Oncogene (2002) 21, 6317 – 6327 ª 2002 Nature Publishing Group All rights reserved 0950 – 9232/02 $25.00 www.nature.com/onc

Thioredoxin reductase regulates AP-1 activity as well as thioredoxin nuclear localization via active in response to ionizing radiation

Shervin Karimpour1,7, Junyang Lou2,7, Lilie L Lin2, Luis M Rene2, Lucio Lagunas2, Xinrong Ma3, Sreenivasu Karra3, C Matthew Bradbury1, Stephanie Markovina2, Prabhat C Goswami4, Douglas R Spitz4, Kiichi Hirota5, Dhananjaya V Kalvakolanu3, Junji Yodoi6 and David Gius*,1

1Radiation Oncology Branch, Radiation Oncology Sciences Program, Center for Cancer Research, National Cancer Institute, National Institutes of Health, Bethesda, Maryland, USA; 2Section of Cancer Biology, Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, Missouri, USA; 3Greenebaum Cancer Center, Department of Microbiology and Immunology, Molecular and Cellular Biology Program, University of Maryland School of Medicine, Baltimore, Maryland, USA; 4Free Radical and Radiation Biology Program, Department of Radiology, University of Iowa, Iowa City, Iowa, USA; 5Department of Anesthesia, Kyoto University Hospital, Kyoto University, Kyoto, Japan; 6Department of Biological Responses, Institute for Virus Research, Kyoto University, Kyoto, Japan

A recently identified class of signaling factors uses TR enzyme as a signaling factor in the regulation of AP- critical motif(s) that act as redox-sensitive 1 activity via a cysteine motif located in the . ‘sulfhydryl switches’ to reversibly modulate specific Oncogene (2002) 21, 6317 – 6327. doi:10.1038/sj.onc. cascades regulating downstream 1205749 with similar redox-sensitive sites. For example, signaling factors such as redox factor-1 (Ref-1) and Keywords: ; thioredoxin; AP-1; transcription factors such as the AP-1 complex both redox; ionizing radiation contain redox-sensitive cysteine motifs that regulate activity in response to . The mammalian thioredoxin reductase-1 (TR) is an oxidoreductase selenocysteine-containing flavoprotein that also appears Introduction to regulate multiple downstream intracellular redox- sensitive proteins. Since ionizing radiation (IR) induces The cytotoxicity of IR is primarily mediated via the oxidative stress as well as increases AP-1 DNA-binding production of reactive oxygen intermediates (ROI) from activity via the activation of Ref-1, the potential roles of intracellular H2O. ROI (in order of sequential reduction - TR and thioredoxin (TRX) in the regulation of AP-1 from O2, superoxide [O2 ], hydrogen peroxide [H2O2], activity in response to IR were investigated. Permanently and hydroxyl radical [.OH]) are normally produced by transfected cell lines that overexpress wild type TR cells as side products of electron transfer reactions demonstrated constitutive increases in AP-1 DNA- during cellular respiration (Halliwell and Gutterridge, binding activity as well as AP-1-dependent reporter 1988). Ideally, a metabolically active cell should strike a expression, relative to vector control cells. In contrast, balance between ROI production and the cellular permanently transfected cell lines expressing a TR gene defense system, resulting in either a neutral with the active site cysteine motif deleted were unable to or slightly reduced cellular environment (Halliwell and induce AP-1 activity or reporter in Gutterridge, 1990). The accumulation of abnormally response to IR. Transient genetic overexpression of high levels of ROI, as a result of any source including either the TR wild type or dominant-negative IR, can create a condition referred to as ‘oxidative demonstrated similar results using a transient assay stress’ that damages cells by lipid peroxidation and system. One mechanism through which TR regulates disruption (Storz et al., 1990). AP-1 activity appears to involve TRX sub-cellular In response to a wide variety of environmental localization, with no change in the total TRX content stresses including IR, a class of proto-oncogenes of the cell. These results identify a novel function of the (including c-Fos and c-Jun) referred to as immediate early response genes are activated (Abate et al., 1991; Gius et al., 1990; Kerppola and Curran, 1995). These genes encode nuclear transcription factors (i.e., AP-1) *Correspondence: D Gius, Section Chief, Molecular Radiation Oncology, Radiation Oncology Branch, Radiation Oncology Sciences involved in the transmission of inter- and intracellular Program, Center for Cancer Research, National Cancer Institute, information through multiple cellular signaling path- National Institutes of Health, 9000 Rockville Pike, Bldg 10, Room ways (Holbrook and Fornace, 1991; Kerr et al., 1992). B3B69, Bethesda, Maryland, 20892-1002301, USA; One possible role for the induction of these transcrip- E-mail: [email protected] 7These two authors contributed equally to this work tion factors is to modulate the expression of specific Received 8 January 2002; revised 28 May 2002; accepted 7 June target genes involved in a protective or reparative 2002 cellular response to the damaging effects of oxidative Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6318 stress induced by exogenous cytotoxic agents such as One intracellular function of TRX is to activate IR (Karin and Smeal, 1992; Kerppola and Curran, protein-nucleic acid interactions of nuclear transcrip- 1995; Xanthoudakis et al., 1992). tion factors via redox regulation (Hirota et al., 1999, c-Fos and c-Jun are members of a multigene family 2000a; Wei et al., 2000). It has been previously shown implicated in a number of signal transduction cascades that following exposure to IR (Wei et al., 2000) or associated with growth, differentiation, neuronal exci- other oxidants (Hirota et al., 1999, 2000a), TRX tation, and cellular stress, and thus provide a useful translocates into the nucleus, where it interacts with model for the investigation of stimulus-evoked altera- another redox-sensitive signaling protein, redox factor- tions in gene expression (Abate et al., 1990, 1991; Kerr 1, and appears to regulate the nuclear transcription et al., 1992). c-Fos and c-Jun comprise part of the factor complex, AP-1. The current study demonstrates mammalian AP-1 and heterodi- the essential role of TR as a redox-sensitive signaling merize to bind to a specific DNA sequence referred to factor functioning upstream of TRX in the regulation as the AP-1 site; this binding activates the expression of AP-1 activity and demonstrates that the active site of downstream target genes (Franza et al., 1988). In cysteine residues are essential in mediating IR-induced this regard, the induction of c-Fos and c-Jun is an ideal activation as well as the regulation of stress-induced paradigm for studying the role of early response genes TRX nuclear translocation. in response to oxidative cellular stress induced by IR. One post-translational mechanism that modulates c- Results Fos/c-Jun DNA-binding activity in vitro involves changes in oxidation-reduction (redox) state of the Characterization of TRX and TR overexpressing cell lines protein complex (Abate et al., 1991; Xanthoudakis et al., 1992). In this process, redox regulation is mediated We have previously demonstrated that AP-1 DNA- by a conserved cysteine residue flanked by basic amino binding activity is induced following exposure to IR via acids lysine and arginine (KCR) located in the basic a pathway that appears to involve redox regulation of DNA-binding domain of both c-Fos and c-Jun (Abate et TRX and Ref-1 (Wei et al., 2000). However, the IR- al., 1990). This specific motif appears to be conserved in induction of AP-1 dependent gene expression and the all of the c-Fos- and c-Jun-related proteins and is also upstream factors regulating TRX as well as the role of present, in a slightly modified form, in several other the active site cysteine residues in the process had not redox sensitive transcription factors (Abate et al., 1990; been resolved. To address these issues, a series of Xanthoudakis et al., 1992; Kerppola and Curran, 1995). permanent cell lines derived from MCF-7 or HeLa Since IR induces AP-1 DNA-binding activity as well as were constructed that stably overexpress: (1) a control alters the intracellular oxidation/reduction state of the expression vector (pcDNA-X); (2) a wild-type TRX- cell via the generation of ROIs, it seems logical to containing vector (pcDNA-TRX); (3) a TRX mutant- investigate if other signaling proteins upstream of AP-1 containing vector (pcDNA-TRX-dm); (4) a wild type containing conserved cysteines regulate AP-1 activity in TR-containing vector (pCXN2-myc-TR-wt); or (5) a response to IR. We have recently shown that thioredox- mutant TR containing vector (pCXN2-myc-mTR). in and Ref-1 undergo changes in redox state that These plasmids have been previously described (Hu et contribute to the activation of AP-1 DNA-binding al., 2001; Ma et al., 2001) activity (Wei et al., 2000). These results suggest a role To confirm the expression of these exogenous genes for additional redox-sensitive signaling pathways in in these permanently transfected cell lines, Western blot regulating cellular responses to IR. analyses were performed. pcDNA-TRX-wt and TR, TRX, and NADPH comprise a highly conserved, pcDNA-TRX-dm cell lines demonstrated a sevenfold ubiquitous system (Mustacich and Powis, 2000) that increase in immunoreactive TRX protein, relative to plays an important role in the redox regulation of the vector control (Figure 1a). In addition, Western multiple intracellular processes, including DNA synth- blot analysis of the pCXN2-myc-TR-wt and pCXN2- esis, cell growth, and resistance to cytotoxic agents that myc-mTR permanent cell lines, using an anti-myc induce oxidative stress and apoptosis (Berggren et al., antibody, showed a similar increase in exogenous TR 1996; Williams et al., 2000; Oberley et al., 2001). TR is a protein levels (Figure 1b). Interestingly, there is a homodimeric selenocysteine-containing protein that consistent increase in exogenous mTR protein levels catalyses the NADPH-dependent reduction of TRX as when compared to exogenous wild type TR protein well as numerous other oxidized cellular proteins levels. The mechanism for this is unclear, however, one (Becker et al., 2000; Poole et al., 2000). There are two could speculate that very high levels of wild type TR confirmed forms of mammalian TRs, TR1 and TR2, may impart a selective growth disadvantage. both of which share in common a conserved -Cys-Val- To confirm that the TR transfected cell lines Asn-Val-Gly-Cys- redox catalytic motif thought to play contained altered TR activity, pCXN2-myc-X, a central role in the redox function of TR (Mustacich pCXN2-myc-TR-wt, or pCXN2-myc-mTR cells were and Powis, 2000; Powis et al., 1998). Interestingly, one harvested and analysed for enzymatic activity as major downstream target of TR, TRX, also contains a previously described (Holmgren and Bjornstedt, 1995). redox-active disulfide/dithiol motif within the conserved These results demonstrated that transfected cell lines active site-Cys-Gly-Pro-Cys- (Hirota et al., 1999; overexpressing wild type TR (pCXN2-myc-TR-wt) Williams et al., 2000). exhibited a threefold increase in TR activity (Figure

