Hypoxia Enhances Sphingosine Kinase 2 Activity and Provokes Sphingosine-1-Phosphate-Mediated Chemoresistance in A549lung Cancer Cells

Published OnlineFirst February 24, 2009; DOI: 10.1158/1541-7786.MCR-08-0156 Published Online First on February 24, 2009

Hypoxia Enhances Sphingosine Kinase 2 Activity and Provokes Sphingosine-1-Phosphate-Mediated Chemoresistance in A549Lung Cancer Cells

Steffen E. Schnitzer, Andreas Weigert, Jie Zhou, and Bernhard Bru¨ne, Goethe-University, Faculty of Medicine, Institute of Biochemistry I/ZAFES, Frankfurt am Main, Germany

Abstract sclerosis, myocardial infarction, or inflammation. Furthermore, Hypoxia and signaling via hypoxia-inducible factor-1 the occurrence of hypoxic regions is a characteristic feature of (HIF-1) is a key feature of solid tumors and is related to solid tumors and is often associated with poor prognosis (1-3). tumor progression as well as treatment failure. Although To circumvent periods of insufficient nutrient and oxygen it is generally accepted that HIF-1 provokes tumor cell supply, cells evolve mechanisms to adapt to such conditions, survival and induces chemoresistance under hypoxia, with the notion that hypoxia-inducible factor-1 (HIF-1) appears HIF-1-independent mechanisms operate as well. We of particular importance. HIF-1 regulates the expression of present evidence that conditioned medium obtained >70 target genes, including proapoptotic ones (4), but also from A549cells, incubated for 24 h under hypoxia, provokes an antiapoptotic phenotype of tumor cells (5-8). In protected naive A549cells from etoposide-induced addition, there are HIF-1-independent pathways that promote cell death. Lipid extracts generated from hypoxia- cell survival under hypoxia by activating, for example, conditioned medium still rescued cells from apoptosis phosphoinositide 3-kinase (PI3K) or extracellular signal- induced by etoposide. Specifically, the bioactive lipid regulated kinase 1/2 [p42/44 mitogen-activated protein kinase sphingosine-1-phosphate (S1P) not only was essential (MAPK; ref. 9)]. for cell viability of A549cells but also protected cells Considerable attention has also been given to signaling from apoptosis. We noticed an increase in sphingosine properties of bioactive lipids, such as ceramide and sphingosine, kinase 2 (SphK2) protein level and enzymatic activity as well as their metabolites ceramide-1-phosphate and sphingo- under hypoxia, which correlated with the release of S1P sine-1-phosphate (S1P). The term ‘‘sphingolipid rheostat’’ into the medium. Knockdown of SphK2 using specific defines the ratio between ceramide/sphingosine and S1P as a small interfering RNA relieved chemoresistance of A549 determinant for survival and death (10). Whereas ceramide and cells under hypoxia and conditioned medium obtained sphingosine induce apoptosis in various cell types (11, 12), S1P from SphK2 knockdown cells was only partially enhances cell survival (13). S1P is produced during sphingolipid protective. Coincubations of conditioned medium with metabolism via the conversion of ceramide to sphingosine,

VPC23019, a S1P1/S1P3 antagonist, reduced protection which is then phosphorylated to S1P by sphingosine kinases of conditioned medium, with the further notion that p42/ (SphK). In mammals, two SphK isoforms have been identified, 44 mitogen-activated protein kinase transmits autocrine SphK1 and SphK2. Although generating the same product, both or paracrine survival signaling downstream of S1P1/S1P3 enzymes show opposed biological functions (14). Whereas receptors. Our data suggest that hypoxia activates SphK1 promotes survival, vascularization, and angiogenesis SphK2 to promote the synthesis and release of S1P, (15, 16), SphK2 activity, mainly based on forced overexpression which in turn binds to S1P1/S1P3 receptors, thus studies, is linked to a proapoptotic outcome (17, 18). Although activating p42/44 mitogen-activated protein kinase to SphK1 is affected by growth factors or cytokines, activity convey autocrine or paracrine protection of A549cells. regulation of SphKs is not fully understood (19-22). SphK1 (Mol Cancer Res 2009;7(3):OF1–9) contains several putative phosphorylation sites and phosphorylation of Ser225 by p42/44 MAPK and cyclin-dependent kinase Introduction 2 enhanced SphK1 activity on tumor necrosis factor-a or Tissue hypoxia provokes a decrease in nutrient supply and is phorbol 12-myristate 13-acetate treatment (23, 24). On phos- associated with pathologic conditions such as stroke, athero- phorylation, SphK1 translocates to the plasma membrane, where it gets in close proximity to its substrate. Activation of SphK2 in Received 3/27/08; revised 10/24/08; accepted 11/10/08; published OnlineFirst 02/24/2009. response to growth factors or phorbol 12-myristate 13-acetate Grant support: Deutsche Forschungsgemeinschaft (BR999, FOG784), Sander via activation of extracellular signal-regulated kinase 1 and Foundation, Deutsche Krebshilfe, EU (PROLIGEN), and LOEWE (LiFF). phosphorylation at Ser351 and Thr578 has also been shown (25). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in Ding et al. identified two more serine residues within SphK2 accordance with 18 U.S.C. Section 1734 solely to indicate this fact. (Ser419 and Ser421) that became phosphorylated by protein Note: S.E. Schnitzer and A. Weigert contributed equally to this work. kinase D, which fostered the nuclear export of SphK2 as an Requests for reprints: Bernhard Bru¨ne, Goethe-University, Faculty of Medicine, Institute of Biochemistry I/ZAFES, Theodor-Stern-Kai 7, 60590 Frankfurt alternative regulatory circuit (26). am Main, Germany. Phone: 49-69-6301-7423; Fax: 49-69-6301-4203. E-mail: Although both SphK enzymes are located intracellularly and [email protected] Copyright D 2009 American Association for Cancer Research. first reports on signaling properties of S1P described intracel- doi:10.1158/1541-7786.MCR-08-0156 lular effects, the identification of direct S1P targets awaits

Mol Cancer Res 2009;7(3). March 2009 OF1 Downloaded from mcr.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Published OnlineFirst February 24, 2009; DOI: 10.1158/1541-7786.MCR-08-0156

OF2 Schnitzer et al.

