Induction of Endoplasmic Reticulum Stress by Sorafenib and Activation of NF-Kb by Lestaurtinib As a Novel Resistance Mechanism in Hodgkin Lymphoma Cell Lines

Published OnlineFirst December 13, 2012; DOI: 10.1158/1535-7163.MCT-12-0532

Molecular Cancer Cancer Therapeutic Insights Therapeutics

Induction of Endoplasmic Reticulum Stress by Sorafenib and Activation of NF-kB by Lestaurtinib as a Novel Resistance Mechanism in Hodgkin Lymphoma Cell Lines

Meike Stefanie Holz1, Angela Janning1, Christoph Renne2, Stefan Gattenlohner€ 3, Tilmann Spieker1, and Andreas Brauninger€ 1,3

Abstract Hodgkin-Reed/Sternberg (HRS) cells of classical Hodgkin lymphoma show aberrant expression and activation of several receptor tyrosine kinases (RTK) in the majority of cases. Therefore, we tested whether tyrosine kinase inhibitors (TKI) already in clinical use or late stages of clinical trials have antiproliferative effects on HRS cell lines and evaluated the targets, affected signaling pathways, and mechanisms of cell death and resistance. Sorafenib and lestaurtinib had antiproliferative effects on HRS cell lines at concentrations achievable in patients. Sorafenib inhibited platelet-derived growth factor receptor (PDGFR) a, TRKA and RON, caused decreases in total and phosphorylated amounts of several signaling molecules, and provoked caspase-3–independent cell death, most likely due to endoplasmic reticulum stress as indicated by upregulation of GADD34 and GADD153 and phosphorylation of PERK. Lestaurtinib inhibited TRKA, PDGFRa, RON, and JAK2 and had only a cytostatic effect. Besides deactivation, lestaurtinib also caused activation of signaling pathways. It caused increases in CD30L and TRAIL expression, and CD30L/CD30 signaling likely led to the observed concomitant activation of extracellular signal–regulated kinase 1/2 and the alternative NF-kB pathway. These data disclose the possible use of sorafenib for the treatment of Hodgkin lymphoma and highlight NF-kB activation as a potential novel mechanism of resistance toward TKIs. Mol Cancer Ther; 12(2); 173–83. 2012 AACR.

Introduction lymphoma cases (5), the receptors are presumed to be In HRS cells of Hodgkin lymphoma, the activation of activated in an autocrine or paracrine fashion. several signaling pathways is caused not only by muta- Several inhibitors of tyrosine kinases are highly effec- tions but also by the activation of receptors like receptor tive in the treatment of malignancies with aberrantly tyrosine kinases (RTK) by ligands expressed by cells of the activated tyrosine kinases (8). Tyrosine kinase inhibitors microenvironment or the HRS cells themselves (1–4). (TKI) are usually most effective in malignancies with Thus, HRS cells express 8 RTKs, platelet-derived growth mutation-activated tyrosine kinases such as lung cancer factor receptor (PDGFR) a, DDR2, EPHB1, RON, TRKA, with mutated EGF receptor (EGFR) treated with gefitinib TRKB, TIE1, and CSF1R, in varying patterns, and (9). However, several TKIs are also used for the treatment PDGFRa and TRKA have been shown to be activated in of cancers in which so far no tyrosine kinase–activating primary cases (5–7). As mutations affecting RTKs could mutations have been identified such as sorafenib in not be detected in HRS cell lines and as activating ligands advanced renal cell carcinoma and hepatocellular carcinoma (10–12). were frequently co-expressed in HRS cells or expressed by a surrounding cells in the infiltrate of primary Hodgkin In previous experiments, the PDGFR inhibitor imatinib and K-252a (5, 13), an inhibitor of TRKA, caused decreases of HRS cell proliferation indicating that also HRS cells could be sensitive to tyrosine kinase inhibition Authors' Affiliations: 1Gerhard-Domagk-Institute for Pathology, although so far no tyrosine kinase–activating mutations Westfalische€ Wilhelms-University, Munster;€ 2Department of Pathology, have been observed. Therefore, we chose 9 TKIs, either Johann Wolfgang Goethe-University, Frankfurt; and 3Department of Pathology, Justus-Liebig-University, Giessen, Germany already approved for cancer therapy or in late stages of clinical trials, to examine their effects on HRS cell lines. We Note: Supplementary data for this article are available at Molecular Cancer Therapeutics Online (http://mct.aacrjournals.org/). show here that sorafenib and lestaurtinib reduced HRS cell proliferation significantly at concentrations achiev- Corresponding Author: Andreas Brauninger,€ Department of Pathology, Justus-Liebig-University Giessen, Langhansstr. 10, Giessen 35392, able in patients, in line with recently published studies (14, Germany. Phone: 49-641-98541130; Fax: 49-641-98541139; E-mail: 15). As sorafenib induced cell death without activation of [email protected] the effector caspase-3 and as lestaurtinib, in contrast to a doi: 10.1158/1535-7163.MCT-12-0532 recently published study (15), affected cell growth but did 2012 American Association for Cancer Research. not induce cell death, we analyzed the targets, the impact

www.aacrjournals.org 173

Holz et al.

on activated signaling pathways, and the modes of cell experiments. Densitometric analysis was conducted with death induction and resistance of both TKIs. the open access image processing program ImageJ.