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6319 as well as a source of reducing equivalents to regulate nuclear transcription factor activity (Sasada et al., 1996; Oberley et al., 2001, Wei et al., 2000). As such, it seemed logical to determine if TR was an upstream signaling factor involved in the activation of AP-1 activity following exposure to IR. To investigate this, the HeLa cell lines stably over- expressing vector only (pCXN2-myc-X), wild type TR gene (pCXN2-myc-TR-wt), or a mutant TR gene (pCXN2-myc-mTR) were used. These cells lines were harvested for EMSA to determine AP-1 DNA-binding activity. In the vector only pCXN2-myc-X cells we observed a 3.6-fold increase in AP-1 DNA-binding activity following IR, as compared to non-irradiated cells (Figure 2a, lanes 1 vs 4). This induction of AP-1 DNA-binding activity is consistent with the results published by others (Hallahan et al., 1993; Martin et al., 1997; Beetz et al., 2000; Wei et al., 2000) and is identical to results observed in the parental HeLa cells used to make the permanent cell lines described above (Wei et al., 2000). In the pCXN2-myc-TR-wt cells, a threefold increase in constitutive AP-1 DNA-binding activity was observed (lane 1 vs 2) with a consistent but minimal 1.3-fold induction following IR (lanes 2 vs 5). Figure 1 Overexpression of wild type and mutant forms of thior- In contrast, no induction of AP-1 DNA-binding edoxin and thioredoxin reductase by permanent cell lines. (a) activity was observed following IR in the pCXN2- 20 mg of MCF7 cell homogenate from each permanently trans- myc-mTR cell lines (Figure 2, lanes 3 vs 6). fected cell line was loaded and run on a 15% polyacrylamide gel followed by transfer to nitrocellulose paper. The blot was then To confirm and expand upon the results above, the incubated with mouse anti-human TRX antibody and HRP-con- role of TR in regulating AP-1-dependent reporter gene jugated mouse secondary antibody with ECL detection on X-ray expression was determined. Therefore, we used a film. (b)20mg of HeLa cell homogenate from each transfected cell transient transfection system with AP-1 reporter line was loaded into each lane and run on a 10% polyacrylamide as described above. (c) Enzymatic activity of thioredoxin reduc- plasmid transfected into vector only, wild-type TR tase in cells expressing various mutants of TR. pCXN2-myc-X, overexpressors, or mutant TR-expressing cells with pCXN2-myc-TR-wt, and pCXN2-myc-mTR cell lines were har- subsequent determination of IR-dependent luciferase vested and analysed as described (Holmgren and Bjornstedt, activity. For these transient transfections, pcDNA-X, 1995). Data indicates mean+s.d. of triplicates. In every case, pcDNA-TRX-wt, and pcDNA-TRX-dm cells were background enzymatic activity (due to endogenous TRX) was subtracted from the experimental values transfected with 2.0 mg of p7x-AP-1-tk-Luc (reporter plasmid), 12.0 mg of pUC (for equivalent transfection of total DNA), and 1.0 mg of pCMV-b-gal plasmid (added into every transfection plate to normalize for 1c), when compared to control (pCXN2-myc-X). In any possible differences in transfection efficiencies contrast, cells that overexpress the mutant TR gene between the individual permanent cell lines). Figure (pCXN2-myc-mTR) exhibited a roughly 65% decrease 2b illustrates the ratio of luciferase to b-galactosidase in TR activity (Figure 1c). It is worth mentioning that activity units for each of the experimental conditions in the changes in TR activity are somewhat less than this experiment. In the vector only pCXN2-myc-X cell expected when considering the significant overabun- line, a 3.5-fold increase in AP-1 dependent gene dance of exogenous to endogenous proteins (as shown in expression is detected following IR exposure, a result Figure 1a – c). This may be due to several factors also observed by other groups (Martin et al., 1997; including altered signaling complex protein – protein Beetz et al., 2000; Wei et al., 2000). In contrast, interactions, possible non-functioning exogenous pCXN2-myc-TR cells demonstrated a threefold protein, and/or altered protein conformation. Finally, increase in basal AP-1-dependent gene expression prior the results of these experiments confirm that: (1) the to IR that marginally increased an additional 1.4-fold mutant TR gene substantially reduces TR activity; and following exposure to IR. Finally, and in contrast to (2) TR overexpressing cells have increased TR activity. the vector only cells, no induction of AP-1-dependent reporter gene expression was observed in mutant pCXN2-myc-mTR cells following IR. These experi- Overexpression of wild-type TR stimulates AP-1 activity ments show that cell lines overexpressing the wild-type while overexpression of mutant TR prevents the induction TR gene initiate a pathway resulting in the constitutive of AP-1 activity in response to IR activation of the AP-1 complex, while cells expressing a The TR/TRX system serves as an electron donor for cysteine mutant TR gene are not able to activate AP-1 multiple enzymes such as reductases and activity in response to IR. These results suggest that

