Sorafenib-Induced Apoptosis of Chronic Lymphocytic Leukemia Cells Is Associated with Downregulation of RAF and Myeloid Cell Leukemia Sequence 1 (Mcl-1)
We have previously shown that sorafenib, a multikinase inhibitor, exhibits cytotoxic effects on chronic lymphocytic leukemia (CLL) cells. Because the cellular microenvironment can protect CLL cells from drug-induced apoptosis, it is important to evaluate the effect of novel drugs in this context. Here we characterized the in vitro cytotoxic effects of sorafenib on CLL cells and the underlying mechanism in the presence of marrow stromal cells (MSCs) and nurselike cells (NLCs). One single dose of 10 µmol/L or the repeated addition of 1 µmol/L sorafenib caused caspase-dependent apoptosis and reduced levels of phosphorylated B-RAF, C-RAF, extracellular signal-regulated kinase (ERK), signal transducer and activator of transcription 3 (STAT3) and myeloid cell leukemia sequence 1 (Mcl-1) in CLL cells in the presence of the microenvironment. We show that the RAF/mitogen-activated protein kinase kinase (MEK)/ERK pathway can modulate Mcl-1 expression and contribute to CLL cell viability, thereby associating sorafenib cytotoxicity to its impact on RAF and Mcl-1. To evaluate if the other targets of sorafenib can affect CLL cell viability and contribute to sorafenib-mediated cytotoxicity, we tested the sensitivity of CLL cells to several kinase inhibitors specific for these targets. Our data show that RAF and vascular endothelial growth factor receptor (VEGFR) but not KIT, platelet-derived growth factor receptor (PDGFR) and FMS-like tyrosine kinase 3 (FLT3) are critical for CLL cell viability. Taken together, our data suggest that sorafenib exerts its cytotoxic effect likely via inhibition of the VEGFR and RAF/MEK/ERK pathways, both of which can modulate Mcl-1 expression in CLL cells. Furthermore, sorafenib induced apoptosis of CLL cells from fludarabine refractory patients in the presence of NLCs or MSCs. Our results warrant further clinical exploration of sorafenib in CLL.
Chronic lymphocytic leukemia (CLL) is the most common leukemia in the Western world, leading to approximately 5,000 deaths annually (1). CLL is characterized by an accumulation of monoclonal mature B cells in the blood, secondary lymphoid tissues and the marrow. Despite major advances in the field, there is no curative therapy for CLL to date, and new strategies are needed (2). Current treatment approaches aim at achieving minimal residual disease, which is associated with superior long-term outcome (3). The frontline therapy for CLL is the purine analog fludarabine. However, 30% of patients treated with fludarabine do not achieve complete remission, even when used in combination with other agents (2). To improve this outcome, other treatment avenues, such as those targeting pathways downstream of the B-cell receptor, are currently being evaluated in preclinical and early clinical trials (3).
More recently, new therapeutic strategies have been designed to abrogate the prosurvival interaction of CLL cells with their microenvironment and the related signaling pathways. Accessory cells such as nurselike cells (NLCs) and marrow stromal cells (MSCs) protect CLL cells from drug-induced apoptosis in vitro (4). Thus, it has been postulated that CLL cells receive survival signals from these accessory cells, which constitute part of the CLL B-cell microenvironment in secondary lymphoid tissue and marrow (5,6). Such niches could protect leukemia cells from spontaneous or drug-induced apoptosis in vivo. Therefore, it is essential when investigating novel CLL drugs in vitro to test them in the context of the microenvironment.
Exposure of CLL cells to proteins released from cells of the microenvironment causes activation of the extracellular signal-regulated kinase (ERK) signaling pathway, which is an important mediator of CLL cell survival (7,8) and thus an attractive drug target. These proteins include chemokine (C-X-C motif) ligand 12 (CXCL12) and chemokine (C-C motif) ligand 19/21 (CCL19/CCL21), which signal through their respective receptors, CXCR4 and CCR7. We have previously shown that sorafenib (BAY 43-9006, Nexavar), an orally active multikinase inhibitor that targets RAF kinases, as well as several receptor tyrosine kinases (9), prevents CXCL12-mediated upregulation of the active form of mitogen-activated protein kinase kinase (MEK) and ERK in CLL cells and causes cell death (7). Sorafenib causes apoptosis in leukemia cell lines and in blast cells from patients with acute myeloid leukemia (10) and displays a broad-spectrum antitumor activity in colon, breast and non-small-cell lung cancer xenograft models (11). Sorafenib was approved by the U.S. Food and Drug Administration for the treatment of patients with advanced renal cell carcinoma and unresectable hepatocellular carcinoma (12). As the first drug to improve the survival of patients with hepatocellular carcinoma, sorafenib is currently being tested in clinical trials for its efficacy in the treatment of other solid tumors such as thyroid carcinoma (12). Here we investigated the mechanism of sorafenib-mediated CLL cytotoxicity in the context of the cellular microenvironment.
