Fullerene derivative prevents cellular transformation induced by JAK2 V617F mutant through inhibiting c-Jun N-terminal kinase pathway
Abstract
The constitutively activated mutation (V617F) of tyrosine kinase Janus kinase 2 (JAK2) is found in the majority of patients with myeloproliferative neoplasms (MPNs). The development of a novel chemical compound to suppress JAK2 V617F mutant-induced onset of MPNs and clarification of the signaling cascade downstream of JAK2 V617F mutant will provide clues to treat MPNs. Here we found that a water-soluble pyrrolidinium fullerene derivative, C60-bis (N, N-dimethylpyrrolidinium iodide), markedly induced apoptosis of JAK2 V617F mutant-induced trans- formed cells through a novel mechanism, inhibiting c-Jun N-terminal kinase (JNK) activation pathway but not gen- eration of reactive oxygen species (ROS). Pyrrolidinium fullerene derivative significantly reduced the protein expression level of apoptosis signal-regulating kinase 1 (ASK1), one of the mitogen-activated protein kinase kinase kinases (MAPKKK), resulting in the inhibition of upstream molecules of JNK, mitogen-activated protein kinase kinase 4 (MKK4) and mitogen-activated protein kinase kinase 7 (MKK7). Strikingly, the knockdown of ASK1 enhanced the sensitivity to pyrrolidinium fullerene derivative-induced apoptosis, and the treatment with a JNK inhibitor, SP600125, also induced apoptosis of the transformed cells by JAK2 V617F mutant. Furthermore, adminis- tration of both SP600125 and pyrrolidinium fullerene derivative markedly inhibited JAK2 V617F mutant-induced tumorigenesis in nude mice. Taking these findings together, JAK2 V617F mutant-induced JNK signaling pathway is an attractive target for MPN therapy, and pyrrolidinium fullerene derivative is now considered a candidate potent drug for MPNs.
1. Introduction
The classic Philadelphia chromosome-negative myeloproliferative neoplasms (Ph-MPNs), which include polycythemia vera (PV), essential thrombocytosis (ET) and primary myelofibrosis (PMF), are clonal disorders arising from hematopoietic stem cells or progenitor cells, and are charac- terized by uncontrolled proliferation of terminally differentiated myeloid cells. Clinical manifestations include variable degrees of erythrocytosis, thrombocytosis and leukocytosis, or cytopenias, extramedullary hemat opoiesis (e.g. splenomegaly), increased risk for thrombosis and transfor- mation to acute myeloid leukemia (AML) [1,2]. In 2005, a novel somatic mutation of the tyrosine kinase Janus kinase 2 (JAK2) gene was identified in more than 90% of PV patients and in approximately 50% of ET and PMF patients [3,4]. This dominant gain-of-function mutation is a guanine to thymidine substitution at nucleotide 1849 of the JAK2 gene, which results in a valine-to-phenylalanine substitution at codon 617 of JAK2. Recently, it has been reported that this point mutation of JAK2 could induce an MPN-like phenotype utilizing conditional knock-in mice, demonstrating that V617F mutation of the JAK2 gene is a cause of MPNs [5].
In hematopoietic cells, JAK2 regulates various cytokine-induced signaling pathways, including erythropoietin (Epo) [6]. JAK2 possesses seven Janus homology (JH) domains, from JH1 to JH7. JH1 is the C-terminal kinase domain and JH2 is a catalytically inactive pseudokinase domain [7]. In the inactive form of JAK2, JH1 is reported to be associated with JH2, which is reported to suppress the kinase activity of JH1 [8]. Although no detailed structural study has been completed, the V617F mutation results in a lack of autoinhibition, possibly by destabilizing the JH1–JH2 interaction [9].
Previously, we found that the JAK2 V617F mutant induces the cytokine-independent survival of erythroid progenitor cells [10]. In addition, we clarified that JAK2 V617F mutant requires Epo receptor (EpoR) as a scaffolding protein to exhibit transforming activity [11]. JAK2-induced phosphorylation of tyrosine residues at 343 and 479 in EpoR is essential for the recruitment and activation of STAT5 and the Akt signaling pathway, respectively. We demonstrated that STAT5 and Akt are critical signal transducers for a proliferative advantage and the transforming ability of JAK2 V617F mutant [12,13]; however, the details of the mechanism by which JAK2 V617F mutant causes tumorigenesis have not yet been brought to light, and it is expected that elucidation of the signaling pathway induced by JAK2 V617F mutant would provide clues to understanding the pathogenic mechanism of MPNs and devel- oping novel therapeutic procedures.