further clarification (15, 27, 28). Besides, S1P is exported from In continuing experiments, we collected medium from A549 the cell by members of the ATP-binding cassette transporter cells incubated for 24 h under normoxia [CM(normoxia)] or family. Once secreted, S1P binds to one of its five G-protein- hypoxia [CM(hypoxia)] and stimulated naive cells with these coupled receptors (S1P1-S1P1), activating downstream path- conditioned media for 24 h under normoxic conditions. We ways such as p38 MAPK, p42/44 MAPK, PI3K, c-Jun noticed that CM(hypoxia) but not CM(normoxia) protected 2+ NH2-terminal kinase, or Ca signaling (29-31). A549 cells from etoposide-induced cell death (Fig. 1C). To More recently, the role of lipids in cell survival under elucidate the nature of the transferable survival factor secreted hypoxia/ischemia was discussed. Two reports showed activa- under hypoxia, we first excluded proteins being involved by tion of SphK1 and the release of S1P under hypoxia (32, 33). treating CM(hypoxia) with 50 Amol/L proteinase K (1 h, 37jC) Anelli et al. identified HIF-2 as a positive regulator of SphK1 followed by protein denaturation (2 h, 100jC). This treatment expression, whereas HIF-1 acted as a competitive inhibitor (33). did not eradicate protection induced by CM(hypoxia) (Fig. 1C). In opposite, Schwalm et al. suggested that both HIF-1 and Taking into account that lipid mediators are associated with cell HIF-2 act in concert to induce SphK1 protein under hypoxic protection, we extracted lipids from conditioned medium and conditions (32). New evidence suggests that SphK1 activity tested for their cytoprotective potential. Lipid extracts from under hypoxia is mediated by reactive oxygen species (34). normoxia-conditioned medium [Lip-CM(normoxia)] did not Irrespective of differences on mechanistic details, increased protect cells from etoposide-induced cell death. However, lipid expression and activation of SphKs and subsequent formation extracts from CM(hypoxia) [Lip-CM(hypoxia)] strongly re- of S1P has been correlated with enhanced survival (29), and an duced caspase-3 activity on etoposide treatment (Fig. 1C). increased synthesis of S1P has been reported under hypoxic These observations suggest that CM(hypoxia) contains a lipid conditions (35). Additionally, N,N-dimethylsphingosine molecule that provokes chemoresistance. (DMS), an inhibitor of both SphK isoforms, was used to To identify the nature of the lipid present in CM(hypoxia), attenuate proliferation of smooth muscle cells under hypoxia we screened several potential candidates described to interfere (35). Moreover, mouse cardiac myocytes lacking functional with apoptosis. Treatment with lysophosphatidic acid SphK1 showed a reduced survival under hypoxia due to a (10 Amol/L), platelet-activating factor (1 Amol/L), or ceram- diminished S1P release, which underscores the importance of ide-1-phosphate (1 Amol/L) for 24 h under normoxia failed to SphK activity for cell survival under hypoxia (16). protect A549 cells from etoposide-induced cell death (Fig. 2A). Our work provides evidence that hypoxia caused an increase Interestingly, S1P (10 Amol/L) reduced caspase-3 activity by in SphK2 protein expression and activity, which provoked the f35% compared with etoposide treatments (Fig. 2A). We formation and release of S1P. Extracellular S1P acted as a further showed that 0.1 to 10 Amol/L S1P dose-dependently transferable survival factor, attenuating induction of apoptosis protected A549 cells from cell death induced by etoposide via binding to S1P1/S1P3 receptors and activation of p42/44 (Fig. 2B). To ascertain whether CM(hypoxia) or Lip-CM(hy- MAPK signaling in A549 cells. poxia) contain S1P, we used 10 Amol/L DMS, an inhibitor of SphK isoenzymes, which is known to reduce cell viability Results at high concentrations, and coincubated DMS with S1P, Previous work led to the identification of a chemoresistant Lip-CM(hypoxia), or CM(hypoxia). As expected, stimulation phenotype in A549 lung carcinoma cells when subjected to of cells with 10 Amol/L DMS for 24 h under normoxia caused hypoxia. Cells showed an abrogated caspase-3 activation and cell death followed by caspase-3 activation (Fig. 2C). Induction chromatin condensation, two hallmarks of apoptosis, in of caspase-3 activity could be prevented by coincubating DMS response to etoposide treatment (5). Stimulation of A549 cells with 10 Amol/L S1P, CM(hypoxia), or Lip-CM(hypoxia) with 150 Amol/L etoposide for 4 h under serum-free conditions (Fig. 2C). These results argue for a pivotal role of S1P in provoked massive cell death followed by caspase-3 activity preserving cell viability and furthermore point to S1P being measurements (specific activity of caspase-3: normoxia, present in CM(hypoxia) or lipid extracts [Lip-CM(hypoxia)]. 2.1 F 1.5 nmol/mg/min; normoxia + etoposide, 68.4 F 18.0 A prerequisite for S1P synthesis is the expression and/or nmol/mg/min). Induction of apoptosis was reduced by f75%, activation of at least one SphK isoform under hypoxia. Because when cells were preincubated for 24 h under hypoxia (specific DMS is an inhibitor of both SphK isoenzymes, we went on to activity of caspase-3; 16.4 F 3.1 nmol/mg/min; Fig. 1A, black identify which isoform might be responsible for hypoxic columns). To exclude drug-specific effects, we replaced protection. Regulation of SphK1 under hypoxia has been etoposide and used staurosporine to induce apoptosis. Treat- correlated to HIF activity, and due to the fact that the promoter ment of A549 cells with 1 Ag/mL staurosporine caused of SphK2 also contains putative hypoxia response elements, we significant caspase-3 activation, which could be prevented by incubated cells under hypoxia or in the presence of the hypoxia- preincubating cells under hypoxia for 24 h (Fig. 1A, white mimetic CoCl2 to test whether HIF transcription factors columns). To rule out that hypoxia-mediated protection from influence mRNA expression of SphKs. In A549 cells, both drug-induced cell death is cell type specific, we confirmed isoforms are expressed at a basal rate, but neither 100 Amol/L mechanisms of hypoxia-induced chemoresistance using human CoCl2 nor 4 h hypoxia caused changes in mRNA transcripts of THP-1 macrophages. Treatment of THP-1 macrophages with SphK1 (Fig. 3A, white columns) or SphK2 (Fig. 3A, black 1 Ag/mL staurosporine for 8 h under serum-free conditions columns). Interestingly, hypoxia (1-4 h) elevated the protein caused caspase-3 activation (normalized to 100%), which was level of SphK2, whereas the protein amount of SphK1 was not reduced by f40% after preincubation of cells for 24 h under induced (Fig. 3B). The expression of HIF-1a served as a hypoxia (Fig. 1B). marker for cellular hypoxia. Blocking protein synthesis with