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6320 environmental stress inducing agents (Hirota et al., A. 1999, 2000a), has been well established. However, the role of the critical cysteines in IR-induced induction of the AP-1 complex has not been investigated. As such, permanently transfected cell lines that express the vector (pcDNA-X), wild type TRX gene (pcDNA- TRX-wt), or cysteine mutant TRX gene (pcDNA- TRX-dm) were harvested without and with exposure to IR and EMSA performed. In the pcDNA-X cells, a threefold increase in AP-1 DNA – binding activity was observed following IR (Figure 3a, lanes 1 vs 4) while pcDNA-TRX-wt cells demonstrated a constitutive threefold increase in DNA-binding activity that increased slightly following IR exposure (lanes 2 vs 5). These results are similar to those observed by Hirota et al. (2000a) using transient co-transfection assays and suggests that AP-1 activity is near- maximally activated in cell lines overexpressing TRX. In contrast, no induction of AP-1 DNA-binding was B. observed in pcDNA-TRX-dm cells (lanes 3 vs 6) exposed to IR. To determine the possible role of TRX in the regulation of AP-1-dependent gene expression, the described cell lines were used in transient assays. Transfection of p7x-AP-1-tk-LUC into pcDNA-X cells resulted in a roughly threefold increase in luciferase activity following exposure to IR, as compared to non- irradiated cells (Figure 3b). In pcDNA-TRX-wt cells, a threefold increase in luciferase activity was observed prior to IR with a small increase observed after exposure. In contrast, minimal induction of luciferase Figure 2 The active site critical cysteine residues in thioredoxin activity was observed in pcDNA-TRX-dm transfected reductase are essential for regulating AP-1 DNA binding and re- with p7x-AP-1-tk-LUC following IR. These results are porter gene expression in response to IR. (a) HeLa cell lines that consistent with those derived from our previous study stably express vector alone (pCXN2-myc-X), wild type TR (pCXN2-myc-TR-wt), or a mutant TR (pCXN2-myc-mTR) were (Wei et al., 2000) and suggest an essential role for the exposed to IR (10 Gy) or sham treated. Cell were harvested after critical cysteine residues of TRX in the regulation of 1 h via subcellular fraction followed by analysis of AP-1 DNA- radiation-induced signaling leading to AP-1 activity. binding activity utilizing EMSA. Arrows indicate the AP-1 com- plex and free unbound AP-1 oligonucleotide. Equal protein was determined using the Bradford protein assay. Sections of fluoro- Co-transfection of wild-type TR activates AP-1 gene grams from native gels obtained using a Storm Phosphorimager expression are shown. (b) pCXN2-myc-X, pCXN2-myc-TR-wt, pCXN2- myc-mTR cells were transfected with 2.0 mg of p7x-AP-1-tk-Luc, To further investigate the potential role of TR and TRX 1 mg of pCMV-b-gal, and 12.0 mg of pUC. Sham treated and ir- on the regulation of AP-1 gene expression a transient co- radiated cells (10 Gy) were harvested 8 h after treatment followed by analysis of luciferase and b-galactosidase activity. The results transfection model was established. This system was are presented as relative-fold induction over the sham treated con- used to expand upon and confirm the results observed in trol. Luciferase activity was normalized to b-galactosidase activity the permanent cells lines shown above. In these and the results are presented as luciferase/b-galactosidase relative experiments, 2.0 mg of p7x-AP-1-tk-LUC and 1 mgof units. All results are the mean of at least three separate experi- pCMV-b-gal were co-transfected into HeLa cells with ments all done in duplicate. The results of individual transfections varied by less than 25% control vector pCXN2, or into cells with an expression vector containing either the wild-type pCXN2-TR-wt or cysteine mutant pCXN2-mTR genes. The total amount of pCXN2- based expression vector DNA was held TR plays a role in the regulation of basal and inducible constant at 6.0 mg per transfection using pCXN2-TR- AP-1 activity. wt, pCXN2-mTR, or the empty vector pCXN2-X. pUC was used to bring each transfection to a total of 15.0 mg of plasmid DNA. Thus, each transfection plate received Overexpression of wild-type TRX stimulates AP-1 identical amounts of each specific type of reporter, activity while overexpression of mutant TRX prevents the expression, and carrier plasmid. In addition, the induction of AP-1 activity in response to IR relative-fold induction was calculated by normalizing The role of TRX in the regulation of AP-1 DNA- luciferase activity to the b-galactosidase activity in an binding activity following IR, as well as other effort to account for any possible differences in