In 1985, fullerene (C60) was discovered as a new type of carbon allotrope that has been utilized for micro-scale devices in electronic and mechanical applications [14]. Chemical modification with several hydro- philic groups increases the solubility of fullerene, and water-soluble ful- lerene derivatives are reported to possess various biological and pharmacological properties, as described below. (1) Pyrrolidinium ful- lerene derivative induced apoptosis of human promyeloleukemia cells, HL60, through the generation of reactive oxygen species (ROS) [15]. (2) Malonic acid fullerene derivative has excellent antioxidant activity [16]. (3) Proline-modified fullerene derivative has inhibitory ac- tivity on human immunodeficiency virus (HIV)-reverse transcriptase [17] (Fig. 1).
In this study, we analyzed the effects of fullerene derivatives on JAK2 V617F mutant-mediated transformation. Pyrrolidinium-type full-erene derivative potently induced apoptotic cell death of JAK2 V617F mutant-transformed cells through inhibition of the JNK activation pathway and prevented JAK2 V617F mutant-induced tumorigenesis in vivo. These observations suggest that JNK could be a therapeutic target for MPNs and show the potential of the fullerene derivative to be a novel anti-MPN drug.
2. Materials and methods
2.1. Reagents
Three types of fullerene derivative were synthesized as previously described [15–17]. Recombinant human erythropoietin (Epo) (ESPO® 3000) was purchased from Kirin Brewery Co. (Tokyo, Japan). Antibodies against c-Myc, Odc and β-actin antibody were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-HA antibody (3 F10) was purchased from Roche (Indianapolis, IN). Other primary antibodies and peroxidase-conjugated secondary antibodies were purchased from Cell Signaling Technology (Danvers, MA).
2.2. Plasmids
Murine JAK2-HA was subcloned into retroviral plasmids, murine stem cell virus (MSCV)-GFP and MSCV-Hygro (Clontech, CA, USA). Mu- rine EpoR c-Flag was subcloned into MSCV-Puro (Clontech, CA, USA) [12]. The sequences of oligonucleotides used for constructing shRNA retroviral vector were as follows: sh-ASK1: 5′-gatccccgtgcagagactga gagtaattcaagagattactctcagtctctgcacttttta-3′ and 5′-agcttaaaaagtgcagaga ctgagagtaatctcttgaattactctcagtctctgcacggg-3′. Underlined sequences cor- respond to the sequence of murine ASK1 (from 4728 to 4746 in the 3′-untranslated region).
2.3. Cell cultures
Ba/F3 cells were infected with empty virus (−), wild-type murine JAK2 c-HA or a mutant of murine JAK2 c-HA (V617F) with full-length murine EpoR c-FLAG as described previously [10,11]. These cells were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (BioWest, France), 100 units/ml penicillin and 100 μg/ml strepto- mycin (Nacalai Tesque), with or without 5 units/ml Epo.
2.4. Immunoprecipitation and western blotting
Nuclear and cytosolic fractions were prepared as described previous- ly [18]. Cells were harvested in ice-cold PBS and lysed in Nonidet P-40 lysis buffer (50 mM Tris–HCl, pH 7.4, 10% glycerol, 50 mM NaCl, 0.5% sodium deoxycholate, 1% Nonidet P-40, 20 mM NaF, 0.2 mM Na3VO4) supplemented with protease inhibitors. Cell lysates were centrifuged at 15,000 rpm for 15 min to remove debris, and the supernatants were incubated with the indicated antibody for 4 h. Immune complexes were precipitated with protein G-Sepharose (Zymed Laboratories Inc.), washed three times with lysis buffer, and then eluted with sample buffer for SDS-PAGE. Eluted proteins were resolved by SDS-PAGE and trans- ferred to polyvinylidene difluoride membranes (Millipore, Billerica, MA). Membranes were probed using the designated antibodies and visu- alized with the ECL detection system (GE Healthcare) [10].
