Endoplasmic Reticulum Stress Stimulates the Expression of
Cyclooxygenase-2 through Activation of NF-
B and pp38
Mitogen-activated Protein Kinase*
Received for publication, March 31, 2004, and in revised form, August 6, 2004 Published, JBC Papers in Press, August 19, 2004, DOI 10.1074/jbc.M403568200
Jui-Hsiang Hung‡§, Ih-Jen Su§¶, Huan-Yao Lei储, Hui-Ching Wang§, Wan-Chi Lin‡, Wen-Tsan Chang‡, Wenya Huang**, Wen-Chang Chang‡‡, Yung-Sheng Chang‡§, Ching-Chow Chen§§, and Ming-Derg Lai‡¶¶
From the ‡Department of Biochemistry, the储Department of Microbiology and Immunology, §Institute of Basic Medicine,
the **Department of Medical Technology, and the ‡‡Department of Pharmacology, College of Medicine, National Cheng
Kung University,¶Division of Clinical Research, National Health Research Institute, Tainan 701, and the §§Department
of Pharmacology, College of Medicine, National Taiwan University, Taipei 100, Taiwan, Republic of China
Expression of mutant proteins or viral infection may interfere with proper protein folding activity in the en-doplasmic reticulum (ER). Several pathways that main-tain cellular homeostasis were activated in response to these ER disturbances. Here we investigated which of these ER stress-activated pathways induce COX-2 and potentially oncogenesis. Tunicamycin and brefeldin A, two ER stress inducers, increased the expression of COX-2 in ML-1 or MCF-7 cells. Nuclear translocation of NF-B and activation of pp38 MAPK were observed dur-ing ER stress. IB␣ kinase inhibitor Bay 11-7082 or IB␣ kinase dominant negative mutant significantly inhib-ited the induction of COX-2. pp38 MAPK inhibitor SB203580 or eIF2␣ phosphorylation inhibitor 2-amin-opurine attenuated the nuclear NF-B DNA binding ac-tivity and COX-2 induction. Expression of mutant hepa-titis B virus (HBV) large surface proteins, inducers of ER stress, enhanced the expression of COX-2 in ML-1 and HuH-7 cells. Transgenic mice showed higher expres-sion of COX-2 protein in liver and kidney tissue express-ing mutant HBV large surface protein in vivo. Similarly, increased expression of COX-2 mRNA was observed in human hepatocellular carcinoma tissue expressing mu-tant HBV large surface proteins. In ML-1 cells express-ing mutant HBV large surface protein, anchorage-inde-pendent growth was enhanced, and the enhancement was abolished by the addition of specific COX-2 inhibitors. Thus, ER stress due either to expression of viral surface proteins or drugs can stimulate the expression of COX-2 through the NF-B and pp38 kinase pathways. Our results provide important insights into cellular carcinogenesis associated with latent endoplasmic reticulum stress.
The endoplasmic reticulum (ER)1stress response is a mech-anism by which cells respond to excess unfolded proteins in the
ER (1–5). The ER stress response is suggested to contribute to several types of human disease, including degenerative neuro-nal disorders (5–7) and type II diabetes (8, 9). ER stress re-sponse is induced by overexpression of exogenous membrane or secretory proteins, such as virus gene-encoded proteins (10 – 12). Hepatitis C virus and Japanese encephalitis virus infection initiate endoplasmic reticulum stress (13–15). Drugs that per-turb ER function (e.g. glycosylation inhibitors such as tunica-mycin (TM) or disulfide bond reducing agents such as 2-mer-captoethanol) can be used to study the following two types of ER signal pathways (1): the unfolding protein response (UPR), and the ER-overloaded response (EOR) pathway.