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6321

Figure 4 Co-transfection of wild type TR and TRX activates AP-1 gene expression. (a) HeLa cells were co-transfected with 1 mg of p7x-AP-1-tk-LUC and either control vector pCXN2, or an expression vector containing either the wild-type (pCXN2- TR-wt) or cysteine mutant (pCXN2-mTR) genes. The total amount of pCXN2-based expression vector DNA was held con- Figure 3 The active site critical cysteine residues in thioredoxin stant at 6.0 mg per transfection. pUC was used to bring each are essential for regulating AP-1 DNA binding and reporter gene transfection to a total of 15.0 mg of plasmid DNA. The relative- expression in response to IR. (a) MCF7 cell lines that stably over- fold induction of luciferase activity was normalized by b-galacto- express either pcDNA-X, pcDNA-TRX-wt, or pcDNA-TRX-dm sidase activity via cotransfection of pCMV-b-GAL. All results are were either exposed to 10 Gy IR or sham treated. Cell were har- presented as relative-fold induction in luciferase activity over the vested after 1 h via subcellular fraction followed by analysis of baseline. (b) HeLa cells were co-transfected with 1 mg of p7x-AP- AP-1 DNA-binding activity utilizing EMSA. Arrows indicate 1-tk-LUC and either pcDNA-X, pcDNA-TRX-wt, or pcDNA- the AP-1 complex and free unbound AP-1 oligonucleotide. (b) TRX-dm in an identical manner described above. (c) HeLa cells pcDNA-X, pcDNA-TRX-wt, pcDNA-TRX-dm cells were trans- were co-transfected with 1 mg of p7x-AP-1-tk-LUC and either fected with 2.0 mg of p7x-AP-1-tk-Luc, 1 mg of pCMV-b-gal, pCXN2 or increasing amounts of pCXN2-TR-wt (0.5 – 2.0 mg) and 12.0 mg of pUC. Sham treated and irradiated cells were har- as shown (1 – 4). pCXN2 was added to each transfection to main- vested 1 h following IR and analysis of reporter gene expression tain the amount of pCXN2 based vector constant at 6.0 mg. Bars using luciferase and b-galactosidase activity was accomplished (5 – 8) represent 1 mg of p7x-AP-1-tk-LUC and 2.0 mg of pCXN2- as described in Figure 2 TR-wt transfected with increasing amounts of pCXN2-myc-mTR (0.5, 1.0, 2.0 or 4.0 mg). The total amount of pCXN2-based ex- pression vector DNA was held constant at 6.0 mg per transfection and pUC was added to bring the total amount of transfected transfection efficiencies. The results of these experiments plasmid DNA to 15.0 mg. All results are the mean of three sepa- rate experiments all done in duplicate. The results of individual demonstrated that the transient over expression of TR transfections varied by less than 25% activates AP-1 gene expression (Figure 4a). To follow this result, similar transient assays were done with the TRX plasmids (pcDNA-X, pcDNA-TRX-wt, or pcDNA-TRX-dm) and similar results were observed concentrations of pCXN2-TR-wt (0.5 – 2.0 mg). For with the wild type TRX expression vector in all experiments the total amount of transfected pCXN2 (Figure 4b). based expression vector was held constant at 6.0 mg. Finally, if the transient overexpression of TR These results demonstrated increased luciferase activity activates AP-1 dependent gene expression, as measured as a function of increasing amounts of transfected TR by luciferase activity, it seems logical to determine if expression vector (Figure 4c). Finally, co-transfection co-transfection of increasing amounts of the TR of 2.0 mg of pCXN2-TR-wt with increasing amounts of cysteine mutant gene would inhibit activity. As such, pCXN2-mTR expression vector inhibited the activation HeLa cells were co-transfected with 1.0 mg of p7x-AP- of the AP-1-dependent reporter cassette that was first 1-tk-LUC reporter cassette and either 6.0 mg of the observed at 2.0 mg of pCXN2-TR-dm and more control expression plasmid (pCXN2) or increasing dramatically at 4.0 mg (Figure 4c). These results, along

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6322 with those presented above, suggest that TR is an upstream signaling factor involved in the regulation of the AP-1 transcription factor.