2.5. Ba/F3 cell growth assay and cell cycle analysis
Transduced Ba/F3 cells were incubated with RPMI 1640 medium supplemented with 1% fetal bovine serum for 24 h. Cell viability was checked by the trypan blue exclusion method [10]. Cell cycle parameters were determined by flow cytometry analysis using FACSCalibur [11].
2.6. DNA fragmentation assay
Genomic DNA was prepared for gel electrophoresis as described previously [11]. Electrophoresis was performed on a 1% agarose gel in Tris/boric acid buffer.
2.7. Transfection and luciferase assay
Cells were transfected with 1 μg of pAP-1-Luc, pSTAT3-Luc or pSTAT5-Luc and 0.5 μg of pRL-TK (Promega, Madison, WI) using FuGENE 6 (Roche). After 36 h, cells were treated with DMSO (0.1%), fullerene derivative I, SP600125 or AG490 for 12 h. Luciferase activi- ties were measured using the Mini Lumat LB9506 (Bertold). Transfec- tion efficacy was normalized with Renilla luciferase activity.
2.8. RNA isolation and reverse transcription–polymerase chain reaction (RT-PCR)
RNA was prepared using an RNA purification kit (Qiagen, Hilden, Germany). RT was performed using an oligo(dT)20 primer and 2 μg of total RNA for first-strand cDNA synthesis, as described previously [12]. PCR was performed at an annealing temperature of 57 °C with 18 amplification cycles for GAPDH (glyceraldehyde-3-phosphate dehydro- genase) and 23 amplification cycles for ASK1, c-Myc, Odc, Cdk4 and SOCS3. PCR products were resolved and electrophoresed on a 1% agarose gel in TAE (Tris-acetic acid-EDTA) buffer. PCR primer sequences were as follows: c-Myc, 5′-TGCGACGAGGAAGAGAATTT-3′ (upstream) and 5′-AACCGCTCCACATACAGTCC-3′ (downstream); Odc, 5′-GCCTGG CTCAGCTATGATTC-3′ (upstream) and 5′-CATCCAAAGGCAAAGTTGGT-
3′ (downstream); Cdk4, 5′-CTGGTACCGAGCTCCTGAAG-3′ (upstream) and 5′-TTGTGCAGGTAGGAGTGCTG-3′ (downstream); SOCS3, 5′-GTT GAGCGTCAAGACCCAGT-3′ (upstream) and 5′-CGTTGACAGTCTTCCGA CAA-3′ (downstream); GAPDH, 5′-ACTCCACTCACGGCAAATTC-3′ (upstream) and 5′-CCTTCCACAATGCCAAAGTT-3′ (downstream).
2.9. Animal tumorigenesis and administration of fullerene derivative and SP600125
BALB/c nude mice aged 6 weeks were injected intravenously (i.v.) with 5 × 106 transduced BaF3 cells expressing JAK2 V617F and EpoR. After injection of cells, the mice were treated daily for 6 days with intraperitoneal injection (i.p.) of pyrrolidinium-type fullerene deriva- tive I (20 mg/kg) or SP600125 (20 mg/kg).
2.10. Histological examination
After sacrifice, the liver of each nude mouse was fixed in 4% para- formaldehyde and then dehydrated gradually in alcohol. The tissues were embedded in paraffin and sectioned at a thickness of 2 μm.
Lymph nodes were embedded in Tissue-Tek OCT compound (Sakura Finetek, Torrance, CA) and frozen in liquid nitrogen. Cryostat sections were cut at 6 μm, air-dried and stored at −20 °C until use. The sections were stained with hematoxylin and eosin and analyzed for the presence of tumor cell infiltration using an Olympus BX50 microscope (Olympus, Tokyo, Japan) with Olympus Micro DP70 software [11,12].
3. Results
3.1. Pyrrolidinium-type fullerene derivative induced apoptosis of Ba/F3-VF cells
First, we tested whether three types of fullerene derivative, pyrrolidinium type, malonic acid type and proline-modified type (called Type I, II and III fullerene derivatives, respectively), could in- hibit the cell growth of Ba/F3 cells transformed by JAK2 V617F mutant (Ba/F3-VF cells) (Fig. 1A). As shown in Fig. 1B, only fullerene de- rivative I decreased the cell viability of Ba/F3-VF cells in a dose-dependent manner. On the other hand, the other types of deriv- ative failed to affect cell viability (Fig. 1B). When Ba/F3-VF cells were treated with fullerene derivative I for 24 h, there was an increased percentage in the sub-G1 phase, which was consistent with apoptotic cells (Fig. 1C). Ladder patterns of DNA internucleosomal fragmenta- tion and activation of caspase-3 were also observed in Ba/F3-VF cells treated with fullerene derivative I, confirming that these cells underwent apoptotic cell death (Fig. 1D, E).