The UPR pathway has three components in mammalian cells: basic leucine zipper transcription factor ATF6, IRE1 RNA-processing enzyme, and ER localized kinase (PERK). ATF6 is synthesized as an ER transmembrane protein and is cleaved to generate cytosolic transcription factors that migrate to the nucleus. ATF6 cooperates with transcription factor NF-Y to bind mammalian ER stress-responsive elements (ERSE) to UPR-responsive gene promoters, such as GRP78 (16 –18). Mammalian IRE1 ribonuclease is activated by accumulation of unfolded protein in the endoplasmic reticulum. By removing a 26-nucleotide intron from XBP-1 mRNA, active IRE1 produces a novel XBP-1 mRNA encoding a transcription factor that can act via ERSE to activate the transcription of many UPR-re-sponsive genes (19 –22). In addition to activation of the ERSE-related transcription pattern, ER stress also alters the trans-lational pattern through PERK. The C-terminal cytoplasmic kinase domain of PERK can directly phosphorylate translation factor eIF2␣ and cause translational repression in response to an upstream ER stress signal (23, 24).
The EOR pathway triggers the activation of transcription factor NF-B (1, 14, 25–28). The activation of NF-B has been suggested to require the release of calcium from the ER and the production of reactive oxygen species (28). STAT3 transcription factor may mediate part of the activation of the JAK-STAT
* This work was supported in part by Grant NHRI-EX92-9116-BP1 from the National Health Research Institute (to H.-Y. L.) and by the Program for Promoting University Academic Excellence Projects 91-B-FA09 –1-4 (to W.-C. C.). 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 accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
¶¶To whom correspondence should be addressed: Dept. of Biochem-istry, College of Medicine, National Cheng Kung University, Tainan 701, Taiwan, Republic of China. Fax: 886-6-2741694; E-mail: [email protected].
1
The abbreviations used are: ER, endoplasmic reticulum; MAPK, mitogen-activated protein kinase; HBV, hepatitis B virus; COX,
cy-clooxygenase; TM, tunicamycin; UPR, unfolding protein response; EOR, ER-overloaded response; ESRE, ER stress-responsive elements; JNK, c-Jun N-terminal kinase; PDTC, pyrrolidine dithiocarbamate; ELISA, enzyme-linked immunosorbent assay; RT, reverse transcription; DMEM, Dulbecco’s modified Eagle’s medium; PBS, phosphate-buffered saline; FITC, fluorescein isothiocyanate; EMSA, electrophoretic mobil-ity shift assay; PG, prostaglandin; 2-AP, 2-aminopurine; pon-A, ponas-terone A; LMB, leptomycin B; ERK, extracellular signal-regulated ki-nase; PERK, phosphorylated ERK; eIF, eukaryotic initiation factor; EIA, enzyme immunoassay; HBsAg, hepatitis B surface antigen; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.
© 2004 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A.
This paper is available on line at http://www.jbc.org
46384
at National Taiwan University on June 18, 2009
www.jbc.org
pathway by reactive oxygen species (14). Activation of NF-B is also reported to be mediated via tumor necrosis factor receptor-associated factor 2 and c-Jun N-terminal kinase (JNK) (29, 30). Although the EOR and UPR pathways are distinct, they are related. Overexpression of IRE1 can activate NF-B, and dom-inant negative IRE1 can inhibit activation of NF-B (29). Phos-phorylation of the␣ subunit of eukaryotic initiation factor 2 by PERK is required for activation of NF-B in response to endo-plasmic reticulum stress (31). Activation of NF-B by ER stress leads to induction of many cellular genes that are largely anti-apoptotic in function. Latent long term expression of many viral surface proteins in mammalian cells may lead to cellular carcinogenesis, which in turn may be partly associated with the ER stress induced by these proteins (12, 14).
Overexpression of cyclooxygenase (COX)-2 has been found in many types of cancer and was linked to disease progression and drug resistance (32–38). Overexpression of COX-2 is sufficient to induce tumorigenesis or sensitize mouse skin for carcinogen-esis (39, 40). Cyclooxygenase-2 expression is regulated through multiple pathways including NF-B, C/EBP transcription fac-tors, and mitogen-activated protein kinases (41– 46). Induction of COX-2 mRNA is regulated by NF-B in macrophages (41– 44). The expression of COX-2 is correlated with the increase of NF-B activity, and induction of COX-2 by interleukin-1 is mediated partly by NF-B in colorectal cancer cells (45, 46).