Materials and Methods Reverse transcription PCR and quantitative real- Cell lines time PCR HRS cell lines (HDLM-2, KM-H2, L-428, L-1236, and U- Total RNA was extracted with the RNeasy Plus Micro H01) were obtained from the DSMZ and cultured as Kit (Qiagen). The cDNA was synthesized using the 1st recommended. Cell line identity was determined by the Strand cDNA Synthesis Kit (Roche Applied Science). For DSMZ by STR DNA fingerprinting and ensured by con- quantitative PCR analysis, the Applied Biosystems Power tinuous culture for less than 6 months. Hematopoietic SYBR Green Master Mix System (Life Technologies) was neoplasm U-2932, SU-DHL-8, SU-DHL-10, KARPAS-422, used. Technical duplicates were conducted for each SU-DHL-1, KARPAS-299, DG-75, RAMOS, DAUDI, RAJI, sample. Primer sequences are given in Supplementary JURKAT, L-363, K-562 and HL-60, carcinoma MCF7, Table S2. Identity of all PCR products was verified by MDA-MB-468, A-549, HTB-56, HELA, and fibroblast sequencing. NIH-3T3 cell lines were obtained from several colleagues (R. Kuppers,€ Essen, Germany; C. Muller-Tidow,€ Munster,€ Affymetrix GeneChip Human 1.0 ST Array analysis Germany; A. Neumann, Munster,€ Germany). As they KM-H2 was incubated for 24 hours with 1 mmol/L were used only for comparisons of the antiproliferative lestaurtinib or dimethyl sulfoxide (DMSO), and total RNA effects of sorafenib and lestaurtinib, they were not authen- was extracted and analyzed with GeneChip Human 1.0 ST ticated by the authors. Numbers of viable cells were arrays. Processing of the microarrays and initial bioinfor- determined with a CASY TT cell counter as technical matic analyses were conducted by the Integrated Func- duplicates (Innovatis AG). tional Genomics Unit of the Interdisciplinary Center for Clinical Research at the Medical Faculty of the University MTT assay for metabolic activity of Munster€ (Munster,€ Germany). After scaling data were Cells were plated at 1 Â 105 cells/mL (L-428 and KM- imported into GeneSpringGX (Agilent Technologies) and H2) or 2 Â 105 cells/mL (L-1236) and incubated with the normalized using the default normalization methods. indicated drug or solvent alone. To quantify enzymatic Mean values of replicate gene arrays were used to deter- reductase activity corresponding to metabolic activity, the mine changes in mRNA expression by pairwise compar- In Vitro Toxicology Assay (Sigma-Aldrich) was used, and ison (P < 0.05) of lestaurtinib-treated samples to control the concentration of the dye converted from MTT was samples. To identify gene ontology (GO) categories, dif- measured in duplicate at 570 nm with a NanoDrop ferentially expressed genes (fold change > 2) were ana- (Peqlab). lyzed with GeneCodis (16). All data about the microarray analysis have been deposited at ArrayExpress under Apoptosis analysis by flow cytometry accession number E-MEXP-3565. For analysis of cell death induction after treatment with lestaurtinib or sorafenib, L-428, L-1236, and KM-H2 cells Results were stained with FITC-AnnexinV (BD Biosciences Phar- Sorafenib and lestaurtinib inhibit proliferation of mingen) and propidium iodide (PI; AppliChem) accord- HRS cell lines ing to the manufacturer’s instructions and analyzed by Screening of 9 TKIs revealed that only sorafenib and flow cytometry. lestaurtinib reduced proliferation of 3 of 5 HRS cell lines after 48 hours in a concentration-dependent manner and Immunoprecipitation of RTKs at concentrations achievable in patients (IC50 for sorafe- Cells were lysed in NP-40–based lysis buffer (Pierce Co- nib: 8 mmol/L for L-1236 and L-428; 5 mmol/L for KM-H2; Immunoprecipitaion Kit, Thermo Scientific) supplemen- IC50 for lestaurtinib: 10 nmol/L for L-1236; 300 nmol/L for ted with 2 mmol/L Na3VO4 and Complete Mini Protease L-428 and KM-H2; Fig. 1A–E and Supplementary Tables Inhibitor Cocktail (Roche Applied Science). Cell lysates S3 and S4; refs. 17, 18). At longer incubation times also, (250 mg protein) were incubated with 1 mg of anti-Ron b (C- HRS cell lines HDLM-2 and U-HO1 showed sensitivities 20) or anti-Trk (C-14) antibodies (Santa Cruz Biotechnol- comparable with other HRS cell lines (Fig. 1F and G), most ogy), and immune complexes were collected by binding to likely due to their longer doubling times. These sensitiv- Dynabeads Sheep anti-Rabbit IgG (Life Technologies). ities were in line with 2 recently published studies for HRS Proteins were eluted by boiling Dynabeads in SDS sample cell lines (14, 15). Incubation of other lymphoma, leukemia buffer. and carcinoma cell lines with 5 mmol/L sorafenib, a concentration at which other cell lines without RTK muta- Western blot analysis tions showed sensitivity (19–21), and 300 nmol/L lestaur- Western blot analysis was conducted as described in tinib, a concentration at which cell lines with FLT3 or JAK2 Supplementary Materials and methods with primary mutations were sensitive (22, 23), revealed that sensitiv- antibodies specified in Supplementary Table S1. Results ities toward the multikinase inhibitor sorafenib were shown here are representative for at least 2 independent comparable for all cell lines (Fig. 1F) and that HRS cell