Mol Cancer Res 2009;7(3). March 2009 Downloaded from mcr.aacrjournals.org on October 2, 2021. © 2009 American Association for Cancer Research. Published OnlineFirst February 24, 2009; DOI: 10.1158/1541-7786.MCR-08-0156

Hypoxia-Induced S1P Mediates Chemoresistance OF3

FIGURE 1. Protection from drug-induced cell death by hypoxia or conditioned medium. A. Measurement of caspase-3 activity in A549 cells incubated for 24 h under normoxia or hypoxia followed by the addition of etoposide (black columns) or staurosporine (white columns). Caspase-3 activity was normalized to untreated normoxic controls and relative induction of caspase-3 activity was determined. B. Measurement of caspase-3 activity in THP-1 macrophages. Cells were incubated under normoxia or hypoxia for 24 h followed by the induction of apoptosis with staurosporine. Caspase-3 activity of cells treated with staurosporine under normoxia was set to 100% and relative protection from apoptosis was determined. C. Measurement of caspase-3 activity in A549 cells pretreated with normoxia-conditioned medium [CM(normoxia)], hypoxia-conditioned medium [CM(hypoxia)], either untreated or incubated with proteinase K followed by heat inactivation, or lipid extracts [Lip-CM(normoxia) and Lip-CM(hypoxia)]. Incubations were done under normoxia for 24 h followed by etoposide treatment. Caspase-3 activity of cells treated with etoposide under normoxia was set to 100% and relative protection from apoptosis was determined. Mean F SE. *, P V 0.05. cycloheximide pointed to protein expression being involved. attenuated the S1P release compared with cells that retain full Cycloheximide, shortly supplied before putting cells to 1% SphK activity. CM(hypoxia) contained 354 nmol/L S1P versus oxygen, blocked expression of SphK2 under hypoxia (Fig. 3C). 188 nmol/L found in CM(hypoxia/DMS). Accordingly, on Accompanied by increased protein expression of SphK2, we treatment of A549 cells with etoposide, chemoresistance was measured an increased SphK2 enzyme activity after 2 to 4 h of attenuated in cells treated with CM(hypoxia/DMS) (Fig. 4A). hypoxia (Fig. 3D, black columns), whereas SphK1 activity was This effect could be reversed by coincubating CM(hypoxia/ not affected (Fig. 3D, white columns). To determine con- DMS) with 1 Amol/L S1P (Fig. 4A). To answer the question sequences of enhanced SphK2 activity, we next quantified whether SphK2 protein expression and its enzymatic activity extracellular concentrations of S1P in the supernatants of caused chemoresistance under hypoxia, we used validated small hypoxia-treated cells. There was an increase of extracellular interfering RNA (siRNA) against SphK2 to specifically inhibit S1P from f37 nmol/L at 4 h under normoxia compared with SphK2 expression. We noted a reduction of SphK2 mRNA by f145 nmol/L at 4 h of hypoxia (Fig. 3E). f50%, whereas transfection with either control siRNA (mock- We then asked whether enhanced SphK activity and the siRNA) or SphK1-specific siRNA did not affect SphK2-mRNA concomitant release of S1P contributed to protection of cells transcripts (Fig. 4B). In A549 cells, with a knockdown of under hypoxia. Therefore, we incubated cells for 24 h under SphK2, hypoxia was less effective in protecting cells from hypoxia in the presence of 5 Amol/L DMS and collected etoposide-induced cell death when compared with naive cells. medium. [CM(hypoxia/DMS)]. Quantification of extracellular Transfection with mock-siRNA or SphK1-siRNA did not S1P showed that inhibition of SphK activity in A549 cells interfere with chemoresistance under hypoxia (Fig. 4C). In

OF4 Schnitzer et al.

extending experiments, we checked whether the release of the survival molecule demands SphK2 expression. A549 cells were transfected with siRNA directed against SphK1, SphK2, or mock-siRNA followed by incubating transfected cells for 24 h under hypoxia to collected conditioned medium. Conditioned medium from cells with a knockdown of SphK2 (siSphK2-CM) failed to induce chemoresistance as potent as cells incubated with conditioned medium obtained from mock-siRNA-transfected cells (simock-CM) or SphK1- siRNA-transfected cells (siSphK1-CM; Fig. 4D). Experiments thus far indicate that activation of SphK2 under hypoxia is a prerequisite for the release of S1P under hypoxia, thus provoking a chemoresistant phenotype in A549 cells via autocrine or paracrine mechanisms. With the following experiments, we asked which S1P receptor mediates protection in response to conditioned medium and investigated intracellular signaling pathways conveying chemoresistance. We used the S1P receptor antagonists VPC23019 (to block S1P1/S1P3) and JTE013 (to block S1P2). Only coincubation of conditioned medium with 1 Amol/L VPC23019 diminished the protective principle of CM(hypoxia), which points to S1P1 or S1P3 being involved (Fig. 5A). We then focused on signaling pathways downstream of S1P1/S1P3 receptor activation. Inhibition of major signaling cascades by LY294002 (PI3K), SB203580 (p38 MAPK), U0126 (p42/44 MAPK), or inhibitors of the JAK/ STAT pathway showed that only inhibition of p42/44 MAPK by U0126 interfered with protection in response to CM(hypoxia) or Lip-CM(hypoxia). Coincubating cells with CM(hypoxia) or Lip-CM(hypoxia), which both protect cells from cell death, and 5 Amol/L U0126 sensitized cells toward etoposide-induced apoptosis (Fig. 5B). To confirm that p42/44 MAPK signaling is activated on CM(hypoxia) and Lip- CM(hypoxia), we checked phosphorylation (that is, activation) of p42/44 MAPK on treatment of A549 cells with S1P, CM(hypoxia), or Lip-CM(hypoxia). As expected, activation of p42/44 MAPK occurred after treatment of cells for 15 min with 1 Amol/L S1P. Exposing cells to CM(hypoxia) or Lip- CM(hypoxia) also produced robust p42/44 MAPK phosphorylation at 5 to 30 min, whereas the amount of total p42/44 MAPK protein and h-tubulin remained unaltered (Fig. 5C).