Because NF-B is induced by ER stress, and NF-B can regulate the expression of COX-2, we hypothesized that ER stress may induce the expression of COX-2 and regulate cellu-lar homeostasis. In this report, we demonstrated ER stress induced by tunicamycin (TM) and brefeldin A leads to in-creased expression of COX-2. The induction of COX-2 was mediated through NF-B and p38 pathways. Furthermore, ER stress induced by expression of hepatitis B virus surface pro-tein also enhanced the expression of COX-2 in vitro. Mutant hepatitis B virus surface protein expression induced the in vivo expression of COX-2 in transgenic mice. Finally, expression of mutant HBV large surface proteins enhanced anchorage-inde-pendent growth of hepatocytes, which is deanchorage-inde-pendent on the induction of COX-2.
EXPERIMENTAL PROCEDURES
Chemicals and Kits—Tunicamycin, brefeldin A, actinomycin D,
pep-statin, sodium orthovanadate, leupeptin, dithiothreitol, ethidium bro-mide, SDS, Bay 11-7082, pyrrolidine dithiocarbamate (PDTC), and leptomycin B were products of Sigma. PD98059, SD203580, and JNK inhibitor II were purchased from Calbiochem. Ponasterone A was from Stratagene (La Jolla, CA). [␥-32P]ATP (6000 Ci/mmol) and ECL West-ern blot detection system were from Amersham Biosciences. The pros-taglandin E2EIA kit was from Cayman Chemical (Ann Arbor, MI). HBsAg enzyme-linked immunosorbent assay (ELISA) kit was from General Biological Corp. (Taipei, Taiwan). RT-PCR reagent and G418 was from Promega (Madison, WI). The NE-PER nuclear and cytoplas-mic extraction reagents kit and Micro BCATMprotein assay reagent kit were from Pierce. Anti-COX-2 and anti-GRP78 were purchased from Transduction Laboratories. Anti-p38, anti-phospho-p38, anti-ERK, ti-phospho-ERK, anti-JNK, anti-phospho-JNK, anti-eIF2␣, and an-tiphospho-eIF2␣ antibodies were from Cell Signaling (Beverly, MA). Anti-IB␣, anti-p65, and anti-RelB antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA). Anti-p50 was from Up-state Biotechnology, Inc. (Lake Placid, NY). Anti-␣-tubulin was from MDBio (Frederick, MD). Anti--actin was from Chemicon (Pittsburgh, PA). Anti-CDK4 was from Sigma. Anti-p50 and anti-p65 antibodies for EMSA supershift were from Santa Cruz Biotechnology. LipofectAMINE 2000, Dulbecco’s modified Eagle’s medium (DMEM), and antibiotic mixture (10,000 units of penicillin, 10,000 mg of streptomycin) were products of Invitrogen. Fetal bovine serum was purchased from Biolog-ical Industries (Beit Haemek, Israel).
Cell Culture and Treatments—ML-1, ML-1 PreS1⌬, ML-1 PreS2⌬, ML-1 vector HuH-7 inducible-PreS1⌬, HuH-7 inducible-PreS2⌬, HuH-7 inducible-vector, and MCF-7 cell lines were maintained at 37 °C in a 5% CO2atmosphere in DMEM supplemented with 10% heat-inactivated
fetal bovine serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100
g/ml streptomycin.
Plasmid and Stable Clone Cell Lines Construction—Plasmid p(3A)
SAg⌬1, p(3A) SAg⌬2, and pTK-neo were from Dr. Ih-Jen Su. ML-1 cells were co-transfected with p(3A) SAg⌬1/pTK-neo or p(3A) SAg⌬2/pTK-neo by using Invitrogen LipofectAMINE 2000 reagent according to the manufacturer’s protocol. Cells were then selected by G418 for 2 weeks. The p(3A) SAg⌬1 or p(3A) SAg⌬2 stable clone cell lines were established by HBsAg ELISA kit. HuH-7 inducible-PreS1⌬, HuH-7 inducible-PreS2⌬, and HuH-7 inducible-vector cell lines were obtained from Dr. Ih-Jen Su. IB kinase dominant negative mutant was kindly provided by Dr. Ching-Chow Chen (47).