174 Mol Cancer Ther; 12(2) February 2013 Molecular Cancer Therapeutics

Effects of Sorafenib and Lestaurtinib on Hodgkin Cell Lines

lines were less sensitive toward lestaurtinib than all other verified as targets in HRS cell lines, whereas TRK and tested cell lines (Fig. 1G). RON were identified as new targets of sorafenib and PDGFRa and RON as new targets of lestaurtinib. Cell death induction by sorafenib but not lestaurtinib in HRS cell lines Sorafenib and lestaurtinib reduce activation of As sorafenib and lestaurtinib inhibited growth of HRS multiple intracellular signaling pathways cell lines, we determined whether this was due to a To analyze the effect of TKIs on signaling pathways cytostatic effect or cell death. Therefore, we incubated usually activated by RTKs, Western blotting was con- L-428, L-1236, and KM-H2 cells with the TKIs and used ducted. Treatment with sorafenib caused the reduction Western blotting (WB) for active, cleaved forms of the of phosphorylation of JAK2, various STATs and RAF1 effector caspase-3 and flow cytometric analysis after in all 3 cell lines, and reduction of AKT1 in L-1236 AnnexinV/PI staining to detect apoptosis (Fig. 2). and KM-H2. However, in all instances, the reduced Sorafenib treatment lead to a substantial decrease of phosphorylation was accompanied by concomitant viability in 3 HRS cell lines as detected by AnnexinV/PI reduction of total protein amounts (Fig. 4A). Extracel- staining but not by Western blotting for activation of the lular signal–regulated kinase (ERK)1/2 were less phos- effector caspase-3 in L-428 and KM-H2. Lestaurtinib phorylated in all cell lines without decreases in total induced apoptosis in L-1236 as detected by AnnexinV/ protein amounts. PI staining and caspase-3 Western blotting but had, in Lestaurtinib caused decreases in phosphorylation of contradiction to previously reported findings (15), only a AKT1, JAK2, and RAF1 in all cell lines with concomitant cytostatic effect on L-428 and KM-H2 cells as no substan- reduction of RAF1 total protein amount (Fig. 4B). The tial apoptosis induction could be detected by either phosphorylation levels of STATs were reduced in L-428 means. and L-1236. Unexpectedly, the phosphorylation of ERK1/ 2 was increased in all 3 cell lines, and the phosphorylation Targets of sorafenib and lestaurtinib in HRS cell lines of STATs was constant in KM-H2 after lestaurtinib treat- To identify the targets of sorafenib and lestaurtinib in ment with elevated total protein amounts of JAK2 and HRS cell lines, we analyzed their effects on RTK phos- STATs in this cell line. phorylation via direct Western blotting or Western blotting after immunoprecipitation (IP) with antibodies Sorafenib provokes endoplasmic reticulum stress in against p-TRK, p-PDGFRa, and p-RON. The relative HRS cell lines phosphorylation of RTKs was calculated in relation to Sorafenib can cause endoplasmic reticulum (ER) stress the total amount of RTK detected by densitometry anal- with a general reduction of translation and caspase-3– ysis of representative Western blotting (Supplementary independent cell death in human leukemia cells (25), and Table S5). ER stress could thus also account for the reduction in total The effects of TKIs on TRKA (hereafter referred to as protein amounts of several signaling molecules and cas- TRK) were analyzed with L-1236 and HDLM-2, which pase-3–independent cell death in HRS cell lines. There- showed that when pretreated with sorafenib or lestaurti- fore, we analyzed the effect of sorafenib on different ER nib, b-NGF failed to induce TRK phosphorylation in both stress markers in HRS cell lines by Western blotting and cell lines (Fig. 3A, Supplementary Table S5). quantitative real-time PCR (qRT-PCR). The phosphory- In L-428, RON is significantly phosphorylated without lation of PERK, one of the ER resident proteins inducing application of its ligand, most likely due to the presence of the stress response (26), and the expression of genes a constitutively active splice variant lacking the fourth induced by ER stress, GADD34 and GADD153 (25–27), extracellular immunoglobulin plexin transcription domain were monitored (Fig. 5). (data not shown) (24). Sorafenib nearly completely abol- All 3 analyzed cell lines had phosphorylated PERK ished the phosphorylation of RON in L-428 (Fig. 3B), and when incubated with solvent alone, which increased upon also lestaurtinib had an inhibitory effect on the phosphor- incubation with 10 mmol/L sorafenib in L-428 and KM- ylation of RON (Fig. 3B and Supplementary Table S5). H2. Incubation with 10 mmol/L sorafenib increased In U-HO1, ligand-induced phosphorylation of PDGFRa GADD34 and GADD135 mRNA levels in all cell lines as was decreased by sorafenib and also by lestaurtinib (Fig. determined by qRT-PCR (7.5- to 16-fold and 14- to 37-fold 3C, Supplementary Table S5). Surprisingly, expression increases for GADD34 and GADD135, respectively). To and phosphorylation of PDGFRa was increased in KM- investigate whether the induction of ER stress by sorafe- H2 cells when treated with lestaurtinib (Fig. 3D), whereas nib was an HRS cell–specific effect, the expression of evaluation of the effects of sorafenib in KM-H2 was GADD34 and GADD153 was also analyzed in HDLM-2, difficult due to the low PDGFRa phosphorylation under U-HO1 and 2 non–Hodgkin lymphoma cell lines with normal culture conditions. After preincubation with les- comparable sensitivity toward the TKIs, L-363 and taurtinib, sorafenib caused a decrease of PDGFRa phos- RAMOS. In all cell lines, 24-hour incubation with 10 phorylation and protein amount (Fig. 3E). mmol/L sorafenib caused an increase of GADD34 and Taken together, the known targets of sorafenib and GADD135 mRNA levels comparable with the effect in lestaurtinib, PDGFRa and TRK, respectively, could be HRS cell lines (Supplementary Fig. S1).

www.aacrjournals.org Mol Cancer Ther; 12(2) February 2013 175

Holz et al.

Sorafenib (µmol/L)

Figure 1. Sensitivity of Hodgkin-Reed/Sternberg cell lines to TKIs. A, chemical structures of compounds used in this study. B, HRS cell lines L-1236 and L-428 (both expressing TRKA, RON, DDR2, and EPHB1), KM-H2 (expressing PDGFRa and DDR2), HDLM-2 (expressing TRKA, DDR2, and EPHB1), and U-HO1 (expressing PDGFRa but not TRKA and RON) were incubated with different TKIs. Percentage of viable cells after 48-hour treatment with 5 mmol/L TKI are shown as mean values of 2 experiments. Reductions below 80% were considered relevant. For lapatinib, sunitinib, and dasatinib, see Supplementary Table S4.

These observations indicate that sorafenib provokes ER HRS cell lines and increased expression of PDGFRa,JAK2, stress, which may cause cell death, in HRS and the other and STATs in KM-H2 indicated that the expression and cell lines tested. The induction of ER stress could thus be a activation of several signaling pathway constituents, which general effect of sorafenib. could support survival, increased upon AKT1 and JAK inhibition in HRS cell lines. To address this issue system- Lestaurtinib activates the alternative NF-kB pathway atically, we conducted differential genome-wide gene in HRS cell lines expression analysis after lestaurtinib treatment of KM-H2. Despite the dephosphorylation of the anti-apoptotic KM-H2 cells were incubated with 1 mmol/L lestaurtinib AKT1 and in contrast to a recent study (15), lestaurtinib or solvent for 24 hours, and differentially expressed did not induce apoptosis in L-428 and KM-H2 in our hands. transcripts were identiﬁed using replicates of Affymetrix The lestaurtinib induced ERK1/2 phosphorylation in all GeneChip Human 1.0 Arrays and the GeneSpringGX

176 Mol Cancer Ther; 12(2) February 2013 Molecular Cancer Therapeutics

Effects of Sorafenib and Lestaurtinib on Hodgkin Cell Lines

DF 5 µmol/L Sorafenib

Lestaurtinib (µmol/L)

EG 300 nmol/L Lestaurtinib

Lestaurtinib (µmol/L)