Materials and Methods
Isolation and Purification of CLL B Cells
Blood samples were collected from patients at the Moores University of California San Diego (UCSD) Cancer Center who satisfied diagnostic and immunophenotypic criteria for common B-cell CLL after providing written informed consent in compliance with the Declaration of Helsinki (13) and the institutional review board of UCSD. Peripheral blood mononuclear cells (PBMCs) were isolated from CLL patients by density centrifugation with Ficoll-Hypaque (Pharmacia, Uppsala, Sweden), resuspended in 90% fetal calf serum (Omega Scientific, Tarzana, CA, USA) and 10% dimethylsulfoxide (DMSO) (Sigma) for viable storage in liquid nitrogen. If not otherwise indicated, the CLL cells were isolated from thawed PBMCs via negative selection using anti-CD2 and anti-CD14 magnetic beads (Miltenyi Biotechnology, Auburn, CA, USA).
Patients deemed refractory to fludarabine were defined as showing less than a partial response after completing a fludarabine regimen (14). Among the four patients selected, three presented with progressive disease and one presented with a stable disease after treatment. Cytogenetic analyses were available for three of the four patients tested and revealed that two patients presented no chromosomal abnormalities and one presented with 12q trisomy.
Sorafenib was purchased from LC Laboratories (Woburn, MA, USA) and solubilized in DMSO. DMSO was used in all experiments as a vehicle control. Fludarabine monophosphate (F-ara-A), the MEK inhibitor PD98059 and the RAF inhibitor GW5074 were obtained from Sigma, whereas the caspase inhibitor Z-VAD-FMK was purchased from BD. The B-RAF and C-RAF inhibitor KG5, the control kinase inhibitor KG1 and vatalanib were provided by D Cheresh (University of California San Diego).
Generation of NLCs
PBMCs were isolated from the blood of normal volunteers (anonymously purchased from the San Diego Blood Bank) over a Ficoll density gradient (GE Healthcare, Piscataway, NJ, USA). CD14+ monocytes were isolated from PBMCs by positive selection using anti-CD14 beads (Miltenyi Biotech) following the manufacturer’s instructions. To generate NLCs, 1.25 × 105/well CD14+ cells were cocultured with 3 × 106/well purified CLL B cells in 1 mL media in a 24-well plate (BD, Franklin Lakes, NJ, USA) in culture media (RPMI 1640 supplemented with 10 mmol/L 4-(2-hydroxyethyl)-1-piperazine ethanesulfonic acid [HEPES] [GIBCO-BRL]), penicillin (100 U/mL)-streptomycin (100 µg/mL) (GIBCO-BRL) and 10% pooled human serum (Omega Scientific) for 12 d. At this point, CLL B cells were gently washed off, and the adherent NLCs were used for coculture experiments using freshly purified CLL cells.
Generation of MSCs
Mononuclear cells from marrow aspirates of CLL patients were isolated after centrifugation over a Ficoll density gradient. The cells were seeded at a density of 2 × 106 cells/mL in DMEM (Mediatech, Manassas, VA, USA) supplemented with 10% fetal calf serum (Omega Scientific), 10 mmol/L HEPES, 100 U/mL penicillin and 100 µg/mL streptomycin and cultured at 5% O2 (using a Sanyo incubator, MCO-18M) for approximately 3 wks, with media renewal every week. The adherent MSCs were expanded and used for the coculture assays after two to six passages. MSCs were seeded between 5,000 and 10,000 cells/well in 24-well plates at least 1 d before the addition of CLL cells. Purified CLL cells were seeded at 1 × 106 cells/mL (1 mL/well) in the same media used to generate MSCs and were cultured at 5% O2 with or without 10 µmol/L sorafenib and the appropriate DMSO control. Measurement of cell viability was performed at the indicated time points as described below. For MSC conditioned media preparation, MSCs were plated between 2,000 and 3,000 cells/cm2 and incubated in the media described above for 7–9 d at 5% O2, at which point the supernatant was collected, spun down and frozen at −20°C until further use. One hundred percent of conditioned media was used for CLL cell stimulation.