Discussion Growing tumors require sufficient blood and nutrient supply. HIF-1 is activated to support the development and progression of tumors by releasing proangiogenic and proproliferative FIGURE 2. S1P protects from etoposide-induced apoptosis. A. cytokines and/or growth factors (3, 36, 37). Another character- Caspase-3 activity assays in A549 cells treated with lysophosphatidic acid (LPA;10Amol/L), platelet-activating factor (PAF;1Amol/L), istic feature of cancer cells is their resistance toward ceramide-1-phosphate (C1P;1Amol/L), or S1P (10 Amol/L) for 24 h under chemotherapeutic agents, a process that can be promoted by normoxia followed by etoposide treatment. Caspase-3 activity of cells treated with etoposide under normoxia was set to 100% and relative intratumoral hypoxia (2, 8). protection from apoptosis was determined. Asterisks, significant differ- In contrast to published data showing that hypoxia failed to ences compared with etoposide treatment under normoxia. B. A549 cells protect A549 cells from etoposide-induced cell death (38), we were treated with different concentrations of S1P (0.1-10 Amol/L) under normoxia 24 h before etoposide treatment. Caspase-3 activity was corroborated our previous results showing the appearance of a determined as described above. Asterisks, significant differences com- chemoresistant phenotype induced by hypoxia (5). These pared with etoposide treatment under normoxia. C. Caspase-3 activity differences can be explained by discrepancies in treatment was measured in A549 cells treated for 24 h under normoxia in the presence of 10 Amol/L DMS either alone or coincubated with S1P (10 conditions. Cosse et al. incubated cells for 16 h under serum- Amol/L), Lip-CM(hypoxia), or CM(hypoxia). Caspase-3 activity was free conditions in the presence of 50 Amol/L etoposide without normalized to the untreated normoxic control and relative induction of caspase-3 activity was determined. Asterisks, significant difference hypoxic preincubations (5). In contrast, we preconditioned cells compared with DMS treatment. Mean F SE. *, P V 0.05. for 24 h under hypoxia before short-time (4 h) etoposide

Hypoxia-Induced S1P Mediates Chemoresistance OF5

FIGURE 3. Hypoxia induces SphK2 protein and its activity and triggers the release of S1P. A. mRNA quantification of SphK1 (white columns) and SphK2 (black columns) in A549 cells after incubation for 4 h under normoxia, in the absence or presence of 100 Amol/L CoCl2, or hypoxia. Basal SphK-mRNA expression under normoxia was set to 100% and relative mRNA expression was determined. B. Expression of HIF-1a, SphK2, SphK1, and actin in A549 cells exposed for 1 to 4 h to hypoxia was determined by Western blot analysis. Actin expression served as a loading control. Representative of three experiments. C. Expression of SphK2 and actin in A549 cells exposed to 10 Amol/L MG132 for 4 h or to hypoxia for 2 and 4 h with or without the previous addition of 10 Ag/mL cycloheximide. Actin expression served as a loading control. Representative of three experiments. D. Relative enzyme activity of SphKs in A549 cells incubated for 2 to 4 h under hypoxia (white column, SphK1; black column, SphK2). Basal SphK activity under normoxia was set to 100% and relative induction was determined. Asterisks, significant differences compared with normoxic controls. E. A549 cells were incubated under serum-free conditions under normoxia (4 h) or hypoxia (2-4 h) and S1P in the supernatants was quantified. Asterisks, significant difference compared with normoxic treatments. Mean F SE. *, P V 0.05. treatments. We now addressed the role of S1P as a potential translation being involved. Interestingly enough, mRNA mediator contributing to chemoresistance under hypoxia. stability regulation of SphK1 has recently been reported to Conditioned medium obtained from cells incubated under account for v-Src-mediated SphK1 overexpression (39), hypoxia contained a lipid that protected A549 cells from cell whereas information for SphK2 is lacking thus far. Moreover, death via autocrine and/or paracrine mechanisms. Circumstan- one has to keep in mind that cancer cells such as A549 often tial evidence supports the concept that S1P is an effective harbor functional mutations in signaling pathways (e.g., of mediator of chemoresistance, with the notion that SphK2 PI3K), which are responsive to hypoxia (9, 40). Hypoxia produces S1P under hypoxia. In A549 cells, both SphK therefore may enhance mRNA transcription and translation or isoforms are active and contribute to cell viability under affect mRNA stability to increase SphK2. Alternatively, unstimulated conditions. We noticed a rapid increase of SphK2 hypoxia stimulates the enzymatic activity of SphKs to increase protein and activity within 1 to 4 h under hypoxia. A recent the S1P concentration. These multiple options of hypoxic publication by Anelli et al. showed up-regulation of SphK1- regulation may explain expression and/or activation of either mRNA and protein and the release of S1P in a HIF-2a- SphK1 or SphK2 under various experimental settings, such as dependent manner (33). In endothelial cells, Schwalm et al. long-lasting hypoxic exposure periods (32), using CoCl2 rather noticed a rapid and long-lasting activation of SphK1 under than authentic hypoxia (33) or referring to HIF-independent hypoxia (2-24 h), which was associated with HIF-1- and HIF-2- regulation of SphK activity as seen by Ader et al. (34) and our dependent signaling (32). More recently, activation of SphK1 study. under hypoxia was related to reactive oxygen species Besides SphK2 protein expression, we observed an increase formation, which in turn facilitated stabilization of HIF-1a in SphK2 activity under short-time hypoxia, which correlated (34). In A549 cells, we excluded the HIF system to regulate with an increase of extracellular S1P. It is discussed that several SphK2, because CoCl2, a potent inducer of HIF activity, stimuli increased SphK activity most likely by achieving post- showed no respective activity. In addition, transcriptional translational protein modification. Tumor necrosis factor-a regulation of SphK2 under hypoxia is unlikely because we triggered phosphorylation of SphK1 by p42/44 MAPK (41), observed induction of SphK2 within 1 h of hypoxia. The fast whereas epidermal growth factor induced a protein kinase Cy- response may point to mechanisms of mRNA stability and dependent phosphorylation and the subsequent translocation of

OF6 Schnitzer et al.