Figure 1. (Continued)C–E, for generation of dose–response curves for sorafenib and lestaurtinib, cells were incubated with the indicated concentrations of the respective TKIs for 72 hours and counted (C, sorafenib; D, lestaurtinib). In addition, the metabolic activity (E) of HRS cell lines treated with a concentration range of lestaurtinib for 72 hours was measured with an MTT assay. Mean values and SDs were calculated from 3 independent experiments. F and G, different lymphoma, leukemia, and carcinoma cell lines were incubated with 5 mmol/L sorafenib (F) or 300 nmol/L lestaurtinib (G) for 72 hours and counted. Shown are the mean values from 2 experiments.

software. Sixty-three genes were at least 2-fold upregu- lestaurtinib-induced upregulation of CD30L resulted also lated and 180 genes were at least 2-fold downregulated in NF-kB activation, we analyzed the expression of tran- (Supplementary Tables S6 and S7). Gene ontology scriptional targets of the classical NF-kB pathway analysis revealed the affiliation of most (116 of 180) down- (ICAM1, LTA, and BIRC3) by qRT-PCR, but observed no regulated genes to biologic processes linked to cell divi- increases in expression (data not shown). In addition, we sion and proliferation (Supplementary Table S8), and analyzed activation of the alternative NF-kB pathway by most of these genes enhance cell-cycle progression. The monitoring the accumulation of p52 by Western blotting. upregulated genes showed no strong affiliation to specific After 24-hour incubation with 1 mmol/L lestaurtinib, biologic processes. increased amounts of p52 were detected in all 3 cell lines We chose 7 upregulated genes that could enhance cell (Fig. 6B). survival (LIPH, PIM2, SOS2, STAT4, TNFSF8, TNFSF10, To investigate whether the lestaurtinib-induced upre- and TRAF3IP3) for further analysis by qRT-PCR with KM- gulation of CD30L was a more general mechanism, 2 H2, L-428, and L-1236 (Fig. 6A). For KM-H2, increases in further HRS and anaplastic large cell lymphoma (ALCL) RNA levels of all genes could be confirmed, showing the cell lines expressing also CD30 were analyzed by qRT- validity of the microarray analysis. The most distinctly PCR for CD30L mRNA changes after 24-hour treatment induced genes in all 3 cell lines were TNFSF8 (6- to 18- with 1 mmol/L lestaurtinib (Supplementary Fig. S2). fold), the CD30 ligand, and TNFSF10 (3- to 13-fold), the An increase of CD30L mRNA amounts could be observed TRAIL-R ligand. in one of the HRS cell lines but not in the second HRS CD30, a cell surface TNF receptor (TNFR) expressed by cell line, nor in the ALCL cell lines, indicating that the most HRS cells and HRS cell lines (28), has been shown to upregulation of CD30L is a HRS cell–specific resistance activate the MEK/ERK pathway and the classical and mechanism. alternative NF-kB pathways upon binding of its ligand Taken together, lestaurtinib induced upregulation of (29–31). The observed increases in ERK1/2 phosphoryla- CD30L and TRAIL in HRS cell lines and the former most tion upon lestaurtinib treatment are thus most likely due likely caused ERK1/2 phosphorylation and activation of to the CD30L upregulation. To determine whether the the alternative NF-kB pathway.

www.aacrjournals.org Mol Cancer Ther; 12(2) February 2013 177

Holz et al.

Caspase-3 Caspase-3 Caspase-3

Cleaved Cleaved Cleaved Caspase-3 Caspase-3 Caspase-3

N PPN

C 10 µmol/L Sorafenib

Figure 2. Assessment of cell death induction in Hodgkin-Reed/Sternberg cell lines by sorafenib and lestaurtinib. A, induction of cell death was monitored by Western blotting for caspase-3 and its active cleaved form after 24- and 72-hour (longer film exposure) treatment with 1 mmol/L lestaurtinib or 10 mmol/L sorafenib. B, flow cytometric analysis of AnnexinV-FITC and PI-stained cells was conducted after 24- and 72-hour incubation with DMSO, 1 mmol/L lestaurtinib, or 10 mmol/L sorafenib. C, representative scatter plots of an AnnexinV-FITC/PI flow cytometric analysis for L-428 treated with DMSO or sorafenib for 72 hours. Note that equal numbers of events were counted, most of which located close to the x-axis in the DMSO control scatter plot example. The percentage of cells stained for Annexin V and/or PI was calculated from 2 independent experiments.

TKIs can increase the effect of conventional nib alone, or the combination of one chemotherapeutic chemotherapeutic drugs with one TKI (Supplementary Fig. S3). The inhibitory The combination of conventional chemotherapeutic effects of vinblastine and doxorubicin were in general drugs with TKIs is a strategy to enhance the effect of enhanced by the simultaneous treatment with sorafenib cancer therapy (32). For example, the EGFR inhibitor or lestaurtinib compared with either TKI or vinblastine or erlotinib is added to conventional chemotherapy for doxorubicin alone. This effect was most evident for the patients with advanced-stage pancreatic cancer (33). combination of sorafenib with either conventional chemo- Therefore, we examined the effect of sorafenib and les- therapeutic drug in L-428 and KM-H2 cells. taurtinib on HRS cell lines in combination with drugs used in conventional Hodgkin lymphoma chemotherapy. Discussion The metabolic activity of L-428, L-1236, and KM-H2 cells Several different RTKs are expressed and activated in was measured (MTT assay) after 72-hour treatment with HRS cells of Hodgkin lymphoma (5), and the inhibition vinblastine and doxorubicin alone, sorafenib or lestaurti- of these RTKs with TKIs could be a novel option for