3.2. Pyrrolidinium-type fullerene derivative induced apoptosis of Ba/F3- VF cells independently of ROS generation
In a previous study, fullerene derivative I was reported to induce apoptosis by ROS generation in HL60 cells [15]. Then, ROS generation in the cells treated with three types of fullerene derivative (I, II, III) was detected by oxidation of DCFH-DA (2′,7′-dichlorodihydrofluorescein- DA). As shown in Fig. 2A, enhanced generation of ROS was observed in Ba/F3-VF cells when treated with fullerene derivative I, but not with ful- lerene derivative II and fullerene derivative III (Fig. 2A). Next, to clarify whether oxidative stress is involved in fullerene derivative I-induced cell death of Ba/F3-VF cells, we tested the effect of an antioxidant, α-tocopherol, on fullerene derivative I-induced ROS generation and apoptosis. Pre-treatment with α-tocopherol significantly suppressed ful- lerene derivative I-induced intracellular oxidative stress in a dose- dependent manner (Fig. 2B). When treated with 300 μM α-tocopherol, intracellular oxidative stress was completely reduced to control levels; however, fullerene derivative I still induced cell death (Fig. 2B, C). In Ba/ F3-VF cells pretreated with α-tocopherol, the increased percentage in the sub-G1 phase, DNA fragmentation and caspase-3 activation were clearly detected by the addition of fullerene derivative I (Fig. 2D, E, F). Therefore, these results suggest that fullerene derivative I induced apo- ptosis through other mechanisms but not ROS generation in Ba/F3 cells transformed by JAK2 V617F mutant.
3.3. Pyrrolidinium-type fullerene derivative specifically inhibited JNK activation in Ba/F3-VF cells
To investigate the mechanism by which fullerene derivative I induces apoptosis in Ba/F3-VF cells, we first tested whether fullerene derivative I could directly inhibit JAK2 V617F mutant. The phosphorylation of Y1007/1008 within the activation loop of JAK2 is a well-known hallmark of JAK2 activation [6]. Unexpectedly, phosphorylation at Y1007/1008 in the JAK2 V617F mutant was still observed upon treatment with fullerene derivative I (Fig. 3A). Next, the effect of fullerene derivative I on down- stream molecules of JAK2 V617F mutant, including STAT3, STAT5 and Akt, was examined. In Ba/F3-VF cells, the constitutive phosphorylation of STAT3, STAT5 and Akt was detected as previously described [10–13]; however; fullerene derivative I had no effect on the phosphor- ylation of these molecules (Fig. 3B). Furthermore, the nuclear localiza- tion of STAT3 and STAT5 was normally observed in the presence of fullerene derivative I (data not shown). It was reported that ERK was ac- tivated by JAK2 V617F mutant in various cells [19]; however, no reported study has tested the abilities of JAK2 V617F mutant to activate other MAP kinase families, p38 and JNK. Therefore, the effect of JAK2 V617F mutant on the activation of MAP kinase families was determined. Epo signifi- cantly induced the activation of ERK and JNK but not p38 in empty virus-infected Ba/F3 cells and Ba/F3 cells expressing wild-type JAK2 (WT). On the other hand, in Ba/F3-VF cells, constitutive activation of ERK and JNK was observed regardless of Epo stimulation (Fig. 3C), suggesting that JAK2 V617F mutant could induce the activation of not only ERK but also JNK. Interestingly, fullerene derivative I significantly inhibited JNK activation induced by JAK2 V617F mutant; however, it had no effect on the activation of ERK. On the other hand, compared with cells treated with DMSO, p38 activation was slightly enhanced by fullerene derivative I. Furthermore, fullerene derivative I inhibited the phosphorylation of a substrate for JNK, c-Jun, confirming that it specifi- cally inhibited JNK activation downstream of JAK2 V617F mutant (Fig. 3D). As shown in Fig. 3E, other types of fullerene derivative showed no effect on the phosphorylation of JNK and c-Jun. When Ba/F3-VF cells were pretreated with α-tocopherol, fullerene derivative I still inhibited the phosphorylation of JNK and c-Jun (Fig. 3F), suggesting that fullerene derivative I selectively inhibited JNK activation independently of ROS generation.