TRX nuclear translocation is altered in TR overexpressing cells We have recently shown that TRX nuclear localization is one step in the activation of the AP-1 complex in response to IR (Wei et al., 2000). In addition, the results presented above demonstrated constitutively active AP-1 activity in permanent cell lines that overexpress the wild-type TR gene. Thus, it seemed logical to determine if one mechanism for the constitutive activation of the AP-1 complex in the TR overexpressing cells might be due to altered TRX sub-cellular localization. As such, pCXN2-myc-X, pCXN2-myc-TR-wt, and pCXN2-myc-mTR cells were harvested without and with exposure to IR and the nuclear immunoreactive TRX levels were determined. The pCXN2-myc-X cells without and with exposure to IR are shown (Figure 5a, lanes 1 vs 4) and demonstrated TRX nuclear translocation following IR exposure, as has been shown by others (Hirota et al., 1999, 2000a; Wei et al., 2000). This result is identical to that observed in the parental HeLa cell line exposed to IR (Wei et al., 2000) and acts as a positive control. Figure 5 TRX immunoreactive protein levels in TR permanently When the pCXN2-myc-TR-wt cells were examined, transfected wild type and mutant cell lines. (a) Nuclear and (b) an increase in TRX immunoreactive nuclear protein whole cell fractions from pCXN2-myc-X, pCXN2-myc-TR-wt, levels was observed with little additional increase or pCXN2-myc-mTR permanently transfected cells were treated following IR (Figure 5a, lanes 2 vs 5). This constitutive without or with IR and harvested after 1 h. 20 mg of cell homo- genate from each transfected cell line was loaded into each lane increase in nuclear TRX protein levels appears to and run on a 10% polyacrylamide gel followed by transfer to ni- closely correspond with the pattern of constitutive trocellulose paper. The blot was then incubated with anti-TRX increase in AP-1 DNA-binding and gene expression antibody and HRP-conjugated anti-mouse secondary antibody activity shown in Figure 2. No increase in nuclear with ECL detection on X-ray film TRX immunoreactive protein levels is observed in pCXN2-myc-mTR cells following IR (lanes 3 vs 6), as opposed to IR-induced TRX translocation seen in the these experiments further suggest that TR overexpres- control, vector only cells (lanes 1 vs 4). To rule out the sion results in TRX nuclear translocation while a TR possibility that the altered nuclear TRX protein levels cysteine mutant prevents TRX nuclear translocation and/or lack of nuclear translocation in the pCXN2- following IR. myc-mTR was due to changes in total TRX protein, whole cell extracts were examined in all three cell lines. Increased AP-1 DNA-binding activity is not due to As shown, no change in total TRX protein levels is increased c-Fos and c-Jun protein levels observed in pCXN2-myc-X, pCXN2-myc-TR-wt, and pCXN2-myc-mTR without or following exposure to IR The results above suggest that cell lines overexpressing (Figure 5b). TR or TRX have constitutively elevated AP-1 To further confirm the location of intracellular TRX, functional activity, however, it is not clear if this pCXN2-myc-X, pCXN2-myc-TR-wt, and pCXN2-myc- phenotype is mediated by increased c-Fos and c-Jun mTR cells were treated without and with IR followed protein levels or via a post translational induction of by examination with an indirect immunofluorescence AP-1 functional activity. To address this issue, method using antibodies raised against TRX. These pcDNA-X, pcDNA-TRX, or pcDNA-TRX-dm cell results (Figure 6) confirm those shown in Figure 5 and lines were harvested and immunoreactive c-Fos and demonstrate constitutive nuclear localization of TRX c-Jun protein levels were determined (Figure 7a). These in pCXN2-myc-TR-wt cells. In addition, the increase in results demonstrate that similar levels of immunoreac- nuclear TRX seen in vector only cells (pCXN2-myc-X) tive c-Fos and c-Jun protein were found in the vector following IR is not observed in pCXN2-myc-mTR control when compared to the wild type or mutant cells. The nuclear localization of TRX following IR has TRX overexpressing cell lines. Finally, similar experi- been previously shown using indirect immunofluores- ments were performed for the pCXN2-myc-X, pCXN2- cence in HeLa cells (Wei et al., 2000). The results of myc-TR-wt, or pCXN2-myc-mTR cell lines. Similar to

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6323

Figure 6 Indirect immunofluorescence cell staining with an anti-thioredoxin antibody in permanently transfected wild type and mu- tant cells. pCXN2-myc-X, pCXN2-myc-TR-wt, or pCXN2-myc-mTR permanently transfected cells were treated without or with IR. Indirect immunofluorescence cell staining with anti-thioredoxin antibody was done 1 h after IR. The images on the left are repre- sentative of non-irradiated cells and the images on the right are after irradiation with 10 Gy the results with the TRX cells, no apparent differences TRX, Ref-1, and AP-1) (Xanthoudakis and Curran, in the c-Fos and c-Jun immunoreactive protein levels 1992; Kirkpatrick et al., 1997; Hirota et al., 2000a; were observed among the TR cell lines (Figure 7b), Wei et al., 2000). These redox-sensitive signaling suggesting that a mechanism other than alterations in proteins and downstream transcription factors might the absolute amounts of c-Fos and c-Jun proteins must therefore play a central role in maintaining the steady account for the observed changes in AP-1 functional state intracellular balance between prooxidant activity. production, antioxidant capacity, and the repair of oxidative damage. IR results in the formation of ROI that are thought to initiate several redox-sensitive Discussion signaling cascades in response to the damaging and cytotoxic effects of IR (Wei et al., 2000). Since IR Many signaling cascades that redirect metabolism in appears to activate redox-sensitive signaling factors, it response to stress are thought to sense changes in is logical to hypothesize that critical cysteine residues cellular oxidation/reduction (redox) status through contained in TR/TRX might mediate these signaling redox sensitive thiol-containing proteins (such as pathways.

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6324 the wild type and mutant TRX genes. Interestingly, this observed increase in AP-1 DNA-binding activity is independent of increased total TRX (Figure 5b) or c-Fos and c-Jun protein levels (Figure 7a,b). Finally, preliminary results suggest that TR may regulate AP-1 activity by a mechanism involving the regulation of TRX sub-cellular localization (Figures 5a and 6). The results of these experiments, combined with earlier results (Wei et al., 2000), strongly support the hypothesis that following exposure to IR, TR mediates an alteration in the redox state of TRX that participates in the activation of AP-1 DNA binding activity and gene expression. In addition, it appears that the critical cysteines in TR and TRX are targets for this signaling process, further suggesting a mechanism involving alterations in the redox status of these proteins. The current work also confirms that TRX is an upstream, cytoplasmic signaling factor regulating both constitutive and IR-inducible AP-1 DNA binding activity in human cancer cells (Hirota et al., 1999, Wei et al., 2000). This observation is consistent with that observed by others (Hirota et al., 2000a,b) using TNFa or other agents to induce AP-1 activity. In this work, the authors used transient co-transfection assays to determine that overexpression of wild type TRX constitutively activated AP-1 gene expression. In addition, we have also now demonstrated that permanent cell lines that express a TRX mutant, lacking the redox sensitive N-terminal cysteines, are unable to induce AP-1 activity following exposure to IR (Figure 3a,b). This finding demonstrates that these redox sensitive cysteine residues are essential for TRX to function in activating AP-1 in response to IR. The Figure 7 c-Fos and c-Jun immunoreactive protein levels in (a) results of these experiments, combined with previous TR and (b) TRX permanently transfected wild type and mutant observations (Wei et al., 2000), clearly establish the cell lines. 20 mg of cell homogenate from each transfected cell line involvement of a redox-sensitive signaling pathway was loaded into each lane and run on a 10% polyacrylamide gel involving both TR and TRX in the regulation of AP-1 followed by transfer to nitrocellulose paper. The blot was then in- in response to IR. cubated with anti-c-Fos and c-Jun antibodies and HRP-conju- gated secondary antibody with ECL detection on X-ray film While the activation of transcription factors, such as AP-1, ultimately occurs within the nucleus, a long- standing hypothesis in radiation biology states that the initiation of cellular signaling pathways resulting from radiation exposure occurs at the cytoplasmic To address this issue, cell lines that overexpress wild membrane (Maity et al., 1994; Curry et al., 1999). type or cysteine mutant forms of TR were used. The TR is a cytoplasmic protein that appears to be part of mutant form of TR lacked critical N-terminal cysteine a signaling pathway involved in the TRX-dependent residues that presumably are involved with the passage activation of the AP-1 complex in cervical tumor cells of electrons from NADPH to TRX, effectively exposed to IR that includes TRX. Thus, it is appealing inhibiting the ability of TR to reduce TRX. The to hypothesize that TR is a signaling factor in a results of the experiments with cells overexpressing cascade that begins with IR-induced free radicals in the wild type TR demonstrated constitutive increases in cytoplasm, then activates transcription factors in the AP-1 DNA-binding activity and reporter gene expres- nucleus, which, in turn, regulate downstream genes sion (relative to vector controls) with little further that protect the cell from the oxidative stress induced induction following exposure to IR (Figure 2a,b). In by free radicals. This raises several interesting questions contrast, cell lines overexpressing mutant TR showed regarding the mechanisms involved in cytoplasmic no increase in constitutive AP-1 DNA-binding activity signaling cascades activated by IR as well as the and reporter gene expression (relative to vector specific factors that pass the signal from the cytoplasm controls) as well as no induction following IR (Figure to the nucleus. The results presented in this report 2a,b). In addition, very similar results were observed identify the cysteine residues located in the N-terminal with the permanently transfected cell lines expressing regions of TR and TRX as critical for IR-induced