178 Mol Cancer Ther; 12(2) February 2013 Molecular Cancer Therapeutics

Effects of Sorafenib and Lestaurtinib on Hodgkin Cell Lines

1 µmol/L Sorafenib10 µmol/L Sorafenib B

µmol/L Lestaurtinib 1 10 µmol/L Sorafenib

Figure 3. Western blot analysis of RTK phosphorylation after incubation with sorafenib or lestaurtinib. A, Western blot analysis of immunoprecipitated RTK after 4-hour incubation of cells in 1% fetal calf serum to increase induction of RTK phosphorylation, 10-minute preincubation with TKIs sorafenib (10 mmol/L) or lestaurtinib (1 mmol/L) and stimulation with 25 ng/mL b-NGF for 2 minutes. B, Western blot analysis of immunoprecipitated RON after 24-hour treatment with 10 mmol/L sorafenib or 1 mmol/L lestaurtinib. C, Western blot analysis of PDGFRa after cells were serum-starved and pretreated with sorafenib or lestaurtinib as described for A and stimulated with 100 ng/mL PDGFAA for 2 minutes. D, Western blot analysis of PDGFRa in KM-H2 after 24-hour incubation with 1 mmol/L lestaurtinib or 10 mmol/L sorafenib. E, Western blot analysis of PDGFRa in KM-H2 cells pretreated with 1 mmol/L lestaurtinib for 24 hours and subsequent incubation with 10 mmol/L sorafenib for 10 minutes or 1 hour. All experiments were carried out at least twice. treatment. In a screening of 9 TKIs, we found that all HRS in line with a recently published study analyzing the effect cell lines tested were sensitive toward the TKIs sorafenib of the TKI on HRS cell lines (14). Sorafenib inhibited not and lestaurtinib at concentrations achievable in patients. only the activation of its known targets RAF1 and Sorafenib, a multikinase inhibitor, induced cell death of PDGFRa (35, 36) but also TRK and RON and thus at least L-428, L-1236, and KM-H2 after prolonged incubation. 2 of the kinases tested here in each HRS cell line, and, given Although other studies investigating sorafenib observed its broad effects on several different cell lines, likely also antiproliferative activity at low nanomolar concentrations other kinases not tested here. when tumor cell lines with mutation activated RTKs were The sorafenib-induced deactivation of intracellular sig- examined (34), the sensitivities observed for HRS cell lines naling molecules was often accompanied by losses of total were comparable with what we found for other cell lines amounts of these proteins, and sorafenib can cause ER and what was found in other studies, analyzing for exam- stress with reduced translation in leukemia cells (25). As ple carcinoma or other lymphoma cell lines (19–21), and is we observed typical features of ER stress also in the

www.aacrjournals.org Mol Cancer Ther; 12(2) February 2013 179

Holz et al.

Figure 4. Western blot analysis of intracellular signaling pathways after TKI treatment. Shown are Western blot analyses for lysates after 24-hour treatment with 10 mmol/L sorafenib (A) or 1 mmol/L lestaurtinib (B) compared with untreated cells. AKT1, JAK2, STATs 3, 5, and 6, RAF1, and ERK1/2 as well as their active phosphorylated forms were detected.

sorafenib-treated HRS cell lines, it is likely that the losses The sensitivity of HRS cells toward lestaurtinib is of total amounts of signaling molecules are due to ER comparable with what was reported for acute myeloid stress induced reduction in translation combined with leukemia cells dependent on mutant FLT3 activity (22) or their fast turnover rates. Furthermore, ER stress caused myeloproliferative neoplasm cell lines with the constitu- cell death in a caspase-3–independent way in leukemia tively active JAK2-mutant V617F (23). However, also all cells (25) and also the sorafenib induced cell death in L-428 other lymphoma, leukemia, and carcinoma cell lines test- and KM-H2 was independent of caspase-3 activation. We ed here were even more sensitive than the HRS cell lines, conclude that cell death in L-428 and KM-H2 might be which could be either due to the interaction of lestaurtinib induced by the same ER stress–dependent mechanisms as with a broad kinase target spectrum, as recently supposed in leukemia cells (25). As induction of ER stress markers in (37), or dependency of all tested cell lines on JAK/STAT response to sorafenib treatment was also observed in 2 signaling. Lestaurtinib was recently reported to induce other non–Hodgkin lymphoma cell lines, induction of ER caspase-3 activity in some HRS cell lines (15). Although stress by sorafenib may be a more general phenomenon. exhibiting strong cytostatic effects, we observed an

AB Figure 5. ER stress after sorafenib treatment. L-1236, L-428 and KM- H2 were treated with sorafenib. A, Western blot analysis of PERK phosphorylation after 24-hour (L-1236 and L-428) or 12-hour (KM-H2) incubation with 10 mmol/L sorafenib or the solvent DMSO. B, mRNA amounts of GADD34 and GADD153 measured by qRT-PCR, normalized to the housekeeping gene PPIA and shown as fold changes compared with control. L-1236, L-428 and KM-H2 cells were incubated with 10 mmol/L sorafenib for 24 hours. Mean values and SDs were calculated from 2 independent experiments.

180 Mol Cancer Ther; 12(2) February 2013 Molecular Cancer Therapeutics

Effects of Sorafenib and Lestaurtinib on Hodgkin Cell Lines

Figure 6. Changes of mRNA expression of selected genes in Hodgkin-Reed/Sternberg cell lines and activation of the alternative NF- kB pathway after lestaurtinib treatment. A, L-1236, L-428, and KM-H2 cells were incubated with 1 mmol/L lestaurtinib for 24 hours. mRNA amounts of LIPH, PIM2, SOS2, STAT4, TNFSF8, TNFSF10 (variant 1/2), TNFSF10 (variant 3), and TRAF3IP3 were measured by qRT-PCR, normalized to the housekeeping gene PPIA and are shown as fold changes compared with control. Values of samples with a Ct > 35 or undetectable ampliﬁcation for some of the DMSO control samples of L-428 and L-1236 for TNFSF8 and TNFSF10 were set to 35 for DDCt calculation and the given fold changes may thus be an underestimation. Mean values and SDs were calculated from 2 independent experiments. B, B Western blot analysis of p52 accumulation in HRS cell lines after incubation with 1 mmol/L lestaurtinib or DMSO for 24 hours.