FIGURE 4. Role of SphK2 in promoting chemoresistance under hypoxia and by CM(hypoxia). A. Media were obtained from cells incubated for 24 h under hypoxia [CM(hypoxia)] or under hypoxia in the presence of 5 Amol/L DMS [CM(hypoxia/DMS)]. Naive cells were incubated for 24 h under normoxia with CM(hypoxia) or CM(hypoxia/DMS) or coincubated with CM(hypoxia/DMS) and 1 Amol/L S1P followed by the addition of etoposide. Caspase-3 activity of cells treated with etoposide under normoxia was set to 100% and relative protection from apoptosis was determined. B. Quantification of SphK2-mRNA content of A549 cells. Cells remain untreated or were transfected with mock-siRNA, SphK1-siRNA, or SphK2-siRNA followed by incubations for 4 h under hypoxia. Basal SphK2-mRNA expression under normoxia was set to 100% and relative SphK2-mRNA expression was determined. Asterisks, significant difference compared with normoxic control. C. Cells remained untreated or were transfected with mock-siRNA, SphK1-siRNA, or SphK2-siRNA and preincubated under normoxia or hypoxia for 24 h followed by etoposide treatment. Caspase-3 activity was determined as described above. D. A549 cells remained as controls or were incubated with conditioned medium for 24 h under normoxia. Conditioned medium was generated under hypoxia by untreated cells [CM(hypoxia)], mock-siRNA-transfected cells (simock-CM), SphK1-siRNA-transfected cells (siSphK1-CM), or SphK2-siRNA-transfected cells (siSphK2-CM) followed by the addition of etoposide. Caspase-3 activity was determined as described above. Mean F SE. *, P V 0.05.

SphK1 to the plasma membrane (42). Phorbol 12-myristate being either SphK1 or SphK2, should be irrelevant as long as 13-acetate triggered p42/44 MAPK and protein kinase D to generated S1P is exported from cells. Extracellular S1P phosphorylate SphK2, thus enhancing its catalytic activity and transmits signals via G-protein-coupled receptors, of which mediating its nuclear export (25, 26). Activation of signaling S1P1/S1P5 have been identified. S1P receptor activation and pathways such as PI3K, extracellular signal-regulated kinase coupling to Gi,Gq,orG12/G13 allows a crosstalk between 1/2, or p38 MAPK, which are active under hypoxia (9, 40), separate G-proteins, which facilitates a fine-tuning of signals might therefore add to SphK2 regulation. Activation of SphK2 transmitted by S1P (44). Experiments using receptor antago- as seen in our study under hypoxia was cytoprotective rather nists point to S1P1/S1P3 being involved. Activation of PI3K than cell destructive as was observed before based on forced and p42/44 MAPK are two pathways known to be activated by SphK2 overexpression (14). Differences might not only be the S1P (31), with the notion that these pathways also promote cell amount of protein being expressed by forced overexpression survival. Kang et al. reported up-regulation of cFLIP and versus a moderated hypoxic increase but also different phosphorylation of Bad downstream of PI3K/Akt signaling localization of the enzyme or the release of S1P to the (45). We showed previously that S1P released from apoptotic extracellular space under hypoxia, which in turn can protect cells enhanced macrophage survival via PI3K, p42/44 MAPK, 2+ cells. Compensatory mechanisms effective under hypoxia may and Ca signaling, caused Bcl-2 and Bcl-XL expression, and operate as well, as it is known that cells generally tolerate inactivated proapoptotic Bad (46). This signaling cascade was, hypoxia very well without entering cell death by apoptosis. at least in part, conveyed via the S1P1 receptor. In terms of cell Along with SphK2 activation, we noticed an increase of S1P viability, S1P1 and S1P2 are the most important S1P receptor in supernatants of hypoxia-treated cells. There is little question isoforms transmitting survival signals (31). Because only that S1P outside cells acts cytoprotective, even being inhibition of S1P1/S1P3, but not S1P2, abrogated CM(hy- considered an essential growth-promoting factor that contrib- poxia)-mediated protection, S1P-dependent survival is likely to utes to tumor progression (43). In this regard, the source of S1P, be transmitted via S1P1 in A549 cells. The fact that S1P,

Hypoxia-Induced S1P Mediates Chemoresistance OF7

CM(hypoxia), and lipid extract [Lip-CM(hypoxia)] induced with 50 nmol/L 12-O-tetradecanoyl-phorbol-13-acetate for phosphorylation (activation of p42/44 MAPK), a signaling 16 h. Cells were incubated in a humidified atmosphere (21% pathway, which can be activated via Gi proteins downstream of O2;5%CO2) and kept at a confluence of V80%. Hypoxic S1P1 receptors, supports this notion. incubations were done in an InVivo2 400 Hypoxia Working We provide evidence that hypoxia induced SphK2 expres- Station (Biotrace) at 1% O2, 94% N2, and 5% CO2. All cell sion, enhanced its activity, and promoted S1P formation. lines where regularly tested to be free of Mycoplasma. Extracellular S1P contributed to chemoresistance via S1P1/ S1P3 receptors and p42/44 MAPK-dependent signaling path- Reagents ways. We have shown previously that cyclooxygenase-2 is All chemicals were of highest grade of purity. U0126, up-regulated under hypoxia in a HIF-1-independent fashion etoposide, staurosporine, Ac-DEVD-Amc, and DMS were (5). It is interesting to note that S1P is a potent inducer of purchased from Axxora. S1P (D-erythro-S1P), NBD-sphingo- cyclooxygenase-2 expression (22, 47), thus opening the sine, VPC23019, and JTE013 were from Avanti Polar Lipids. possibility that S1P is responsible for cyclooxygenase-2 SphK1 and SphK2 antibodies were purchased from Abgent. induction and the formation of antiapoptotic prostaglandin E2 Antibodies against phospho-p42/44 (Thr202/Tyr204) and p42/44 under hypoxia. This might help to understand how signaling were from Cell Signaling Technology. The HIF-1a antibody pathways cooperate in cell protection within hypoxic tumor was from BD Biosciences, whereas the h-tubulin and actin regions. antibodies were from Sigma-Aldrich. Horseradish peroxidase- labeled secondary antibodies were from Amersham Biosciences Materials and Methods and fluorescence-labeled secondary antibodies were from Cell Culture Rockland Immunochemicals. A549 human lung carcinoma cells were maintained in DMEM/Ham’s F-12 (1:1) supplemented with 100 units/mL Preparation of Conditioned Medium and Extraction of penicillin, 100 Ag/mL streptomycin, 1% L-glutamine, and 10% Phospholipids FCS. THP-1 human monocytes were cultured in RPMI 1640 Cells were cultured to 80% confluence in gas-permeable cell supplemented with 100 units/mL penicillin, 100 Ag/mL culture flasks and incubated for 24 h under normoxia (21% O2) streptomycin, 1% L-glutamine, and 10% FCS. Differentiation or hypoxia (1% O2). If applicable, cells were stimulated of THP-1 monocytes to macrophages was done by treatment as indicated. Thereafter, medium was collected, centrifuged