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6325 activation of AP-1 activity. Thus, it would appear that transcription factors via TR, TRX, and Ref-1. Taken these critical cysteine residues are targets for the together with previous investigations (Xanthoudakis et passage of redox-sensitive cellular signals to transcrip- al., 1992; Gallegos et al., 1996; Kirkpatrick et al., 1997; tion factors in response to stress. In this model, subtle Hirota et al., 1999; Wei et al., 2000), these results support changes in cellular redox potential induced by a the concept that a common central pathway(s) mediating stressing agent could alter the flow of electrons through cellular responses to IR or other types of environmental the cysteine residues of TR and TRX resulting in or metabolic oxidative stress (Abate et al., 1990; Tuttle et profound changes in protein activity. These critical al., 1992; Xanthoudakis and Curran, 1996; Spitz et al., cysteine(s) would appear to act as redox-sensitive 2000) may involve a series of redox-sensitive thiol ‘sulfhydryl switches’ that reversibly modulate protein containing signaling factors. activity (Xanthoudakis et al., 1992; Gallegos et al., 1996, Kirkpatrick et al., 1997; Hirota et al., 2000a,b) Materials and methods and allow signal transduction cascades to redirect metabolism in response to radiation-induced stress Cell culture and IR conditions using redox-sensitive transcription factors. The results presented in this study are consistent with HeLa (human cervical carcinoma) cell lines stably over- expressing TR were cultured in minimum essential medium the hypothesis that IR induces oxidative stress via the (alpha modification), containing 5% heat-inactivated (HI; formation of ROI, initiating a cellular response pathway 568C, 30 min) fetal bovine serum (FBS), penicillin (100 mg/ in the cytoplasm. In this model, IR is hypothesized to ml), streptomycin (100 U/ml), and G418 (300 mg/ml) in a increase the flow of electrons from NADPH to TR, humidified 378C incubator with 5% CO2 and air. MCF-7 resulting in an increase in reduced TRX in the nucleus, (human breast adenocarcinoma) cells stably overexpressing which passes the redox signal to the nuclear redox factor- TRX were grown in minimum essential medium containing 1 protein (Ref-1), which subsequently activates AP-1 Earle’s salts, 5% HI FBS, penicillin (100 mg/ml), streptomy- DNA binding activity and gene transcription (Figure 8). cin (100 U/ml), and G418 (1 mg/ml) in a humidified 378C This model would appear to present a possible incubator with 5% CO2 and air. Prior to irradiation, HeLa mechanism linking IR-induced changes in intracellular and MCF-7 cells were serum-starved for 48 h in medium containing 0.5% HI FBS in the absence of G418. A Pantak metabolism through the Pentose Cycle to activation of 250 kVp X-ray generator was used to deliver 10 Gy of unfiltered X-rays in a humidified 378C incubator with 5% CO2 and air. Unirradiated, sham-treated control cells were placed into a similar nearby incubator during IR exposures.

Expression plasmids and stably transfected cell lines The TRX wild type and cysteine to serine mutant expression plasmids have been previously described. The pcDNA3-X, pcDNA-TRX-wt, and pcDNA-TRX-dm were cotransfected with pcDNA-Neo3.1 into MCF7 cells followed by selection with 1 mg/ml G418 (Hu et al., 2001). All TRX and TR over expressing cell lines used in the study are pools of at least 150 independent clones.

Preparation of whole cell and subcellular fractions, SDS – PAGE and immunoblot analysis Whole cell, nuclear, and cytoplasmic extracts were prepared using a modification of the method previously described (Dignam, 1990; Curry et al., 1999). Protein concentrations were determined using the Bradford assay (BioRad Labs., Hercules, CA, USA) using a Beckman DU-640 spectro- Figure 8 Hypothetical model of the signaling pathway involved photometer. Following preparation and protein analysis in ionizing radiation-induced alterations in AP-1 DNA binding and AP-1 dependent reporter gene expression. We have previously samples were stored at 7808C and thawed on ice shown that following exposure to IR, thioredoxin is translocated immediately before use. Constitutive expression of TRX or into the nucleus, forms a physical interaction with Ref-1 and Ref- c-myc-tagged TR in the stably transfected cells was verified 1 appears to regulate AP-1 DNA-binding activity (Wei et al., by immunoblot analysis. Following harvest of whole cell 2000). In addition, the ability of both TRX and Ref-1 immuno- extracts, equal protein amounts (10 – 20 mg/sample) were precipitated from irradiated cells to stimulate AP-1 DNA-binding mixed with Laemmli lysis buffer, boiled for 5 min, and in vitro was found to be sensitive to alterations in thiol oxidation/ loaded into denaturing SDS-polyacrylamide gels for electro- reduction status (Wei et al., 2000). The current report clearly de- phoresis. Following separation, protein samples were monstrates that the critical N-terminal cysteine residues in both transferred to nitrocellulose membranes using an Owl semi- TRX and TR are essential for the regulation of AP-1 DNA-bind- ing activity and in vivo reporter gene expression following expo- dry transfer apparatus (Owl Scientific, Portsmouth, NH, sure to radiation. In this model, it appears that the critical USA) and blocked for 1 h in a 5% milk in PBS – 0.05% cysteines in TR, TRX, Ref-1 and c-fos/c-jun are involved in a re- Tween (PBS-T) solution. Filters with MCF-7 samples were dox sensitive signaling cascade regulating the activity of the AP-1 incubated overnight with a mouse anti-human TRX mono- nuclear transcription factor following exposure to radiation clonal antibody (Serotec, Raleigh, NC, USA) diluted 1 : 250