induction of apoptosis by lestaurtinib only in the already is usually essential for HRS cell survival (42, 43), without preapoptotic L-1236, in contrast to the caspase-3 activa- inducing apoptosis, we conducted a systematic screen for tion in L-428 observed by Diaz and colleagues (15). upregulated constituents of other signaling pathways and Lestaurtinib inhibited activation of its known targets found that lestaurtinib caused signiﬁcant increases of TRK and JAK2 and phosphorylation of AKT1 in all and CD30L and TRAIL expression in all 3 analyzed HRS cell phosphorylation of several STATs in L-428 and L-1236. lines. Furthermore, it inhibited ligand-induced phosphoryla- The activation of CD30 by CD30L has previously been tion of PDGFRa in one cell line and reduced the consti- shown to cause ERK phosphorylation and NF-kB activa- tutive phosphorylation of RON in L-428. In KM-H2, tion in HRS cell lines (29–31), and, indeed, we observed however, lestaurtinib treatment caused increases in total increased ERK1/2 phosphorylation and increases in amounts of JAK2 and STATs, and the amounts of phos- the activity of the alternative NF-kB pathway upon les- phorylated STAT3 and 5 remained constant, despite com- taurtinib treatment in all analyzed HRS cell lines. TRAIL plete JAK2 dephosphorylation. We observed also usually induces apoptosis in tumor cells via the TRAIL-R– increases in expression and phosphorylation of PDGFRa associated DISC. However, in HRS cells, apoptosis induc- in KM-H2 upon lestaurtinib treatment, and thus it is tion via TRAIL-R–associated caspase-8 is prevented by conceivable that the activated PDGFRa directly or via the strong c-FLIP expression (44, 45), and in cancer cells activation of other intracellular tyrosine kinases, for exam- resistant toward TRAIL-induced apoptosis, TRAIL ple, SRC family kinases, contributed to the STAT phos- increased survival in a NF-kB–dependent fashion (46). It phorylation under lestaurtinib treatment (38). Such main- is thus conceivable that in HRS cell lines, the lestaurtinib tenance of activation of a signaling pathway on which induced increases in TRAIL expression also contribute to tumor cells are "dependent" by increased expression and NF-kB activation. Furthermore, activation of the alterna- activation of another tyrosine kinase upon pharmacologic tive NF-kB pathway may also be increased by the inhibition of the "driver" tyrosine kinase is a known observed TRAF3IP3 upregulation, which by releasing mechanism of secondary TKI resistance. NIK from TRAF3 may cause activation of IKKa with Another way of tumor cells to escape the pharmacologic subsequent p100 processing (47, 48). inhibition of antiapoptotic signaling pathways is the acti- It is thus likely that in HRS cells, the inhibitory effects of vation of alternative pathways (39–41). As lestaurtinib lestaurtinib on the survival enhancing AKT1 are compen- treatment caused deactivation of AKT1, whose activity sated by activation of CD30 signaling via CD30L

www.aacrjournals.org Mol Cancer Ther; 12(2) February 2013 181

Holz et al.

upregulation and activation of ERK1/2 and NF-kB and conventional chemotherapeutics had additive effects, also by TRAIL-mediated NF-kB activation. As the com- likely due to the induction of cell death by 2 different bined application of the MEK/ERK inhibitor PD0325901 mechanisms. This indicates a potential use of sorafenib in with lestaurtinib did not result in apoptosis (data not Hodgkin lymphoma therapy in combination with con- shown), activation of NF-kB pathway may be the more ventional chemotherapeutic drugs. Regarding the effects relevant mechanism of apoptosis prevention. Interesting- of lestaurtinib, our results indicate that HRS cells can ly, this effect seems to be HRS cell–specific, as no lestaur- escape apoptosis by TKI-mediated AKT1 inhibition via tinib-induced upregulation of CD30L transcription was increases in the TNFR-mediated activation of the alterna- observed in other CD30-expressing cell lines. tive NF-kB pathway. This mode of TKI resistance differs The addition of TKIs to cancer treatment with conven- from the usually observed upregulation of other tyrosine tional chemotherapeutic drugs has beneficial effects in kinases or increased signaling via other tyrosine kinase- several instances. Therefore, we combined lestaurtinib activated pathways and seems to be an HRS cell–specific and sorafenib with doxorubicin and vinblastine and resistance mechanism. indeed observed increased inhibitory effects on metabolic activity compared with the single agents. Disclosure of Potential Conflicts of Interest The combination of conventional chemotherapeutic No potential conflicts of interest were disclosed. drugs with sorafenib showed additive inhibitory effects in those cell lines possibly undergoing completely cas- Authors' Contributions Conception and design: M.S. Holz, C. Renne, A. Brauninger€ pase-3–independent cell death in response to sorafenib, L- Development of methodology: M.S. Holz, A. Janning 428 and KM-H2. Apoptosis via caspase-3 is one major cell Acquisition of data (provided animals, acquired and managed patients, death pathway caused by conventional chemotherapeutic provided facilities, etc.): M.S. Holz, A. Janning, T. Spieker Analysis and interpretation of data (e.g., statistical analysis, biostatis- agents (49). Potentially, the combination of the 2 drugs is tics, computational analysis): M.S. Holz, C. Renne, A. Brauninger€ most efficient when different mechanisms of cell death are Writing, review, and/or revision of the manuscript: M.S. Holz, C. Renne, € triggered, the apoptotic pathway through caspase-3 by a A. Brauninger Administrative, technical, or material support (i.e., reporting or orga- conventional chemotherapeutic agent and caspase-inde- nizing data, constructing databases): S. Gattenlohner€ pendent cell death by sorafenib. Combining lestaurtinib with conventional chemother- Acknowledgments apeutic drugs did not significantly alter the effect of The authors thank Inka Buchroth and Sandra Muhleis€ for excellent lestaurtinib alone in L-428 and KM-H2. Here, lestaurtinib technical assistance. mediated NF-kB activation may provide protection from moderate concentrations of conventional chemothera- Grant Support This work was supported by the Deutsche Forschungsgemeinschaft peutics. (grant no. BR1238/6-3 to A. Brauninger).€ In summary, we show here that lestaurtinib inhibits The costs of publication of this article were defrayed in part by the proliferation, and sorafenib induces cell death in several payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate HRS cell lines. The sorafenib-induced cell death was this fact. caspase-3 independent, in line with the induction of ER stress, which can cause cell death via caspase-3–indepen- Received May 25, 2012; revised October 12, 2012; accepted November 5, dent mechanisms. The combination of sorafenib with 2012; published OnlineFirst December 13, 2012.