FIGURE 5. S1P1/S1P3 mediate chemoresistance via extracellular signal-regulated kinase 1/2 signaling in response to conditioned medium. A. A549 cells were incubated for 24 h under normoxia or with CM(hypoxia) in the presence or absence of either 100 nmol/L JTE013 (S1P2 antagonist) or 1 Amol/L VPC23019 (S1P1/S1P3 antagonist) before the induction of apoptosis with etoposide. Caspase-3 activity of cells treated with etoposide under normoxia was set to 100% and relative protection from apoptosis was determined. B. The specific p42/44 MAPK inhibitor U0126 (5 Amol/L) was added 30 min before incubating cells with either CM(hypoxia) or Lip-CM(hypoxia) for 24 h under normoxia followed by etoposide treatment. Caspase-3 activity was determined as described above. Mean F SE. *, P V 0.05. C. Western analysis in A549 cells treated with S1P (1 Amol/L), CM(hypoxia), or Lip-CM(hypoxia) under normoxia for 5 to 30 min. Phospho-p42/44 (Thr202/Tyr204), total p42/44, and h-tubulin were detected by Western blot analysis. Representative of three experiments.

OF8 Schnitzer et al.

(5 min, 4jC, 500 Â g) and filtered (0.22 Am) to remove cell either 150 Amol/L etoposide or 1 Ag/mL staurosporine for debris, transferred into fresh tubes, and stored at 4jC. For 4 h (A549) or 1 Ag/mL staurosporine for 8 h (THP-1). Caspase- phospholipid extraction, 1 mL conditioned medium was mixed 3 activity was quantified following the cleavage of Ac-DEVD- with 3.75 mL chloroform/methanol/12 N HCl (v/v/v, 2/4/0.1) AMC as described previously (5). Briefly, cells were resus- and vortexed vigorously. Thereafter, 1.25 mL chloroform was pended in assay buffer (100 mmol/L HEPES, 0.1% CHAPS, added followed by vortexing. Finally, 1 mL double-distilled 10% sucrose, 1 mmol/L DTT), sonicated on ice (15 s), and Â H2O was added followed by subsequent agitation for 5 min and centrifuged at 15,000 g for 15 min. The cell supernatant centrifugation for 10 min at 1,000 rpm to achieve phase (50 Ag) was diluted with assay buffer to a final volume of separation. The lower chloroform phase was transferred into a 140 AL, Ac-DEVD-Amc substrate (0.2 mmol/L) was added, fresh tube and traces of water were removed by adding sodium and caspase-3 activity was determined for 1 h at 30jC using sulfate. The lipid phase was filtered and concentrated using the Mithras LB940 multimode reader (Berthold; extinction, a speed-vac and extracts were resuspended in ethanol. The 360 nm; emission, 460 nm). volume of lipid extract used for stimulation of the cells contained the amount of lipids extracted from 2 mL conditioned SphK Activity Assay medium. Activity of SphKs was followed as phosphorylation of NBD-sphingosine to NBD-S1P as described previously (48). Transfection with siRNA For preparation of whole-cell lysates, cells were harvested, Cells [3 Â 105 (caspase-3 activity assay) or 1 Â 106 (mRNA washed with ice-cold PBS, and resolved in 100 AL ice-cold extraction)] were plated in cell culture dishes. The following SphK lysis buffer [20 mmol/L Tris-HCl, 20% (v/v) glycerol, day, 100 nmol/L of either specific siRNA or siCONTROL 1 mmol/L h-mercaptoethanol, 1 mmol/L EDTA, 15 mmol/L nontargeting Duplex 1 (mock-siRNA) was transfected with NaF, 40 mmol/L h-glycerophosphate, protease inhibitor mix DharmaFECT1 transfection reagent following the manufac- (pH 7.4)]. Subsequently, lysates were sonicated on ice for 5 s turer’s guideline. Transfected cells were incubated under followed by centrifugation (12,000 Â g,30min,4jC). The normoxia for 48 h in medium containing the transfection supernatants (10 AL) were then added to 100 AL SphK assay mixture. Before the experiments, medium was changed and buffer [SphK1: 50 mmol/L HEPES, 15 mmol/L MgCl2, 0.5% cells were stimulated as indicated. DharmaFECT1 and Triton X-100, 10 mmol/L NaF, 1.5 mmol/L semicarbazide siCONTROL nontargeting Duplex 1 were purchased from (pH 7.4); SphK2: 50 mmol/L HEPES, 15 mmol/L MgCl2, Dharmacon. Validated SphK1- or SphK2-specific siRNA was 1 mol/L KCl, 10 mmol/L NaF, 1.5 mmol/L semicarbazide purchased from Qiagen. (pH 7.4)], together with 10 Amol/L NBD-sphingosine and 1 mmol/L ATP, for 2 h at 30jC. Addition of 100 AL alkaline potassium phosphate buffer followed by chloroform/methanol/ Western Blot Analysis HCl (500 AL, 2:1:0.1, v/v/v) allowed to isolate the product NBD- Briefly, cells (1 Â 106 per 10 cm dish) were incubated as S1P in the aqueous phase. The aqueous phase (100 AL) was then indicated, scraped off the plates, lysed in 100 AL lysis buffer analyzed for NBD-S1P-dependent fluorescence in the Mithras (100 mmol/L HEPES, 0.1% CHAPS, 10% sucrose, protease LB 940 multimode (extinction, 485 nm; emission, 538 nm). inhibitor mix, 1 mmol/L DTT), sonicated on ice (10 s), and centrifuged (15,000 Â g, 15 min). Protein (120 Ag) was mixed with 4Â SDS-PAGE sample buffer [125 mmol/L Tris-HCl, 2% Quantification of S1P SDS, 20% glycerine, 1 mmol/L DTT, 0.002% bromophenol Quantification of S1P in cell culture supernatants was blue (pH 6.9)], denaturated for 5 min, and centrifuged achieved by the S1P ELISA kit (Echelon Biosciences). (15,000 Â g, 5 min). Proteins were separated by 10% SDS- Supernatants were collected from A549 cells and incubated PAGE. Gels were washed with blotting buffer [25 mmol/L Tris- under normoxia (4 h) or hypoxia (2-4 h) under serum-free HCl, 192 mmol/L glycine, 20% methanol (pH 8.3)] for 5 min. conditions. Before incubation, cells were washed twice in PBS Gels were blotted on Hybond C extra nitrocellulose membrane to remove traces of serum. Additionally, we determined S1P (Amersham Biosciences) by a semidry electrophoretic transfer concentrations in CM(hypoxia) or CM(hypoxia/DMS). Quan- cell (Bio-Rad). Unspecific binding was blocked with 5% milk/ tification of S1P was done following the manufacturer’s TBS [50 mmol/L Tris-HCl, 140 mmol/L NaCl, 0.05% Tween guidelines. Absorbance (450 nm) was measured by using the 20 (pH 7.2)] for 1 h. All antibodies used were incubated Apollo-8 plate reader (Berthold). following the manufacturer’s guidelines. Afterwards, mem- branes were washed three times for 5 min each with PBS. For Statistical Analysis protein detection, blots were incubated with horseradish Each experiment was done at least three times. Represen- peroxidase-labeled secondary antibodies (1:2,000 in 5% milk/ tative blots are shown. Data are presented as mean F SE. Tween 20-TBS) for 2 h and washed once for 5 min with Tween Statistical analysis was done using a one-way ANOVA 20-TBS and twice for 5 min each with PBS followed by combined with Tukey’s post-test (GraphPad Prism version enhanced chemiluminescence or fluorescence detection using 4.00). Unless indicated otherwise, differences were considered Odyssey infrared imaging system (LI-COR Biotechnology). significant at P V 0.05.

