressed pyrrolidinium fullerene and crizotinib-induced ROS generation in a dosedependent manner (Fig 2A and 2B). We then investigated the effects of these anti-oxidants on the viability of Ba/F3 cells expressing NPM-ALK, which were treated with crizotinib or pyrrolidinium fullerene. Similar to the results shown in Fig 1, crizotinib markedly reduced the viability of Ba/F3 cells expressing NPM-ALK; however, α-tocopherol, but not edaravone significantly recovered the viability of crizotinib-treated cells (Fig 2C and 2D). On the other hand, the pyrrolidinium fullerene-induced reduction in cell viability appeared to be canceled by the co-treatment with α-tocopherol and edaravone (Fig 2C and 2D). These results clearly suggest that crizotinib-induced cell death and its cancelation by α-tocopherol were the most unlikely to depend on intracellular ROS levels. We subsequently examined the effects of antioxidants on the crizotinib and pyrrolidinium fullerene-induced accumulation of sub-G1 phase cells in Ba/F3 cells expressing NPM-ALK. As shown in Fig 2E and 2F, α-tocopherol, but not edaravone consistently reduced increases in the accumulation of sub-G1 phase in Ba/F3 cells expressing NPM-ALK induced by crizotinib. On the other hand, the pyrrolidinium fullereneinduced accumulation of sub-G1 phase cells was reduced by the treatment with α-tocopherol and edaravone. In addition, the crizotinib-induced internucleosomal DNA fragmentation was prevented by the co-treatment with α-tocopherol ( Fig 2G). Collectively, these results strongly suggest that ROS are not critical for crizotinib-induced apoptosis in cells expressing NPM-ALK. They also indicate that α-tocopherol rescued crizotinib-induced apoptosis in cells transformed by NPM-ALK regardless of its ability to scavenge ROS.

α-Tocopherol indirectly attenuated the inhibition of NPM-ALK phosphorylation by crizotinib
In order to elucidate the mechanisms by which α-tocopherol specifically attenuates the effects of crizotinib, we examined whether the uptake of crizotinib into Ba/F3 cells expressing NPM-ALK was affected by α-tocopherol. After the treatment with crizotinib in the absence and presence of α-tocopherol, the intracellular concentration of crizotinib was measured by LC-MS. However, similar amounts of crizotinib were taken up by cells regardless of the cotreatment with α-tocopherol ( Fig 4A). We then investigated whether α-tocopherol attenuated crizotinib-induced inhibitory effects on kinase activity by NPM-ALK. Cells expressing NPM-ALK were treated with crizotinib in the presence or absence of α-tocopherol, and NPM-ALK was then immunoprecipitated with the anti-Flag antibody in order to evaluate its kinase activity using an in vitro kinase assay with radioactive ATP. The autophosphorylation of NPM-ALK was completely suppressed by crizotinib, and this crizotinib-induced inhibition was canceled by the co-treatment with α-tocopherol in a dose-dependent manner (Fig 4B). Furthermore, the treatment with α-tocopherol markedly attenuated the crizotinib-induced inhibition of STAT3 phosphorylation (Fig 4C). In an attempt to gain further insights into the mechanisms by which α-tocopherol suppressed crizotinib-induced inhibitory effects on NPM-ALK, we tested whether α-tocopherol canceled the inhibition of NPM-ALK by crizotinib in vitro. Immunoprecipitated NPM-ALK was treated with crizotinib in the presence or absence of α-tocopherol. As shown in Fig 4D, although crizotinib completely inhibited the phosphorylation of immunoprecipitated NPM-ALK, the co-treatment with α-tocopherol failed to attenuate the inhibitory effects of crizotinib. These results suggest that the cancelation of the effects of crizotinib by α-tocopherol was the most unlikely to be due to the direct effects of α-tocopherol on the NPM-ALK protein.