References 1. Carbone A, Gloghini A, Gruss HJ, Pinto A. CD40 ligand is constitutively 7. Lamprecht B, Walter K, Kreher S, Kumar R, Hummel M, Lenze D, et al. expressed in a subset of T cell lymphomas and on the microenviron- Derepression of an endogenous long terminal repeat activates the mental reactive T cells of follicular lymphomas and Hodgkin's disease. CSF1R proto-oncogene in human lymphoma. Nat Med 2010;16: Am J Pathol 1995;147:912–22. 571–9. 2. Jucker M, Abts H, Li W, Schindler R, Merz H, Gunther A, et al. 8. Noble ME, Endicott JA, Johnson LN. Protein kinase inhibitors: insights Expression of interleukin-6 and interleukin-6 receptor in Hodgkin's into drug design from structure. Science 2004;303:1800–5. disease. Blood 1991;77:2413–8. 9. Herbst RS, Fukuoka M, Baselga J. Geﬁtinib–a novel targeted approach 3. Chiu A, Xu W, He B, Dillon SR, Gross JA, Sievers E, et al. Hodgkin to treating cancer. Nat Rev Cancer 2004;4:956–65. lymphoma cells express TACI and BCMA receptors and generate 10. Bracarda S, Caserta C, Sordini L, Rossi M, Hamzay A, Crino L. Protein survival and proliferation signals in response to BAFF and APRIL. kinase inhibitors in the treatment of renal cell carcinoma: sorafenib. Blood 2007;109:729–39. Ann Oncol 2007;18 Suppl 6:vi22–5. 4. Blume-Jensen P, Hunter T. Oncogenic kinase signalling. Nature 2001; 11. Albiges L, Oudard S, Negrier S, Caty A, Gravis G, Joly F, et al. Complete 411:355–65. remission with tyrosine kinase inhibitors in renal cell carcinoma. J Clin 5. Renne C, Willenbrock K, Kuppers€ R, Hansmann ML, Brauninger€ A. Oncol 2012;30:482–7. Autocrine- and paracrine-activated receptor tyrosine kinases in classic 12. Llovet JM, Ricci S, Mazzaferro V, Hilgard P, Gane E, Blanc JF, et al. Hodgkin lymphoma. Blood 2005;105:4051–9. Sorafenib in advanced hepatocellular carcinoma. N Engl J Med 6. Renne C, Willenbrock K, Martin-Subero JI, Hinsch N, Doring C, 2008;359:378–90. Tiacci E, et al. High expression of several tyrosine kinases and 13. Renne C, Minner S, Kuppers€ R, Hansmann ML, Brauninger€ A. activation of the PI3K/AKT pathway in mediastinal large B cell Autocrine NGFbeta/TRKA signalling is an important survival factor lymphoma reveals further similarities to Hodgkin lymphoma. Leuke- for Hodgkin lymphoma derived cell lines. Leuk Res 2008;32: mia 2007;21:780–7. 163–7.