Measurement of Cell Viability
Purified CLL cells were cultured at 1 × 106 cells/mL in 24-well plates (BD) under various conditions. Determination of CLL cell viability was on the basis of the analysis of mitochondrial transmembrane potential (ΔΨm) using 3,3′-dihexyloxacarbocyanine iodide (DiOC6) (Invitrogen) and cell membrane permeability to propidium iodide (PI) (Sigma). For viability assays, 100 µL of the cell culture was collected at the indicated day and transferred to polypropylene tubes containing 100 µL of 40 µmol/L DiOC6 and 10 µg/mL PI in culture media. The cells were then incubated at 37°C for 15 min and analyzed within 30 min by flow cytometry using a FACSCalibur (Becton Dickinson). Fluorescence was recorded at 525 nm for DiOC6 and at 600 nm for PI.
Data were analyzed using the FlowJo 7.2.2 software (Tree Star). The percentage of viable cells was determined by gating on PI-negative and DiOC6 bright cells. When CLL cells were collected from cocultures in the presence of NLCs or MSCs, there were typically <5% supportive cells followed along with CLL cells. These supportive cells were excluded from the analysis on the basis of their significantly larger size using forward and side scatter analysis.
For the evaluation of signal transducer and activator of transcription 3 (STAT3) expression levels after sorafenib exposure in the presence of MSCs, the culture media were replaced 1 d before the experiment on MSCs. On the day of the experiment, CLL cells were first serum-starved for 2 h in RPMI, followed by a pretreatment of 30 min with 10 µmol/L sorafenib or DMSO. At that point, CLL cells were spun down, and the cell pellet was resuspended in 24-h MSC-conditioned media, to which 10 µmol/L sorafenib or DMSO was added, and CLL cells were cocultured with MSCs for another 30 min. At that point, the CLL cells were collected for protein extraction as described below.
For the study of prosurvival proteins and modulation of the RAF/MEK/ERK pathway by sorafenib, CLL cells were exposed to 30 nmol/L CXCL12 (prepared as described by Messmer et al. ), NLCs or MSCs without prior starvation, and at the time of coculture, 10 µmol/L sorafenib or DMSO was added for 24 h. CLL cells were collected and the adherent NLCs or MSCs were left behind in the wells, as confirmed by bright-field microscopy. CLL cells were lysed for 20 min on ice with radioimmunoprecipitation assay lysis buffer (10 mmol/L Tris, pH 7.4, 150 mmol/L NaCl, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 5 mmol/L ethylenedi-aminetetraacetic acid supplemented with 1 mmol/L phenylmethylsulfonyl fluoride, Halt phosphatase inhibitor [Thermo Fisher Scientific, Rockford, IL, USA], 1 mmol/L sodium vanadate, 1 mmol/L sodium fluoride and protease inhibitor cocktail [Roche Applied Science, Indianapolis, IN, USA]). Protein concentration was determined using the detergent compatible (DC) protein assay (Bio-Rad, Hercules, CA, USA). The lysates were snap-frozen and stored at −80°C. Equal amounts of protein lysates (∼20 or 30 µg) were separated by gel electrophoresis using a NuPAGE Novex 4–12% Bis-Tris Midi Gel (Invitrogen) and transferred to polyvinylidene fluoride membranes (Bio-Rad). Membranes were washed with 1× Tris-buffered saline tween-20 (TBST), blocked for 1 h at room temperature in 5% milk/TBST and probed overnight for phospho-B-RAF (Ser445), B-RAF, Bcl-XL, Bcl-2 interacting mediator of cell death (Bim), phospho-C-RAF (Ser338), C-RAF, phospho-p44/p42 (ERK1/2; Thr202/Tyr204), myeloid cell leukemia sequence 1 (Mcl-1), phospho-STAT3 (Tyr705), STAT3, β-actin or GAPDH, using antibodies from Cell Signaling Technology (Danvers, MA, USA), for Bcl-2 using an antibody from Santa Cruz Biotechnology (Santa Cruz, CA, USA) and for poly(ADP-ribose) polymerase (PARP) using an antibody from BD Biosciences. The next day, membranes were washed with 1× TBST and incubated with goat-anti-rabbit or anti-mouse horseradish peroxidase-conjugated secondary antibodies (Santa Cruz Biotechnology) diluted to 1:12,000 to 1:15,000 in 5% milk/TBST for 1 h at room temperature. Antibodies were detected either using an enhanced chemiluminescence (ECL) detection kit (GE Healthcare) or SuperSignal West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific). For densitometry analysis, the intensity of each band was determined using the free National Institutes of Health ImageJ software (http://rsbweb.nih.gov/ij), divided by the intensity of control protein (as indicated in the figure legends).