The co-treatment with α-tocopherol swept the effect of crizotinib on the subcellular localization of NPM-ALK Ba/F3 cells were infected with an empty retrovirus (-) and retroviruses harboring NPM-ALK and its kinase dead mutant, in which the lysine residue at 210 in the ATP-binding site was mutated to arginine (K210R). As shown in Fig 5A and 5B, while NPM-ALK induced the phosphorylation of STAT3 and cytokine-independent proliferation, NPM-ALK (K210R) failed to induce STAT3 activation and confer cytokine independence to Ba/F3 cells. It has been reported that NPM-ALK was localized both in cytosol and nucleus [5,[21][22][23]. We investigated the subcellular localization of NPM-ALK and NPM-ALK (K210R). Interestingly, whereas NPM-ALK was localized in both cytosol and nucleus in Ba/F3 cells, the nuclear localization of NPM-ALK (K210R) was completely disappeared (Fig 5C), suggesting that kinase activity of NPM-ALK affected its cellular localization. Strikingly, the nuclear localization of NPM-ALK was clearly disappeared by the treatment with crizotinib (Fig 5D,  α-Tocopherol significantly attenuated the anti-tumor activity of crizotinib in nude mice transplanted with cells expressing NPM-ALK In order to evaluate the anti-tumor effects of crizotinib and its cancellation by α-tocopherol, we attempted to establish an assay system by subcutaneously (s.c.) inoculating transduced Ba/ F3 cells expressing NPM-ALK into nude mice. As shown in Fig 6A, whereas Ba/F3 cells expressing NPM-ALK exhibited significant tumor-forming activity, tumor formation was not observed in nude mice inoculated with Ba/F3 cells expressing NPM-ALK (K210R). In these nude mice, abnormally enlarged livers and lymph nodes were observed when mice were transplanted with Ba/F3 cells expressing NPM-ALK, but not NPM-ALK (K210R), suggesting that the kinase activity of NPM-ALK is required for tumorigenesis ( Fig 6A). We then tested the in was examined in a flow cytometric analysis. (A-F) Values are given as the mean ± SD of three independent experiments. **P < 0.01; *P < 0.05 significantly different from the control group; ## P < 0.01; # P < 0.05 significantly different from the group incubated with 0.5 μM crizotinib or 25 μM pyrrolidinium fullerene. (G) To evaluate the apoptotic cell death, the chromatin DNA was isolated from cells and subjected to agarose gel electrophoresis.
https://doi.org/10.1371/journal.pone.0183003.g002  Nude mice were orally administered crizotinib (1 mg/mouse) and α-tocopherol (4 mg/mouse) for 7 consecutive days after the transplantation of Ba/F3 cells expressing NPM-ALK. The dosages of crizotinib and α-tocopherol for the treatment of mice were established by correlative calculations according to the effective dose in human patients with NSCLC and tolerable upper intake levels (ULs) for adults aged 19 years and older [24][25][26]. In contrast to mice treated with vehicle, tumor formation and enlargements in the liver and lymph nodes were effectively suppressed by the administration of crizotinib. On the other hand, the simultaneous administration of crizotinib and α-tocopherol significantly attenuated the inhibitory effects of crizotinib on tumor formation and enlargements in the liver and lymph nodes in mice receiving Ba/F3 cells expressing NPM-ALK (Fig 6B). In addition, liver sections were prepared and stained with hematoxylin and eosin. The density of cells was significantly greater, and the arrangement of hepatocytes was disrupted more in the livers of mice inoculated with Ba/F3 cells expressing NPM-ALK than in the livers of mice inoculated with control Ba/F3 cells and Ba/F3 cells expressing the kinase dead mutant of NPM-ALK (K210R) (Fig 6C), suggesting that these differences were due to the infiltration of NPM-ALK-induced tumor cells. We then attempted to investigate the effects of crizotinib and α-tocopherol on these abnormalities in the livers of mice transplanted with Ba/F3 expressing NPM-ALK. The density of cells and disarrangement of hepatocytes were significantly less with the treatment with crizotinib than with the vehicle. However, in the livers of mice simultaneously administered α-tocopherol, abnormalities such as the enhanced density and disarrangement of hepatocytes were similar to those in the livers of mice treated with vehicle ( Fig 6D). Taken together, our results strongly support crizotinib being useful for the treatment of NPM-ALK-related tumors; however, its therapeutic effects will be abrogated by a co-treatment with α-tocopherol.