182 Mol Cancer Ther; 12(2) February 2013 Molecular Cancer Therapeutics

Effects of Sorafenib and Lestaurtinib on Hodgkin Cell Lines

14. Ullrich K, Wurster KD, Lamprecht B, Kochert K, Engert A, Dorken B, activity in Hodgkin's lymphoma cells. Mol Immunol 2009;46: et al. BAY 43–9006/Sorafenib blocks CSF1R activity and induces 2441–8. apoptosis in various classical Hodgkin lymphoma cell lines. Br 32. Takimoto CH, Awada A. Safety and anti-tumor activity of sorafenib J Haematol 2011;155:398–402. (Nexavar) in combination with other anti-cancer agents: a review of 15. Diaz T, Navarro A, Ferrer G, Gel B, Gaya A, Artells R, et al. Lestaurtinib clinical trials. Cancer Chemother Pharmacol 2008;61:535–48. inhibition of the Jak/STAT signaling pathway in Hodgkin lymphoma 33. Bareschino MA, Schettino C, Troiani T, Martinelli E, Morgillo F, Ciar- inhibits proliferation and induces apoptosis. PLoS One 2011;6: diello F. Erlotinib in cancer treatment. Ann Oncol 2007;18 Suppl 6: e18856. vi35–41. 16. Carmona-Saez P, Chagoyen M, Tirado F, Carazo JM, Pascual-Mon- 34. Lierman E, Lahortiga I, Van Miegroet H, Mentens N, Marynen P, Cools tano A. GENECODIS: a web-based tool for finding significant concur- J. The ability of sorafenib to inhibit oncogenic PDGFRbeta and FLT3 rent annotations in gene lists. Genome Biol 2007;8:R3. mutants and overcome resistance to other small molecule inhibitors. 17. Strumberg D, Clark JW, Awada A, Moore MJ, Richly H, Hendlisz A, Haematologica 2007;92:27–34. et al. Safety, pharmacokinetics, and preliminary antitumor activity of 35. von Bubnoff N, Gorantla SP, Engh RA, Oliveira TM, Thone S, Aberg E, sorafenib: a review of four phase I trials in patients with advanced et al. The low frequency of clinical resistance to PDGFR inhibitors in refractory solid tumors. Oncologist 2007;12:426–37. myeloid neoplasms with abnormalities of PDGFRA might be related to 18. Marshall JL, Kindler H, Deeken J, Bhargava P, Vogelzang NJ, Rizvi N, the limited repertoire of possible PDGFRA kinase domain mutations in et al. Phase I trial of orally administered CEP-701, a novel neurotrophin vitro. Oncogene 2011;30:933–43. receptor-linked tyrosine kinase inhibitor. Invest New Drugs 2005; 36. Lierman E, Folens C, Stover EH, Mentens N, Van Miegroet H, Scheers 23:31–7. W, et al. Sorafenib is a potent inhibitor of FIP1L1-PDGFRalpha and the 19. Liu L, Cao Y, Chen C, Zhang X, McNabola A, Wilkie D, et al. Sorafenib imatinib-resistant FIP1L1-PDGFRalpha T674I mutant. Blood 2006; blocks the RAF/MEK/ERK pathway, inhibits tumor angiogenesis, and 108:1374–76. induces tumor cell apoptosis in hepatocellular carcinoma model PLC/ 37. Davis MI, Hunt JP, Herrgard S, Ciceri P, Wodicka LM, Pallares G, et al. PRF/5. Cancer Res 2006;66:11851–8. Comprehensive analysis of kinase inhibitor selectivity. Nat Biotechnol 20. Takezawa K, Okamoto I, Yonesaka K, Hatashita E, Yamada Y, Fukuoka 2011;29:1046–51. M, et al. Sorafenib inhibits non-small cell lung cancer cell growth by 38. Wang YZ, Wharton W, Garcia R, Kraker A, Jove R, Pledger WJ. Acti- targeting B-RAF in KRAS wild-type cells and C-RAF in KRAS mutant vation of Stat3 preassembled with platelet-derived growth factor beta cells. Cancer Res 2009;69:6515–21. receptors requires Src kinase activity. Oncogene 2000;19:2075–85. 21. Ramakrishnan V, Timm M, Haug JL, Kimlinger TK, Halling T, Wellik LE, 39. Carracedo A, Ma L, Teruya-Feldstein J, Rojo F, Salmena L, Alimonti A, et al. Sorafenib, a multikinase inhibitor, is effective in vitro against non- et al. Inhibition of mTORC1 leads to MAPK pathway activation through Hodgkin lymphoma and synergizes with the mTOR inhibitor rapamy- a PI3K-dependent feedback loop in human cancer. J Clin Invest cin. Am J Hematol 2012;87:277–83. 2008;118:3065–74. 22. Brown P, Meshinchi S, Levis M, Alonzo TA, Gerbing R, Lange B, 40. Chandarlapaty S, Sawai A, Scaltriti M, Rodrik-Outmezguine V, Grbo- et al. Pediatric AML primary samples with FLT3/ITD mutations vic-Huezo O, Serra V, et al. AKT inhibition relieves feedback suppres- are preferentially killed by FLT3 inhibition. Blood 2004;104: sion of receptor tyrosine kinase expression and activity. Cancer Cell 1841–9. 2011;19:58–71. 23. Hexner EO, Serdikoff C, Jan M, Swider CR, Robinson C, Yang S, et al. 41. Nguyen TK, Jordan N, Friedberg J, Fisher RI, Dent P, Grant S. Inhibition Lestaurtinib (CEP701) is a JAK2 inhibitor that suppresses JAK2/ of MEK/ERK1/2 sensitizes lymphoma cells to sorafenib-induced apo- STAT5 signaling and the proliferation of primary erythroid cells ptosis. Leuk Res 2010;34:379–86. from patients with myeloproliferative disorders. Blood 2008;111: 42. Dutton A, Reynolds GM, Dawson CW, Young LS, Murray PG. Con- 5663–71. stitutive activation of phosphatidyl-inositide 3 kinase contributes to the 24. Collesi C, Santoro MM, Gaudino G, Comoglio PM. A splicing variant of survival of Hodgkin's lymphoma cells through a mechanism involving the RON transcript induces constitutive tyrosine kinase activity and an Akt kinase and mTOR. J Pathol 2005;205:498–506. invasive phenotype. Mol Cell Biol 1996;16:5518–26. 43. Georgakis GV, Li Y, Rassidakis GZ, Medeiros LJ, Mills GB, Younes A. 25. Rahmani M, Davis EM, Crabtree TR, Habibi JR, Nguyen TK, Dent P, Inhibition of the phosphatidylinositol-3 kinase/Akt promotes G1 cell et al. The kinase inhibitor sorafenib induces cell death through a cycle arrest and apoptosis in Hodgkin lymphoma. Br J Haematol process involving induction of endoplasmic reticulum stress. Mol Cell 2006;132:503–11. Biol 2007;27:5499–513. 44. Mathas S, Lietz A, Anagnostopoulos I, Hummel F, Wiesner B, Janz M, 26. Novoa I, Zhang Y, Zeng H, Jungreis R, Harding HP, Ron D. Stress- et al. c-FLIP mediates resistance of Hodgkin/Reed-Sternberg cells to induced gene expression requires programmed recovery from trans- death receptor-induced apoptosis. J Exp Med 2004;199:1041–52. lational repression. EMBO J 2003;22:1180–7. 45. van Houdt IS, Muris JJ, Hesselink AT, Kramer D, Cillessen SA, Moes- 27. Wang XZ, Lawson B, Brewer JW, Zinszner H, Sanjay A, Mi LJ, et al. bergen LM, et al. Expression of c-FLIP is primarily detected in diffuse Signals from the stressed endoplasmic reticulum induce C/EBP- large B-cell lymphoma and Hodgkin's lymphoma and correlates with homologous protein (CHOP/GADD153). Mol Cell Biol 1996;16: lack of caspase 8 activation. Histopathology 2007;51:778–84. 4273–80. 46. Ehrhardt H, Fulda S, Schmid I, Hiscott J, Debatin KM, Jeremias I. TRAIL 28. Drexler HG, Minowada J. Hodgkin's disease derived cell lines: a induced survival and proliferation in cancer cells resistant towards review. Hum Cell 1992;5:42–53. TRAIL-induced apoptosis mediated by NF-kappaB. Oncogene 2003; 29. Zheng B, Fiumara P, Li YV, Georgakis G, Snell V, Younes M, et al. MEK/ 22:3842–52. ERK pathway is aberrantly active in Hodgkin disease: a signaling 47. Dadgostar H, Doyle SE, Shahangian A, Garcia DE, Cheng G. T3JAM, a pathway shared by CD30, CD40, and RANK that regulates cell pro- novel protein that specifically interacts with TRAF3 and promotes the liferation and survival. Blood 2003;102:1019–27. activation of JNK(1). FEBS Lett 2003;553:403–7. 30. Wright CW, Rumble JM, Duckett CS. CD30 activates both the canon- 48. He JQ, Saha SK, Kang JR, Zarnegar B, Cheng G. Specificity of TRAF3 ical and alternative NF-kappaB pathways in anaplastic large cell in its negative regulation of the noncanonical NF-kappa B pathway. J lymphoma cells. J Biol Chem 2007;282:10252–62. Biol Chem 2007;282:3688–94. 31. Guo F, Sun A, Wang W, He J, Hou J, ZhouP,etal.TRAF1isinvolved 49. Debatin KM. Apoptosis pathways in cancer and cancer therapy. in the classical NF-kappaB activation and CD30-induced alternative Cancer Immunol Immunother 2004;53:153–9.

www.aacrjournals.org Mol Cancer Ther; 12(2) February 2013 183

Induction of Endoplasmic Reticulum Stress by Sorafenib and Activation of NF- κB by Lestaurtinib as a Novel Resistance Mechanism in Hodgkin Lymphoma Cell Lines

Meike Stefanie Holz, Angela Janning, Christoph Renné, et al.

Mol Cancer Ther 2013;12:173-183. Published OnlineFirst December 13, 2012.

Updated version Access the most recent version of this article at: doi:10.1158/1535-7163.MCT-12-0532

Supplementary Access the most recent supplemental material at: Material http://mct.aacrjournals.org/content/suppl/2012/12/13/1535-7163.MCT-12-0532.DC1

Cited articles This article cites 49 articles, 20 of which you can access for free at: http://mct.aacrjournals.org/content/12/2/173.full#ref-list-1

Citing articles This article has been cited by 1 HighWire-hosted articles. Access the articles at: http://mct.aacrjournals.org/content/12/2/173.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at Subscriptions [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mct.aacrjournals.org/content/12/2/173. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.