Pon-A-inducible Expression of the Mutant Type
Pre-S—HuH-7-induc-ible PreS 1⌬, HuH-7-inducible PreS2⌬, and HuH-7-inducible vector cell lines were gifts from Dr. Ih-Jen Su. The Pre-S plasmid constructs contained ponasterone A (pon-A)-controlling elements. HuH-7 cells were co-transfected with these constructs and the vector pERV3, and stable clones were selected by G418 and hygromycin. The stable clones were treated with 10Mponasterone A for 0, 24, 48, 72, and 96 h, and then cell lysates of these were extracted for Western blotting.
Preparation of Cytosolic and Nuclear Extracts—ML-1 cells (1⫻ 106) in 10-cm dishes were incubated for 0, 6, 12, 18, and 24 h in serum DMEM containing 2.5g/ml tunicamycin. After treatment, the cells were washed with cold PBS, collected with a cell scraper, harvested by centrifugation, and then by using an NE-PER nuclear and cytoplasmic extraction reagents kit treated to extract their cytosolic and nuclear proteins.
Western Blot Analysis—Cell lysates were prepared by treating cells
with 2⫻ SDS lysis buffer (0.1M Tris (pH 6.8), 0.4% SDS, and 20% glycerol). The protein concentration of the supernatant was measured using a Micro BCATMprotein assay reagent kit. About 15–25g of cell lysates were separated by SDS-PAGE with 10% acrylamide and trans-ferred onto polyvinylidene fluoride membranes (Pierce). Following blocking with 5% nonfat dry milk for 1 h at room temperature and washing with Tween 20 with Tris-buffered saline (TTBS), the polyvi-nylidene fluoride membranes were incubated overnight at 4 °C with primary antibody in TTBS containing 1% bovine serum albumin. The second anti-mouse antibody-horseradish peroxidase conjugate (1:2000 dilution) was subsequently incubated with membranes for 1 h at room temperature and washed extensively for 40 –50 min with TTBS at room temperature. The blots were probed with the ECL Western blot detec-tion system according to the manufacturer’s instrucdetec-tions.
Immunofluorescence—5⫻ 105cells/well were plated in 4-well cham-bers in DMEM and treated with 2.5g/ml tunicamycin for 6 h. Cells for immunofluorescence microscopy of NF-B were fixed with 3.7% paraformaldehyde for 5 min and washed three times with PBS. Cells were then treated with ice-cold acetone for 1 min and washed three times with PBS. Cells were stained for NF-B translocation using anti-p65 antibody overnight at 4 °C and then anti-rabbit FITC-conju-gated antibody for 1 h. The negative control was cells stained with FITC-conjugated antibody alone. After staining with antibody, cells were viewed with a fluorescence microscope.
Electrophoretic Mobility Shift Assay (EMSA)—Oligonucleotides
cor-responding to the NF-B consensus sequences in the murine cox-2 promoter (5⬘-GAGGTGAGGGGATTCCCTTAGTTAG-3⬘) were synthe-sized, annealed, and end-labeled with [␥-32
P]ATP (6000 Ci/mmol; Amersham Biosciences) by T4 polynucleotide kinase. Nuclear protein (5 g) was incubated for 30 min at room temperature with 2 g of poly(dI-dC) from Amersham Biosciences, 4 l of gel shift binding 5⫻ buffer (20% glycerol, 5 mMMgCl2, 2.5 mMEDTA, 2.5 mMDTT, 50 mMTris-HCl (pH 7.5), 250 mM NaCl), and 100,000 cpm (1 ng) of a 32P-labeled oligonucleotide in a final volume of 20l. Supershift antibodies (2 g) were added as indicated. DNA-protein (NF-B) complexes were re-solved at 180 V for 4 h in a TBE-buffered, native 5% polyacrylamide gel, dried, and visualized with autoradiography using a Fuji Imaging plate in BAS-IP MS 2025 machine.