Caspase-3Activity Assay Disclosure of Potential Conflicts of Interest Apoptosis was induced under serum-free conditions using No potential conflicts of interest were disclosed.

Hypoxia-Induced S1P Mediates Chemoresistance OF9

26. Ding G, Sonoda H, Yu H, et al. Protein kinase D-mediated phosphorylation References and nuclear export of sphingosine kinase 2. J Biol Chem 2007;282:27493 – 502. 1. Semenza GL. Targeting HIF-1 for cancer therapy. Nat Rev Cancer 2003;3: 721 – 32. 27. Cuvillier O, Pirianov G, Kleuser B, et al. Suppression of ceramide-mediated programmed cell death by sphingosine-1-phosphate. Nature 1996;381:800 – 3. 2. Zhou J, Schmid T, Schnitzer S, Brune B. Tumor hypoxia and cancer progression. Cancer Lett 2006;237:10 – 21. 28. Edsall LC, Cuvillier O, Twitty S, Spiegel S, Milstien S. Sphingosine kinase expression regulates apoptosis and caspase activation in PC12 cells. J Neurochem 3. Rankin EB, Giaccia AJ. The role of hypoxia-inducible factors in tumorigen- 2001;76:1573 – 84. esis. Cell Death Differ 2008;15:678 – 85. 29. Ahmad M, Long JS, Pyne NJ, Pyne S. The effect of hypoxia on lipid 4. Semenza GL. Hypoxia, clonal selection, and the role of HIF-1 in tumor phosphate receptor and sphingosine kinase expression and mitogen-activated progression. Crit Rev Biochem Mol Biol 2000;35:71 – 103. protein kinase signaling in human pulmonary smooth muscle cells. Prostaglandins 5. Schnitzer SE, Schmid T, Zhou J, Brune B. Hypoxia and HIF-1A protect A549 Other Lipid Mediat 2006;79:278 – 86. cells from drug-induced apoptosis. Cell Death Differ 2006;13:1611 – 3. 30. Kim DS, Kim SY, Lee JE, et al. Sphingosine-1-phosphate-induced ERK 6. Liu L, Ning X, Sun L, et al. Hypoxia-inducible factor-1A contributes to activation protects human melanocytes from UVB-induced apoptosis. Arch hypoxia-induced chemoresistance in gastric cancer. Cancer Sci 2008;99:121 – 8. Pharm Res 2003;26:739 – 46. 7. Song X, Liu X, Chi W, et al. Hypoxia-induced resistance to cisplatin and 31. Radeff-Huang J, Seasholtz TM, Matteo RG, Brown JH. G protein mediated doxorubicin in non-small cell lung cancer is inhibited by silencing of HIF-1A signaling pathways in lysophospholipid induced cell proliferation and survival. gene. Cancer Chemother Pharmacol 2006;58:776 – 84. J Cell Biochem 2004;92:949 – 66. 8. Sullivan R, Pare GC, Frederiksen LJ, Semenza GL, Graham CH. Hypoxia- 32. Schwalm S, Doll F, Romer I, Bubnova S, Pfeilschifter J, Huwiler A. induced resistance to anticancer drugs is associated with decreased senescence and Sphingosine kinase-1 is a hypoxia-regulated gene that stimulates migration of requires hypoxia-inducible factor-1 activity. Mol Cancer Ther 2008;7:1961 – 73. human endothelial cells. Biochem Biophys Res Commun 2008;368:1020 – 5. 9. Lee SM, Lee CT, Kim YW, Han SK, Shim YS, Yoo CG. Hypoxia confers 33. Anelli V, Gault CR, Cheng AB, Obeid LM. Sphingosine kinase 1 is up- protection against apoptosis via PI3K/Akt and ERK pathways in lung cancer regulated during hypoxia in U87MG glioma cells. Role of hypoxia-inducible cells. Cancer Lett 2006;242:231 – 8. factors 1 and 2. J Biol Chem 2008;283:3365 – 75. 10. Spiegel S, Merrill AH, Jr. Sphingolipid metabolism and cell growth 34. Ader I, Brizuela L, Bouquerel P, Malavaud B, Cuvillier O. Sphingosine regulation. FASEB J 1996;10:1388 – 97. kinase 1: a new modulator of hypoxia inducible factor 1A during hypoxia in 11. Pettus BJ, Chalfant CE, Hannun YA. Ceramide in apoptosis: an overview and human cancer cells. Cancer Res 2008;68:8635 – 42. current perspectives. Biochim Biophys Acta 2002;1585:114 – 25. 35. Yun JK, Kester M. Regulatory role of sphingomyelin metabolites in hypoxia- 12. Phillips DC, Martin S, Doyle BT, Houghton JA. Sphingosine-induced induced vascular smooth muscle cell proliferation. Arch Biochem Biophys 2002; apoptosis in rhabdomyosarcoma cell lines is dependent on pre-mitochondrial Bax 408:78 – 86. activation and post-mitochondrial caspases. Cancer Res 2007;67:756 – 64. 36. Liao D, Johnson RS. Hypoxia: a key regulator of angiogenesis in cancer. 