3.4. Pyrrolidinium-type fullerene derivative reduced the protein expression level of ASK1, resulting in inhibition of MKK4 and MKK7 in Ba/F3-VF cells
To investigate further the inhibitory mechanism of JNK by fullerene derivative I, we first examined the effect of upstream molecules, MAPKKs such as MKK4 and MKK7. While Epo induced the activation of MKK4 and MKK7 in both empty virus-infected Ba/F3 cells and Ba/F3 cells expressing wild-type JAK2 (WT), constitutive activation of MKK4 and MKK7 was observed in Ba/F3-VF cells (Fig. 4A). ASK1, one of the MAPKKKs, is well known as being located upstream of MKK4 and MKK7 [20]. Interestingly, fullerene derivative I significantly reduced the protein expression level of ASK1 in a dose-dependent manner. In addition, phosphorylation of MKK4 and MKK7 was reduced by fullerene derivative I in a dose-dependent manner (Fig. 4B). The reduction of ASK1 and the consequent inhibition of MKK4 and MKK7 were observed in the cells treated with fullerene derivative I but not other types of fullerene derivative (Fig. 4C). Furthermore, fullerene derivative I-induced re- duction of ASK1 expression and inhibition of MKK4 and MKK7 were not affected by the pretreatment with α-tocopherol (Fig. 4D). Since ful- lerene derivative I had no effect on the expression level of ASK1 mRNA in Ba/F3-VF cells (Fig. 4E), it was suggested that the down-regulation of ASK1 seems to be due to translational or post-translational events, but not to transcription steps. It has been reported that ASK1 was degraded through the ubiquitin-proteasome system [21]. Therefore, the effect of fullerene derivative I on ASK1 expression was examined in the ab- sence and presence of proteasome inhibitor, MG132. As expected, the down-regulation of ASK1 expression by fullerene derivative I was par- tially inhibited in Ba/F3-VF cells treated with MG132, suggesting that ful- lerene derivative I induced degradation of ASK1 by proteasome (Fig. 4F).
3.5. Knockdown of ASK1 drastically accelerates pyrrolidinium-type fullerene derivative-induced apoptosis of Ba/F3-VF cells
The results shown in Fig. 4 suggested the possibility that the reduction of ASK1 is the mechanism by which fullerene derivative I causes cell toxicity to Ba/F3-VF cells. To test this possibility, endoge- nous ASK1 was knocked down in Ba/F3-VF cells using shRNA. As a control, we used the shRNA expression vector against firefly lucifer- ase (sh-Luc). The shRNA against ASK1 (sh-ASK1) effectively reduced the expression of ASK1 in Ba/F3-VF cells. Remarkably, the knockdown of ASK1 reduced phosphorylation of its downstream components, such as MKK4, MKK7, JNK and c-Jun (Fig. 5A). Then, the viability of cells infected with sh-Luc or sh-ASK1 was investigated in the pres- ence and absence of fullerene derivative I. Knockdown of ASK1 slight- ly reduced cell viability of Ba/F3-VF cells and markedly enhanced their sensitivity to fullerene derivative I-induced cell death (Fig. 5B).