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6326 in 2.5% milk in PBS-T. HeLa blots were hybridized 1 mg of the b-galactosidase expression plasmid (pCMV-b-gal) overnight with a 1 : 500 dilution of mouse monoclonal was used. Thirty-six hours after transfection, cells were antibody against c-myc (Zymed Labs, San Francisco, CA, exposed to 10 Gy IR and harvested after 8 h. Luciferase USA). Immunoreactive c-Fos and c-Jun protein levels were activity was determined using a luminometer (Zylux Corp, determined using 1 : 1000 dilutions of polyclonal antibodies to Maryville, TN, USA). b-galactosidase activity was also c-Fos (AP1; Oncogene Science, Cambridge, MA, USA) and determined (Promega, Madison, WI, USA), and the c-Jun (AP2; Oncogene Science). The membranes were washed relative-fold induction of luciferase activity was calculated three times for 15 min each in PBS-T and incubated with by normalizing to b-galactosidase activity. appropriate ECL-sensitive, horseradish -conju- gated secondary antibodies (Santa Cruz Biotech, Santa Indirect cellular immunofluorescence Cruz, CA, USA) diluted 1 : 500 in 2.5% milk in PBS-T. Bands were analysed using an ECL protocol (Amersham pCXN2-myc-X, pCXN2-myc-TR-wt, and pCXN2-myc-mTR Pharmacia, Newark, NJ, USA) and visualized on radio- permanently transfected HeLa cells were seeded onto 1.5 ml of graphic film (Eastman-Kodak, Rochester, NY, USA). growth media in two-chambered Lab-Tek slides (Nalge Nunc, Figures are representative of the outcome in at least two Rochester, NY, USA) at a density of 16104 cells per chamber. independent experiments. Cells were allowed to attach overnight in a 378C, 5% CO2 incubator, after which the serum concentration was reduced to 0.5% FBS and culture resumed for 48 h. Cells were then Thioredoxin reductase enzymatic assay exposed to 10 Gy of IR or sham-irradiated and returned to Thioredoxin reductase activity was determined by method 378C to allow for TRX nuclear localization. Following a 1-h previously developed (Holmgren and Bjornstedt, 1995). recovery period, culture media was aspirated and cells were Briefly, cell extracts from log-phase cells were prepared by washed two times sequentially in 16PBS; this serial wash step freeze-thaw lysis, quantified as described (Curry et al., 1999), was repeated following each stage of the slide processing. Cells and 20 mg of extract was incubated with insulin, NADPH, were fixed in 4% paraformaldehyde/PBS for 3 min and and thioredoxin in 0.2 M HEPES, pH 7.6 for 20 min at 378C. permeabolized in 0.1% NP-40/PBS for 5 min. Non-specific Reactions were terminated after the addition of 6 M binding was blocked for 1 h with 1% BSA/PBS, then cells were guanidinium hydrochloride/0.4 mg/ml dithiobis (2-nitroben- incubated for 4 h at room temperature with a mouse anti- zoic acid) prepared in 0.2 M Tris, pH 8.0. In each case, a human TRX antibody diluted 1 : 50 in 1% BSA/PBS. Cells corresponding experimental sample without TRX was used to were then incubated with a fluorescein-conjugated anti-mouse correct for the basal level of TR activity (due to endogenous secondary antibody (Santa Cruz Biotech, Santa Cruz, CA, TRX and NADPH). Also, reactions without cell extracts and USA) diluted 1 : 50 in 1% BSA/PBS for 1 h in the dark at room reactions with pure TR in place of cell extracts were used as temperature. 4’-6-Diamidino-2-phenylindole-2HCl (DAPI) negative and positive controls, respectively. Triplicate was used at a concentration of 1 mg/ml for 30 min as a nuclear samples were measured for enzymatic activity by spectro- counterstain, and anti-fade reagent (Dako Corp., Carpinteria, photometric absorbance at 412 nm. CA, USA) was applied dropwise prior to overlaying the slide surfaces with coverslips. Cells were analysed on a Leica (Wetzlar, Germany) DMRXA fluorescent microscope at Electromobility shift assay (EMSA) 4006 resolution using a green filter for fluorescein visualiza- Protein samples were prepared, stored, and quantified by tion and a blue filter for DAPI. Digital images at both methods depicted earlier. No reducing agents were added to the wavelengths were acquired from fluorescence-excited fields EMSAs or the buffers used to make nuclear extracts. EMSAs using a Photometrics Sensys (Roper Scientific, Tuscon, AZ, were performed as previously described utilizing a 32P- USA) CCD camera, imported into IP Labs image analysis radiolabeled oligonucleotide corresponding to the consensus software package (Scanalytics, Inc., Fairfax, VA, USA), and AP-1 DNA-binding site (Wei et al., 2000). Nuclear extracts merged using Adobe Photoshop to confirm proximity of the (10 mg) were incubated with poly dI-dC for 10 min on ice, nucleus. Only fluorescein-excited images, representing TRX followed by the addition of radiolabeled oligonucleotide antibody hybridization, are shown. (100 000 c.p.m. of radiolabeled probe per reaction) and incubation at 258C for 20 min. Samples were electrophoresed on a 4.5% non-denaturing PAGE gel, dried, exposed to a phosphorimager screen, and analysed using a STORM 840 Phosphorimager (Molecular Dynamics, Sunnyvale, CA, USA). Acknowledgements The authors are indebted to Dr Jeffery Russell for the Plasmids, co-transfections, and luciferase assays invaluable technical expertise offered at stages throughout HeLa and MCF7 cells were plated at 26106 cells per this project. D Gius was supported by grants from the 100-mm plate, serum starved (0.5% fetal calf serum) for National Institute of Health (1 K08 CA72602-01), and the 8 h, and transfected via calcium phosphate precipitation (Wei American Cancer Society (ACS-IRG-58-010-43 and ACS et al., 2000). In each transfection, 2 mg of p7x-AP-1-tk-LUC RPG-00-292-01-TBE). DR Spitz was supported by NIH containing seven copies of the AP-1 DNA-binding site R01 HL51469. NIH CA69593 supported P Goswami. DV upstream of the luciferase gene in ptk-LUC (Stratagene, Kalvakolanu was supported by grants from the National Inc, La Jolla, CA, USA) was added. As an internal control, Institute of Health (R01 CA 78282 and CA 71401).