Data are represented as means ± SD. Data were analyzed for statistical significance using the paired Student t test. P values <0.05 were considered statistically significant (*P < 0.05, **P < 0.01, ***P < 0.001).
Sorafenib Causes Apoptosis of CLL Cells in the Presence of NLCs and MSCs
CLL Cells from Fludarabine-Refractory Patients Are Sensitive to Sorafenib-Mediated Cytotoxicity, Even in the Presence of the Protective Microenvironment
Sorafenib-Mediated Apoptosis in CLL Cells is Caspase-Dependent
Sorafenib Decreases Mcl-1 Expression in CLL Cells in the Presence of NLCs and MSCs
Sorafenib Reduced the Levels of Active B-RAF, C-RAF and ERK in CLL Cells
CLL Cells Are Not Sensitive to Inhibition of KIT, PDGFR and FLT3 but Are Sensitive to Inhibition of RAF and VEGFR
Overall, in comparison to the other inhibitors, sorafenib was still the most potent drug-inducing CLL cell apoptosis. Because the inhibition of KIT, PDGFR and FLT3 did not affect CLL cell survival, our results suggest that RAF and VEGFR are the most likely active targets of sorafenib in CLL.
Abundant evidence shows that CLL cells can be rescued by accessory cells from spontaneous and drug-induced apoptosis (16,31,32) and protect CLL cells from fludarabine-induced apoptosis in vitro (4). Thus, it is essential to evaluate potential therapeutics in CLL accessory cell cocultures. Here, we show preclinical data reporting the sensitivity of CLL cells to sorafenib-mediated cytotoxicity when cocultured in the presence of accessory cells, suggesting that sorafenib might be a potent new therapeutic for CLL. Sorafenib has a reported elimination half-life of 20.0–27.4 h (33). When given twice daily at 400 mg, the maximum plasma concentration reaches 8.5 µmol/L after 28 d (33), 9.7 µmol/L after 7 d (34) and 9.9 µmol/L after 6 h (35). We show that a single dose of 10 µmol/L sorafenib, a level achievable in vivo, dramatically induces caspase-dependent apoptosis in CLL cells in the presence of NLCs and MSCs. Moreover, sorafenib effectively induced apoptosis of CLL cells isolated from fludarabine-refractory patients even when cocultured with NLCs and MSCs, further suggesting its potential for clinical use as second-line therapeutic strategy.
Sorafenib-induced cytotoxicity of CLL cells appears to be mediated by its impact on the antiapoptotic Bcl-2 family member protein Mcl-1, which was downregulated in all CLL samples cultured alone and in the presence of accessory cells. Mcl-1 plays an important role in CLL cell survival, since silencing of Mcl-1, but not that of Bcl-XL or XIAP (X-linked inhibitor of apoptosis), reduced CLL cell viability (24,31). Consistent with our observations, sorafenib has been shown to induce apoptosis along with a reduction of Mcl-1 protein levels in leukemia cell lines (10) and in CLL cells cultured in absence of the microenvironment (36). In addition, the enforced expression of Mcl-1 in cell lines reduced sorafenib-mediated apoptosis (10), supporting the notion that Mcl-1 downmodulation is likely contributing to sorafenib-mediated cytotoxicity in primary CLL.
In other cancers, sorafenib-induced apoptosis has been shown to involve the downregulation of Mcl-1 via RAF/MEK/ERK-dependent as well as -independent pathways depending on the tumor type (37). We found that sorafenib reduced the activation of B-RAF and C-RAF as well as its downstream mediator ERK in CLL cells cultured in the presence of NLCs or MSCs. Because sorafenib was shown in vitro not to be a direct inhibitor of the activity of MEK and ERK (28), it strongly suggests that the impact of sorafenib on ERK is related to the inhibition of its upstream mediators B- and C-RAF. Several lines of evidence demonstrate a link between ERK and Mcl-1 expression in CLL cells. We showed that treatment of CLL cells with the MEK inhibitor PD98059 inhibited CXCL12-included Mcl-1 upregulation, showing that MEK signaling contributes to Mcl-1 expression in CLL cells. Our results also show that the same strategy to inhibit MEK in CLL cells led to a downregulation of Mcl-1, even in the presence of MSCs, further supporting the regulatory role of the RAF/MEK/ERK pathway on Mcl-1 expression. Comparable observations were made in melanoma cells, where inhibition of MEK using PD98059 also caused downregulation of Mcl-1 (38). In addition, it was been shown that ERK activation can lead to Mcl-1 phosphorylation, which in turn increases its stability (39). Moreover, we show that the RAF inhibitor GW5074 reduced Mcl-1 expression and viability in CLL cells in the presence of the microenvironment, further supporting the functional link between RAF, Mcl-1 expression and viability of CLL cells. Thus, we reasoned that the impact of sorafenib on RAF activity contributes to Mcl-1 downregulation and consequently CLL cell death.