Transgenic Mice Tissue Protein Extraction—The transgenic mouse
liver, kidney, and muscle tissues were gifts from Dr. Ih-Jen Su. The Pre-S2⌬ transgenic mice were constructed by injection of Pre-S2⌬ gene fragment into the male pronucleus of fertilized mouse ova. Microinjec-tion was performed in Fvb/n mice. After 15 months, liver, kidney, or muscle tissue from Pre-S2⌬ transgenic mice was homogenized in RIPA buffer (50 mM Tris-HCl (pH 7.4), 1% Nonidet P-40, 0.25% sodium deoxycholate, 150 mMNaCl, 1 mMEDTA, 1 mMphenylmethylsulfonyl fluoride, 1g/ml aprotinin, leupeptin, and pepstatin, 1 mMNa3VO4, and 1 mMNaF). Homogenates were centrifuged at 15,000⫻ g for 10 min at 4 °C, and the supernatants were collected. Total protein concentrations of the tissue lysates were quantified using the Micro BCATM
protein
at National Taiwan University on June 18, 2009
www.jbc.org
assay reagent kit following the manufacturer’s instructions.
RT-PCR—After treatment, the cells were washed with cold PBS and
then cells were harvested. Total RNA was extracted from ML-1 cells using VIOGENE (total RNA extraction kit) according to the manufac-turer’s instruction. The cDNA was reverse-transcribed from 1g of total RNA using oligo(dT) primers and Moloney murine leukemia-virus transcriptase. The 5⬘ and 3⬘ primers for mouse cox-2-specific gene were 5⬘-ACT CAC TCA GTT TGT TGA GTC ATT-3⬘ (sense) and 5⬘-TTT GAT TAG TAC TGT AGG GTT ATT-3⬘ (antisense). The cycling parameters were as follows: 1 min at 94 °C for denaturation, 1 min at 52 °C for primer annealing, and 1 min at 72 °C for polymerization. Meanwhile, the same amount of cDNA was amplified for 25 cycles using specific glyceraldehyde-3-phosphate dehydrogenase primers: 5⬘-TGAAGGTCG-GTGTGAACGGATTTGGC-3⬘ (sense) and 5⬘-CATGTAGGCCATGAG-GTCCACCAC-3⬘ (antisense). The products were visualized after elec-trophoresis on a 1.5% agarose gel containing ethidium bromide. The signal level of the bands was quantified densitometrically.
RT-PCR for Human HBsAg Type II Hepatoma—The HBsAg type II
cells were obtained by micro-laser dissection, and total RNA was ex-tracted for RT-PCR. The cDNA was reverse-transcribed from 1g of total RNA using oligo(dT) primers and Moloney murine leukemia-virus transcriptase. The 5⬘ and 3⬘ primers for the human COX-2-specific gene were 5⬘-TTC AAA TGA GAT TGT GGG AAA ATT GCT-3⬘ (sense) and 5⬘-AGA TCA TCT GTG CCT GAG TAT CTT-3⬘ (antisense). The cycling parameters are as follows: 1 min at 94 °C for denaturation, 1 min at 52 °C for primer annealing, and 1 min at 72 °C for polymerization. Meanwhile, the same amount of cDNA was amplified for 25 cycles by using specific-actin primers: 5⬘-ATC ATG TTT GAG ACC TTC AA-3⬘ (sense) and 5⬘-CAT CTC TTG CTC GAA GTC CA-3⬘ (antisense). The products were visualized after electrophoresis on a 1.5% agarose gel containing ethidium bromide. The signal level of the bands was quan-tified densitometrically.
PGE2EIA Immunoassay—5⫻ 10
5cells/well were plated in 6-well dishes in DMEM and cultured for 24 h. PGE2levels in the supernatant conditioned medium were then assayed for using a prostaglandin (PG) E2EIA kit.