References

Abate C, Patel L, Rauscher III FJ and Curran T. (1990). Abate C, Marshak DR and Curran T. (1991). Oncogene, 6, Science, 249, 1157 – 1161. 2179 – 2185.

Oncogene Radiation-induced regulation of AP-1 by thioredoxin reductase and thioredoxin S Karimpour et al 6327 Becker K, Gromer S, Schirmer RH and Mu¨ ller S. (2000). Karin M and Smeal T. (1992). Trends Biochem. Sci, 17, 418 – Eur. J. Biochem., 267, 6118 – 6125. 422. Beetz A, Peter RU, Oppel T, Kaffenberger W, Rupec RA, Kerppola T and Curran T. (1995). Nature, 373, 199 – 200. Meyer M, van Beuningen D, Kind P and Messer G. (2000). Kerr LD, Inoue J and Verma IM. (1992). Curr. Opin. Cell Int. J. Radiat. Biol., 76, 1443 – 1453. Biol., 4, 496 – 501. Berggren M, Gallegos A, Gasdaska JR, Gasdaska PY, Kirkpatrick DL, Ehrmantraut G, Stettner S, Kunkel M and Warneke J and Powis G. (1996). Anticancer Res., 16, Powis G. (1997). Oncol. Res., 9, 351 – 356. 3459 – 3466. Maity A, McKenna WG and Muschel RJ. (1994). Radiother. Curry HA, Clemens RA, Shah S, Bradbury CM, Botero A, Oncol., 31, 1 – 13. Goswami P and Gius D. (1999). J. Biol. Chem., 274, Ma X, Karra S, Lindner DJ, Hu J, Reddy SP, Kimchi A, 23061 – 23067. Yodoi J and Kalvakolanu DV. (2001). Oncogene, 20, Dignam JD. (1990). Meth. Enzymol., 182, 194 – 203. 3703 – 3715. Gallegos A, Gasdaska JR, Taylor CW, Paine-Murrieta GD, Martin M, Vozenin MC, Gault N, Crechet F, Pfarr CM and Goodman D, Gasdaka PY, Berggren M, Briehl MM and Lefaix JL. (1997). Oncogene, 18, 981 – 989. Powis G. (1996). Cancer Res., 56, 5765 – 5770. Mustacich D and Powis G. (2000). Biochem. J., 346, 1–8. Franza Jr BR, Rauscher III FJ, Josephs SF and Curran T. Oberley TD, Verwiebe E, Zhong W, Kang SW and Rhee SG. (1988). Science, 239, 1150 – 1153. (2001). Free Radic. Biol. Med., 30, 412 – 424. Gius D, Cao XM, Rauscher III FJ, Cohen DR, Curran T and Poole LB, Reynolds CM, Wood ZA, Karplus PA, Ellis HR Sukhatme VP. (1990). Mol. Cell. Biol., 10, 4243 – 4255. and Calzi ML. (2000). Eur. J. Biochem., 267, 6126 – 6133. Hallahan DE, Gius D, Kuchibhotla J, Sukhatme V, Kufe Powis G, Kirkpatrick DL, Angulo M and Baker A. (1998). DW and Weichselbaum RR. (1993). J. Biol. Chem., 268, Chem. Biol. Interact., 112, 23 – 34. 4903 – 4907. Sasada T, Sono H and Yodoi J. (1996). J. Toxicol. Sci., 21, Halliwell B and Gutteridge JM. (1988). Hum. Toxicol., 7, 7– 285 – 287. 13. Spitz DR, Sim JE, Ridnour LA, Galoforo SS and Lee YJ. Halliwell B and Gutteridge JM. (1990). Methods Enzymol., (2000). Ann. NY Acad. Sci., 899, 349 – 362. 186, 1 – 85. Storz G, Tartaglia LA and Ames BN. (1990). Science, 248, HirotaK,MurataM,SachiY,NakamuraH,TakeuchiJ, 189 – 194. Mori K and Yodoi J. (1999). J. Biol. Chem, 274, 27891 – Tuttle SW, Varnes ME, Mitchell JB and Biaglow JE. (1992). 27897. Int. J. Radiat. Onc. Biol. Phys., 22, 671 – 675. Hirota K, Matsui M, Murata M, Takashima Y, Cheng FS, Wei SJ, Botero A, Hirota K, Bradbury CM, Markovina S, Itoh T, Fukuda K and Yodoi J. (2000a). Biochem. Biophys. Laszlo A, Spitz DR, Goswami PC, Yodoi J and Gius D. Res. Comm., 274, 177 – 182. (2000). Cancer Res., 60, 6688 – 6695. Hirota K, Nakamura H, Arai T, Ishii H, Bai J, Itoh T, Williams CH, Arscott DL, Mu¨ ller S, Lennon BW, Ludwig Fukuda K and Yodoi J. (2000b). Biochem. Biophys. Res. ML, Wang PM, Veine DM, Becker K and Heiner Schirmer Comm., 275, 825 – 830. R. (2000). Eur. J. Biochem., 267, 6110 – 6119. Holbrook NJ and Fornace Jr AJ. (1991). New Biol., 3, 825 – Xanthoudakis S and Curran T. (1992). EMBO J., 11, 653 – 833. 664. Holmgren A and Bjornstedt M. (1995). Meth. Enzymol., 252, Xanthoudakis S, Miao G, Wang F, Pan Y-C and Curran T. 199 – 208. (1992). EMBO J., 11, 3323 – 3335. Hu J, Ma X, Lindner DJ, Karra S, Hofmann ER, Reddy SP Xanthoudakis S and Curran T. (1996). Adv. Exp Med. Biol., and Kalvakolanu DV. (2001). Oncogene, 20, 4235 – 4248. 387, 69 – 75.

Oncogene