13. Hait NC, Oskeritzian CA, Paugh SW, Milstien S, Spiegel S. Sphingosine Cancer Metastasis Rev 2007;26:281 – 90. kinases, sphingosine 1-phosphate, apoptosis and diseases. Biochim Biophys Acta 37. Murdoch C, Giannoudis A, Lewis CE. Mechanisms regulating the 2006;1758:2016 – 26. recruitment of macrophages into hypoxic areas of tumors and other ischemic 14. Maceyka M, Sankala H, Hait NC, et al. SphK1 and SphK2, sphingosine tissues. Blood 2004;104:2224 – 34. kinase isoenzymes with opposing functions in sphingolipid metabolism. J Biol 38. Cosse JP, Sermeus A, Vannuvel K, Ninane N, Raes M, Michiels C. Chem 2005;280:37118 – 29. Differential effects of hypoxia on etoposide-induced apoptosis according to the 15. Olivera A, Kohama T, Edsall L, et al. Sphingosine kinase expression cancer cell lines. Mol Cancer 2007;6:61. increases intracellular sphingosine-1-phosphate and promotes cell growth and 39. Sobue S, Murakami M, Banno Y, et al. v-Src oncogene product increases survival. J Cell Biol 1999;147:545 – 58. sphingosine kinase 1 expression through mRNA stabilization: alteration of AU- 16. Tao R, Zhang J, Vessey DA, Honbo N, Karliner JS. Deletion of the rich element-binding proteins. Oncogene 2008;27:6023 – 33. sphingosine kinase-1 gene influences cell fate during hypoxia and glucose 40. Xu L, Pathak PS, Fukumura D. Hypoxia-induced activation of p38 mitogen- deprivation in adult mouse cardiomyocytes. Cardiovasc Res 2007;74:56 – 63. activated protein kinase and phosphatidylinositol 3¶-kinase signaling pathways 17. Liu H, Toman RE, Goparaju SK, et al. Sphingosine kinase type 2 is a putative contributes to expression of interleukin 8 in human ovarian carcinoma cells. Clin BH3-only protein that induces apoptosis. J Biol Chem 2003;278:40330 – 6. Cancer Res 2004;10:701 – 7. 18. Igarashi N, Okada T, Hayashi S, Fujita T, Jahangeer S, Nakamura S. 41. Billich A, Bornancin F, Mechtcheriakova D, Natt F, Huesken D, Baumruker Sphingosine kinase 2 is a nuclear protein and inhibits DNA synthesis. J Biol T. Basal and induced sphingosine kinase 1 activity in A549 carcinoma cells: B A Chem 2003;278:46832 – 9. function in cell survival and IL-1 and TNF- induced production of inflammatory mediators. Cell Signal 2005;17:1203 – 17. 19. Shu X, Wu W, Mosteller RD, Broek D. Sphingosine kinase mediates vascular endothelial growth factor-induced activation of ras and mitogen-activated protein 42. Paugh BS, Paugh SW, Bryan L, et al. EGF regulates plasminogen activator D kinases. Mol Cell Biol 2002;22:7758 – 68. inhibitor-1 (PAI-1) by a pathway involving c-Src, PKC , and sphingosine kinase 1 in glioblastoma cells. FASEB J 2008;22:455 – 65. 20. Francy JM, Nag A, Conroy EJ, Hengst JA, Yun JK. Sphingosine kinase 1 expression is regulated by signaling through PI3K, AKT2, and mTOR in human 43. Visentin B, Vekich JA, Sibbald BJ, et al. Validation of an anti-sphingosine-1- coronary artery smooth muscle cells. Biochim Biophys Acta 2007;1769:253 – 65. phosphate antibody as a potential therapeutic in reducing growth, invasion, and angiogenesis in multiple tumor lineages. Cancer Cell 2006;9:225 – 38. 21. Sarkar S, Maceyka M, Hait NC, et al. Sphingosine kinase 1 is required for migration, proliferation and survival of MCF-7 human breast cancer cells. FEBS 44. Sanchez T, Hla T. Structural and functional characteristics of S1P receptors. Lett 2005;579:5313 – 7. J Cell Biochem 2004;92:913 – 22. 22. Pettus BJ, Bielawski J, Porcelli AM, et al. The sphingosine kinase 1/ 45. Kang YC, Kim KM, Lee KS, et al. Serum bioactive lysophospholipids sphingosine-1-phosphate pathway mediates COX-2 induction and PGE2 produc- prevent TRAIL-induced apoptosis via PI3K/Akt-dependent cFLIP expression and tion in response to TNF-A. FASEB J 2003;17:1411 – 21. Bad phosphorylation. Cell Death Differ 2004;11:1287 – 98. 23. Kohama T, Olivera A, Edsall L, Nagiec MM, Dickson R, Spiegel S. 46. Weigert A, Johann AM, von Knethen A, Schmidt H, Geisslinger G, Brune B. Molecular cloning and functional characterization of murine sphingosine kinase. Apoptotic cells promote macrophage survival by releasing the antiapoptotic J Biol Chem 1998;273:23722 – 8. mediator sphingosine-1-phosphate. Blood 2006;108:1635 – 42. 24. Pitson SM, Moretti PA, Zebol JR, et al. Activation of sphingosine kinase 1 by 47. Johann AM, Weigert A, Eberhardt W, et al. Apoptotic cell-derived ERK1/2-mediated phosphorylation. EMBO J 2003;22:5491 – 500. sphingosine-1-phosphate promotes HuR-dependent cyclooxygenase-2 mRNA stabilization and protein expression. J Immunol 2008;180:1239 – 48. 25. Hait NC, Bellamy A, Milstien S, Kordula T, Spiegel S. Sphingosine kinase type 2 activation by ERK-mediated phosphorylation. J Biol Chem 2007;282: 48. Billich A, Ettmayer P. Fluorescence-based assay of sphingosine kinases. Anal 12058 – 65. Biochem 2004;326:114 – 9.

Hypoxia Enhances Sphingosine Kinase 2 Activity and Provokes Sphingosine-1-Phosphate-Mediated Chemoresistance in A549 Lung Cancer Cells

Mol Cancer Res Published OnlineFirst February 24, 2009.

Updated version Access the most recent version of this article at: doi:10.1158/1541-7786.MCR-08-0156

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 Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://mcr.aacrjournals.org/content/early/2005/12/31/1541-7786.MCR-08-0156.citation. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.