α-Tocopherol significantly attenuated the effect of crizotinib on NPM-ALK-positive cells derived from human ALCL patients
Ki-JK and SUDHL-1 cell lines were derived from ALCL patients [27,28], which have a t(2;5) (p23;q35) translocation producing NPM-ALK. We examined whether α-tocopherol suppresses the inhibitory effects of crizotinib on the proliferation of Ki-JK cells and SUDHL-1 cells. These cell lines were treated with crizotinib in the presence and absence of α-tocopherol, and intracellular ROS levels, cell viability, and the percentages of cells in the sub-G1 phase were then evaluated. As shown in Fig 7A, the amounts of intracellular ROS were slightly reduced when these two cell lines were treated with a high dose of α-tocopherol. On the other hand, the treatment with crizotinib reduced intracellular ROS levels slightly in Ki-JK cells and markedly in SUDHL-1 cells; however, α-tocopherol did not have any effects on crizotinibinduced reductions in ROS levels in these cells (Fig 7A). Although crizotinib markedly antibody (bottom). The relative phosphorylation level of NPM-ALK is shown in the graph. Values are given as the mean ± SD of three independent experiments. ***P < 0.001 significantly different from the control group; ### P < 0.001 significantly different from the group incubated with 0.5 μM crizotinib. (C) Cell lysates were immunoblotted with an antiphospho-STAT3 antibody (Tyr705), anti-STAT3 antibody, or anti-β-actin antibody. The relative phosphorylation level of STAT3 is shown in the graph. Values are given as the mean ± SD of three independent experiments. ***P < 0.001 significantly different from the control group; # P < 0.05 significantly different from the group incubated with 0.5 μM crizotinib. (D) Ba/F3 cells expressing NPM-ALK were lysed and the NPM-ALK protein was immunoprecipitated with an anti-Flag antibody. Immunoprecipitated NPM-ALK was incubated with crizotinib (0.5 μM) in combination with αtocopherol (6.25, 25, 100 μM) for 15 min, and then assayed for kinase activity. Phosphorylated NPM-ALK was detected by autoradiography (upper). The immunoprecipitates were immunoblotted with anti-Flag antibody (bottom). The relative phosphorylation level of NPM-ALK is shown in the graph. Values are given as the mean ± SD of three independent experiments. ***P < 0.001 significantly different from the control group. https://doi.org/10.1371/journal.pone.0183003.g004 Vitamin E inhibits the anti-tumor activty of crizotinib decreased the viabilities of Ki-JK cells and SUDHL-1 cells, α-tocopherol significantly rescued these cells from crizotinib-induced cytotoxicity in a dose-dependent manner (Fig 7B). In addition, crizotinib significantly induced the accumulation of sub-G1 phase cells in these cell lines; however, this effect was canceled by the treatment with α-tocopherol (Fig 7C). Although crizotinib inhibited the phosphorylation of STAT3, the treatment with α-tocopherol markedly attenuated the crizotinib-induced inhibition of STAT3 phosphorylation in both Ki-JK cells and SUDHL-1 cells. Furthermore, crizotinib induced the activation of caspase 3, the treatment with α-tocopherol markedly attenuated the crizotinib-induced activation of caspase 3 in both Ki-JK cells and SUDHL-1 cells (Fig 7D), suggesting that crizotinib-induced apoptosis was canceled by the treatment with α-tocopherol. These results demonstrate that the anti-tumor activity of crizotinib against NPM-ALK is also canceled by α-tocopherol in cells derived from human patients with ALCL.

Discussion
The results of the present study strongly support the usefulness of crizotinib in the treatment of NPM-ALK-related tumors. Crizotinib is an orally available ALK inhibitor that inhibits ALK by competitively binding within the ATP-binding pocket, and also inhibits receptor tyrosine kinases including c-Met and ROS1 [9,10]. We demonstrated that the simultaneous administration of α-tocopherol markedly inhibited the therapeutic efficacy of crizotinib; however, we were unable to elucidate the precise mechanism by which α-tocopherol abrogates the inhibitory effects of crizotinib on NPM-ALK.
In the present study, we demonstrated that other anti-oxidants including edaravone, βtocopherol, δ-tocopherol, γ-tocopherol, α-tocotrienol, and Trolox failed to cancel the effects of crizotinib (Figs 1 and 3), suggesting that α-tocopherol inhibits the anti-tumor effects of crizotinib in a ROS level-independent manner. It has been shown that α-tocopherol negatively regulates the proliferation of vascular smooth muscle cells and aggregation of platelets by inhibiting PKC independently of its antioxidant properties [29]. Previous findings demonstrated that the α-tocopherol-induced inhibition of PKC was suppressed by a co-treatment with β-tocopherol when administered at a similar concentration to that of α-tocopherol [30]. We examined the effects of a co-treatment with α-tocopherol and β-tocopherol on cytotoxicity induced by crizotinib in Ba/F3 cells expressing NPM-ALK. The results obtained revealed that β-tocopherol failed to buffer the inhibitory effects of α-tocopherol on crizotinib-induced cytotoxicity (S1 Fig). A previous study reported that α-tocopherol specifically binds to the αtocopherol transfer protein (TTP) and is involved in the intracellular trafficking of α-tocopherol [31]. α-Tocopherol has also been reported to activate p38 MAP kinase and its downstream transcription factor, MITF in order to positively regulate the function of osteoclasts in bone tissue [32]. In an attempt to clarify the functional involvement of p38 and TTP in the effects of α-tocopherol, we utilized the p38 inhibitor, SB203580, and siRNA against TTP. Although the treatment with SB203580 reduced the viability of Ba/F3 cells expressing NPM-ALK, the αtocopherol-induced cancelation of cytotoxicity induced by crizotinib was not affected by Flag antibody, anti-Lamin B or anti-β-tublin. The relative expression levels of NPM-ALK and NPM-ALK (K210R) in cytosol and nucleus are shown in the graphs. Values are given as the mean ± SD of three independent experiments. ***P < 0.01 significantly different from the group of NPM-ALK. (D) Ba/F3 cells expressing NPM-ALK were treated with crizotinib (0.5 μM) or in combination with α-tocopherol (25 μM) or β-tocopherol (25 μM) for 24 hr. Cytosol fraction and nuclear fraction were prepared and immunoblotted with anti-Flag antibody, anti-Lamin B or anti-β-tublin. The relative expression levels of NPM-ALK in cytosol and nucleus are shown in the graphs. Values are given as the mean ± SD of three independent experiments. ***P < 0.01 significantly different from the control group; ### P < 0.001 significantly different from the group incubated with 0.5 μM crizotinib.