Because sorafenib is a multikinase inhibitor, we evaluated which of its targets in addition to RAF (VEGFR, PDGFR, KIT, FLT3) are critical for CLL cell viability using a set of kinase inhibitors. KG5 is a kinase inhibitor of RAF, PDGFR α and β, FLT3 and KIT (29), whereas KG1 targets PDGFR α and β, FLT3 and KIT but not RAF (29). Vatalanib targets KIT, PDGFR and VEGFR (40). Our results show that sorafenib, KG5 and vatalanib induced apoptosis of CLL cells, but KG1 failed to do so. Because the common targets of the active drugs are RAF and VEGFR, these results suggest that PDGFR, FLT3 and KIT are unlikely to be not critical for CLL cell viability and that sorafenib most likely causes CLL cell apoptosis via its inhibition of RAF and VEGFR. Although the role of VEGFR in CLL remains controversial, the majority of the evidence points toward the involvement of VEGFR to CLL cell viability. Huber et al. (36) have shown that bevacizumab, a monoclonal antibody against VEGF, did not induce apoptosis in CLL cells, and immunoprecipitation of VEGFR in CLL cells treated with sorafenib showed no effects on phosphotyrosine. However, Lee et al. (41) showed that inhibition of VEGFR signaling in CLL cells decreases Mcl-1 levels and induces cell death when 10-fold higher levels of bevacizumab were used, supporting the role of autocrine VEGF in CLL survival. In line with these findings, VEGFR signaling was demonstrated to support CLL cell survival through the upregulation of XIAP and Mcl-1 (30). In CLL, VEGFR signaling is not mediated through the activation of ERK or AKT (30), but through the activation of STAT3, which physically associates to VEGFR and translocate to the nucleus after activation (41). The blockade of VEGFR signaling in CLL using monoclonal antibodies or specific receptor tyrosine kinase inhibitor (SU11657) was shown to inhibit STAT3 activation and to induce apoptosis, marked by Mcl-1 downregulation (41). Comparable findings were also recently reported regarding the effect of vatalanib on CLL cells in vitro, which was shown to directly reduce the activation of VEFGR, to induce CLL cell apoptosis and to modulate Mcl-1 expression levels in a dose-dependent manner (42). Overall, these studies indicate that VEGFR signaling plays a role in CLL cell survival, which involves the activation of STAT3 and Mcl-1. We show that sorafenib downregulates STAT3 as well as Mcl-1 and have demonstrated a functional link between RAF and Mcl-1 expression in CLL cells, suggesting that sorafenib downregulated Mcl-1 expression by interfering with the VEGFR/STAT3 and the RAF/MEK/ERK pathways.
On the basis of the reported in vivo pharmacokinetic data and safety profile in the treatment of solid tumors (33), sorafenib represents a promising therapeutic agent for CLL. The dramatic reduction in CLL cell viability at 10 µmol/L and lower concentrations of sorafenib in vitro, even in the presence of a supportive microenvironment, substantiates its use in CLL. The sensitivity of CLL cells derived from fludarabine-refractory patients further suggests that sorafenib could represent a viable option as a second-line therapy for CLL. A phase I/II clinical trial is being initiated to assess the effect of sorafenib in CLL patients.
The authors declare that they have no competing interests as defined by Molecular Medicine, or other interests that might be perceived to influence the results and discussion reported in this paper.
This work was supported by the Lymphoma Research Foundation (grant CLL-07-029 to TM Handel, TJ Kipps and D Messmer), Le Fond de la Recherche en Santé du Québec (to J-F Fecteau) and NIH R01AI37113 to TM Handel. The authors would like to thank Andrew Abriol Santos Ang, Vania Frias and Colette Yee for their excellent technical assistance and Bradley Messmer for critical reading of the manuscript.
- 13.Kipps TJ. (2001) Chronic Lymphocytic Leukemia and Related Diseases. In: Williams Hematology. Beutler E, Lichtman MA, Coller BA, Kipps TJ, Seligsohn U (eds.). New York: McGraw-Hill, Medical Publishing Division, pp. 1163–94.Google Scholar
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, and provide a link to the Creative Commons license. You do not have permission under this license to share adapted material derived from this article or parts of it.
The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.
To view a copy of this license, visit (http://creativecommons.org/licenses/by-nc-nd/4.0/)