3.8. Pyrrolidinium-type fullerene derivative effectively inhibited JAK2 V617F mutant-induced tumorigenesis in nude mice
To examine the effect of pyrrolidinium-type fullerene derivative on tumorigenesis in vivo, nude mice were intravenously injected with Ba/F3 cells transformed by JAK2 V617F mutant and then treated with intraperitoneal injection of fullerene derivative I or SP600125 daily for 6 days. The nude mice treated with fullerene derivative I exhibited no obvious alteration, including in terms of life span (data not shown) and the size and weight of the liver and spleen (Fig. 8A, B). In nude mice receiving Ba/F3-VF cells, the spleen and liver were abnormally enlarged. Compared with Ba/F3-VF cell-transplanted nude mice treated with olive oil (vehicle), the en- largement of the spleen and liver was significantly suppressed by administration of fullerene derivative I and SP600125 (Fig. 8A, B). In this study, we took advantage of a retroviral expression vector, MSCV-IRES-GFP, and sorted the infected cells on the basis of GFP ex- pression [10]. To examine the invasive ability of Ba/F3-VF cells, splenocytes were prepared from Ba/F3-VF cell-transplanted mice, and the presence of GFP-positive cells was analyzed by flow cytometry analysis. GFP-positive cells were detected in splenocytes prepared from Ba/F3-VF cell-inoculated mice with vehicle treatment. Strikingly, the population of GFP-positive cells was significantly re- duced by administration of fullerene derivative I and SP600125 (Fig. 8C). Furthermore, while the infiltration of tumor cells into the liver was obse-rved in Ba/F3-VF cell-transplanted mice with vehicle treatment, tumor cells in the liver were decreased by administration of fullerene derivative I and SP600125 (Fig. 8D). Additionally, the life span of nude mice receiving Ba/F3-VF cells was effectively pro- longed when treated with both fullerene derivative I and SP600125 compared with that with vehicle treatment (Fig. 8E). These results indicate that JNK plays a central role in JAK2 V617F mutant-induced tumorigenesis, showing the potential use of fullerene derivatives for effective treatment against MPNs.
4. Discussion
In this study, we first found that the JAK2 V617F mutant induced con- stitutive activation of JNK, and this led to anti-apoptotic and oncogenic activity. Furthermore, a specific JNK inhibitor, SP600125, significantly induced apoptosis of cells expressing the JAK2 V617F mutant (Fig. 6). A number of studies have provided evidence that JNK is a key regulator of apoptosis; however, it has also been reported that JNK could function as a pro-apoptotic protein. This discrepancy is now understood to be dependent on the cell type and stimuli [23,24]. As shown in Fig. 1E, fullerene derivative I also induced the activation of caspase-3, which is a key regulator of apoptosis, and this is known to be strictly regulated by mitochondrial Bcl-2 family proteins. JNK directly phosphorylates and regulates the expression of several members of the Bcl-2 protein family, such as BAX and BAD, as well as 14-3-3 proteins. Phosphorylated 14-3-3 proteins release pro-apoptotic proteins, including BAX and Forkhead transcription factor FOXO, from inactive complexes, thereby facilitating JNK-mediated apoptosis [30]. On the other hand, the molecular mechanism by which JNK suppresses apoptosis is poorly understood. Phosphorylation of BAD by PKA and Akt causes its cytoplasmic seques- tration by 14-3-3ζ. In contrast, JNK phosphorylates BAD at a different res- idue, Thr201, and this phosphorylation is reported to result in reduction of the association of BAD with Bcl-XL, which can be a trigger of apoptotic cell death [31]. Although we do not have any data suggesting that fuller- ene derivative I affects the phosphorylation of Bcl-2 families, the data on the induction of apoptosis and inhibition of JNK activation encouraged us to speculate that fullerene derivative could regulate the action of Bcl-2 family proteins.