https://doi.org/10.1371/journal.pone.0183003.g005 Interestingly, higher levels of α-tocopherol were needed for rescuing STAT3 phosphorylation in human ALCL cell lines compared to Ba/F3 cells expressing NPM-ALK (Figs 4C and  7D). Although IC 50 of crizotinib against c-Met and ROS is higher than against ALK, crizotinib inhibits not only ALK, but also c-Met and ROS [9,10]. Therefore, it is suggested that other tyrosine kinases which induces phosphorylation of STAT3 could be involved in the oncogenicity of ALCL-derived cell lines. Indeed, it was reported that the phosphorylation of STAT3 was not completely inhibited by the knockdown of NPM-ALK using siRNA in SUDHL-1 cells [33]. Although speculative, the involvement of other tyrosine kinases such as c-Met and ROS might be a reason for the requirement of high dosage of α-tocopherol for rescuing STAT3 phosphorylation in SULH-1 cells and Ki-JK cells. Furthermore, whereas crizotinib slightly induced ROS generation in Ba/F3 cells, intracellular ROS levels was decreased by the treatment with crizotinib in human ALCL cell lines (Figs 2B and 7A). Since it was reported that c-Met harbors the ability to control the ROS generation and contributes in the oncogenicity of ALCL-derived cell lines [34,35], it is also suggested that the involvement of c-Met could be a reason for the different responses against the treatment of crizotinib in between ALCL-derived cell lines and Ba/F3 cells.
As shown in Fig 5, NPM-ALK was localized in the cytosol and the nucleus as previously reported [5,21,22]. Remarkably, the nuclear localization of NPM-ALK required its kinase activity, and crizotinib effectively suppressed the nuclear localization of NPM-ALK. Wild type NPM is mainly localized in nucleoli, however some population shuttled between nucleolus, nucleoplasm, and cytosol [23]. Since NPM forms mono-oligomer, it is not surprising that NPM-ALK could interact with NPM. The localization of NPM-ALK is most likely regulated by fused-NPM region and interacted endogenous NPM, therefore partial population of NPM-ALK was detected in nucleolus. Several groups reported that the nucleolar localization of NPM-ALK was not required for the cellular transformation provoked by NPM-ALK [5,21]. N-terminal NPM of NPM-ALK could be replaced with the portion of the unrelated translocated promoter region (TPR) protein that activates the TPR-MET fusion kinase by mediating dimerization through its leucine zipper motif [5]. Whereas NPM-ALK was localized in nucleolus, nucleoplasm, and cytoplasm, the artificial fusion protein TPR-ALK was detectable in only cytoplasm [5,21]. In spite of its different localization, TPR-ALK exhibited the comparable transforming activity with NPM-ALK. However, their reports failed to contradict the possibility that undetectable small population of TPR-ALK could be localized in nucleus, and contribute to the cellular transformation. N-terminal NPM region of NPM-ALK contains nuclear export sequence (NES), however lacks nuclear localization sequence (NLS), suggesting that nuclear localization of NPM-ALK should be triggered by interacting protein of NPM-ALK. Although we do not have any direct evidences supporting the possible contribution of nuclear NPM-ALK in Ba/F3 cells to cellular transformation, it will be important to identify the interacting proteins with nuclear NPM-ALK to elucidate the mechanism how NPM-ALK induces the cellular transformation. In addition, the mutation analysis of NES of NPM-ALK will be the clue for the investigation for the function of nuclear NPM-ALK. NPM-ALK; n.d., not detected. (B) Nude mice were transplanted with Ba/F3 cells expressing NPM-ALK and orally administered crizotinib (1 mg/mouse) in combination with/without α-tocopherol (4 mg/mouse) for 7 consecutive days. Tumor, liver, and lymph node weights were measured 16 days post-inoculation. Values are given as the mean ± SEM of three independent experiments. **P < 0.01; *P < 0.05 significantly different from the vehicle-treated group; # P < 0.05 significantly different from the crizotinib-treated group. (C, D) Sixteen days post-inoculation, liver sections were stained with hematoxylin and eosin (magnification: ×400). https://doi.org/10.1371/journal.pone.0183003.g006 Vitamin E inhibits the anti-tumor activty of crizotinib Clinical resistance mutations of EML4-ALK to crizotinib including L1196M, C1156Y, G1269A, L1152R, G1202R, F1174C, I1171T, and S1206Y have been reported in patients with NSCLC [3,36]. Although we have no direct evidence, the treatment with α-tocopherol may have caused structural alterations in NPM-ALK, similar to the NPM-ALK mutants exhibiting resistance to crizotinib. Second generation ALK inhibitors such as alectinib have recently been developed to overcome acquired resistance to crizotinib [37]. We found that α-tocopherol failed to attenuate cytotoxicity by alectinib against Ba/F3 cells expressing NPM-ALK (S3 Fig). In addition, although we investigated the effects of α-tocopherol on the JAK2 inhibitor, ruxolitinib-induced apoptosis in Ba/F3 cells transformed by a constitutive active mutant of JAK2 (V617F), α-tocopherol failed to rescue ruxolitinib-induced apoptosis in these cells (S3 Fig). Therefore, these results suggest that the inhibitory effects of α-tocopherol on crizotinib depend on its chemical structure.
However, it remains unclear whether the α-tocopherol-mediated inactivation of crizotinib occurs enzymatically or is mediated by non-enzymatic steps.
Vitamin E, which is composed of tocopherols and tocotrienols, is one of the most popular supplements in the world; more than 10% of adults take at least 400 IU of vitamin E daily [38]. α-Tocopherol is the major component of vitamin E and is preferentially absorbed by and accumulates in the human body. Although a number of issues have yet to be clarified, our results indicate that the anti-tumor activities of well-known compounds may be abrogated by environmental factors, such as the intake of supplements, in patients with tumor-related diseases. (B) After 48 hr, total RNA was extracted and RT was performed using an oligo (dT) 20 primer. Quantitative real-time PCR was performed using an iCycler detection system (Bio-Rad, Berkeley, CA, USA). GAPDH mRNA was analyzed as an internal control. Values are the mean ± S. D. of three independent experiments. Ã P < 0.05 (C) After 48 hr, transfected cells were treated with crizotinib (0.5 μM) in combination with α-tocopherol (6.25, 25, 100 μM) for 24 hr. Cell viabilities were assessed by a WST assay. Values are given as the mean ± SD of four Analytical values are given as the mean ± SD of three independent experiments. **P < 0.01; *P < 0.05 significantly different from the control group; ## P < 0.01; # P < 0.05 significantly different from the group incubated with 0.5 μM crizotinib. (D) Cell lysates were immunoblotted with an anti-phospho-STAT3 antibody (Tyr705), anti-STAT3 antibody, anti-cleaved caspase 3 antibody or anti-β-actin antibody. The relative phosphorylation level of STAT3 and the relative expression level of cleaved caspase 3 are shown in the graphs. Values are given as the mean ± SD of three independent experiments. ***P < 0.001 significantly different from the control group; ### P < 0.001; # # P < 0.01 significantly different from the group incubated with 0.5 μM crizotinib.  1 μM) in combination with α-tocopherol (6.25, 25, and 100 μM) for 24 hr. Cell viabilities were evaluated by a WST assay. Values are given as the mean ± SD of four independent experiments. ÃÃ P < 0.01 (B) Ba/F3 cells expressing the erythropoietin receptor (EpoR) and JAK2 V617F mutant were treated with ruxolitinib (0.3 μM) in combination with α-tocopherol (6.25, 25, 100 μM) for 24 hr. Cell viabilities were evaluated by a WST assay. Values are given as the mean ± SD of four independent experiments. ÃÃ P < 0.01 significantly different from the control group; ## P < 0.01 significantly different from the group incubated with 0.3 μM ruxolitinib. (DOCX)