Although fullerene derivative I induced intracellular ROS, the ROS was negligible for JAK2 V617F-induced transformation and anti-apoptotic actions. Strikingly, the inhibition of JNK activation by fullerene derivative I through the down-regulation of ASK1 expression was not affected by α-tocopherol (Fig. 4D), and this fits well with the results of fullerene derivative I-induced apoptosis (Fig. 2C–F). Furthermore, the amount of ROS production in cells expressing JAK2 V617F mutant was comparable with that in parental Ba/F3 cells (data not shown), suggesting that ROS production might be dispensable for the JAK2 V617F mutant-mediated signaling pathway and that fullerene derivative I may utilize a novel mechanism to induce apoptotic cell death without ROS production (Fig. 9). As expected, treatment with JNK inhibitor SP600125 gave similar results, strongly suggesting that fullerene derivative I-induced apoptosis is attributable to the inhibition of JNK; however, it is still under debate how the fullerene derivative inhibits the JNK pathway. It has been shown that thioredoxin in a reduced form binds to and inhibits ASK1 and that ROS activates ASK1, in part by oxidizing thioredoxin, to release thioredoxin from ASK1 [32]. It seems to be a contradiction that ROS induced the activation of JNK and p38 mediated by ASK1. ASK1 phosphor- ylates MKK4/7 and MKK3/6, which phosphorylate and activate JNK and p38, respectively [20]. Interestingly, fullerene derivative I specifically inhibited the activation of JNK through down-regulation of ASK1 expres- sion and inhibition of MKK4 and MKK7, but increased the phosphoryla- tion level of p38 (Fig. 3D). While phosphorylation of ASK1 at threonine 838 was observed in Ba/F3-VF cells but not in Ba/F3 cells expressing wild-type JAK2 (WT) (data not shown), phosphorylation of p38 was always detected regardless of the JAK2 activation state (Fig. 3C). There- fore, it is thought that it does not depend on ASK1 activity for the p38 ac- tivation pathway observed in Ba/F3-VF cells, although the involvement of ASK1 in ROS-mediated p38 activation is unknown. There is one report showing that JAK2 controls the expression level of ASK1. Yu et al. reported that JAK2 induced phosphorylation at tyrosine 718 and degradation of ASK1. It was further shown that tyrosine-phosphorylated ASK1 binds to the suppressor of cytokine signaling −1 (SOCS1), a subunit of ubiquitin ligase responsible for ASK1 degradation [33]. However, we observed that the expression level of ASK1 was not changed in parental Ba/F3
cells, Ba/F3 cells expressing wild-type JAK2 (WT) and Ba/F3-VF cells (data not shown). Therefore, in our system, it is suggested that JAK2 activ- ity is required for the activation of ASK1, but is not responsible for its deg- radation. Interestingly, since the down-regulation of ASK1 by fullerene derivative I was partially canceled by the treatment with MG132, it is suggested that fullerene derivative I enhanced the degradation of ASK1 through the proteasome system (Fig. 4F). The cellular inhibitor of apopto- sis protein 1 (c-IAP1) was identified as the ubiquitin protein ligase for ASK1 ubiquitination in the TNFα signaling pathway [21]. From the obtained results, there are two possibilities: that fullerene derivative I could reinforce ubiquitination of ASK1 by c-IAP1 or that fullerene deriva- tive I could directly enhance proteasome activity. It is necessary to analyze these possibilities in the future. In fact, when ASK1 was knocked down, phosphorylation of MKK4, MKK7, JNK and c-Jun was reduced in Ba/ F3-VF cells; however, the viability of Ba/F3-VF cells was only reduced slightly (Fig. 5A, B). As one of the reasons for this result, it is thought that the knockdown efficiency of ASK1 is insufficient to induce apoptotic cell death of Ba/F3-VF cells. Moreover, as another possibility, it is thought that there is another target of fullerene derivative I in addition to the down-regulation of ASK1.
The role of the JNK pathway in tumorigenesis is supported by the high level of JNK activity found in several human cancer cell lines [34]. Nateri et al. developed a mouse model for intestinal cancer and showed that the phosphorylation of c-Jun by JNK contributes to tu- morigenesis [34]. Furthermore, a proto-oncogene product, c-Myc, was identified as a target gene of AP-1 [25] and, in our study, treat- ment with fullerene derivative I and JNK inhibitor effectively reduced the expression of c-Myc and its target genes (Fig. 7). Previously, we observed that c-Myc contributed to JAK2 V617F mutant-induced tu- morigenesis (data not shown), suggesting that the importance of the JNK-c-Jun pathway in the oncogenic signals induced by JAK2 V617F may be due to the induction of c-Myc (Fig. 9).
Since MKK4- and MKK7-deficient mice exhibited embryonic le- thality because of the failure of liver generation [35–37], it is possible that fullerene derivative I and SP600125 cause toxicity to not only tumor cell-transplanted nude mice but also to healthy controls. How- ever, we did not observe that these chemical compounds caused any toxicity to the healthy controls, and effectively suppressed JAK2 V617F mutant-induced tumorigenesis and prolonged the survival of nude mice transplanted with tumor cells expressing JAK2 V617F. Although we have no data explaining why these compounds exh-ibited no toxicity, the findings strongly suggest that JNK inhibitor,Mycro 3 including pyrrolidinium-type fullerene derivative, could be used as a potent ther- apeutic drug for MPNs.