HBsAg ELISA Kit—Cells were washed with cold PBS and then
col-lected with a cell scraper and harvested by centrifugation. The superna-tant was removed; 100l of H2O was added, and then the cells were freeze-thawed three times at⫺80 °C. The level of HBsAg was determined using an ELISA kit following the manufacturer’s instructions.
Statistical Analysis—Results were presented as the mean⫾ S.D.,
and statistical comparisons were made using the Student’s t test. Sig-nificance was defined at the p⬍ 0.05 or 0.01 level.
RESULTS
ER Stress Induced the Expression of COX-2—Cells of the
mouse liver immortalized cell line ML-1 were treated with 2.5 MTM, and the expression of COX-2 was determined by West-ern blotting and RT-PCR. COX-2 mRNA started to increase 3 h after tunicamycin treatment (Fig. 1A), and COX-2 protein was induced 6 –12 h after treatment with either tunicamycin or another ER stress inducer brefeldin A. The expression of GRP78, an unfolded protein response chaperone, was enhanced in response to tunicamycin (Fig. 1B). Similarly, the expression of COX-2 protein was enhanced by tunicamycin in MCF-7 human breast cancer cells (Fig. 1C).
COX-2 Induction Was Transcription-dependent and Nuclear Export-dependent—To elucidate the mechanism of COX-2
in-duction, ML-1 cells were treated with tunicamycin with or without the transcriptional inhibitor actinomycin D. Actinomy-cin D alone did not alter COX-2 and did abolish tunicamyActinomy-cin- tunicamycin-induced COX-2 induction (Fig. 2A). The expression of COX-2 may be regulated at the nuclear export level (48); therefore, we studied whether the induction of COX-2 by tunicamycin can be inhibited by leptomycin B (LMB), a nuclear export inhibitor. LMB inhibited the COX-2 induction in a dose-dependent fash-ion within the range 2.5 to 10 M (data not shown). At the concentration of 5M, LMB can inhibit COX-2 induction more than 90% throughout the 24-h time course (Fig. 2B). On the other hand, the induction of GRP78 was inhibited by actino-mycin D but not leptoactino-mycin.
COX-2 Induction Is NF-B-dependent—Transcription factor
NF-B can regulate the expression of COX-2 and may be in-duced by ER stress. ML-1 hepatocytes were treated with tuni-camycin for the indicated time, and the amount of nuclear translocation of NF-B was determined by Western blot anal-ysis of the nuclear and cytosolic fraction. Increased nuclear translocation of the p50 and p65 subunits of NF-B was ob-served after tunicamycin treatment in ML-1 cells with different kinetics (Fig. 3A). Degradation of cytosolic IB␣ during ER stress may explain the nuclear translocation of NF-B (Fig. 3A). Nuclear translocation of the p65 subunit of NF-B was further confirmed by immunofluorescence in ML-1 cells and MCF-7 cells (Fig. 3B). Two forms of NF-B complexes were detected with gel shift analysis, and the upper gel-shifted band of NF-B DNA binding activity was strongly induced at 1.5 and 3 h after ER stress (Fig. 3C). Anti-p65 antibody supershifted the upper shifted band but did not affect the lower gel-shifted band. In contrast, anti-p50 antibody supergel-shifted the lower gel-shifted band completely, and partly decreased the intensity of the upper gel-shifted band. Control anti-c-Jun an-tibody did not affect the NF-B DNA binding activity. There-fore, the lower gel-shifted band may consist of p50/p50 ho-modimer only, and the upper gel-shifted band may contain both p65/p65 homodimer and p65/p50 heterodimer. Activation of NF-B appears to be mediated by the increase of p65/p65 and p65/p50 DNA binding activity (Fig. 3C). To determine whether NF-B is essential for the induction of COX-2, ML-1 cells were
FIG. 1. Elevated expression of COX-2 in response to
endo-plasmic reticulum stress. A, ML-1 cells were treated with
tunica-mycin for 0, 3, 6, 12, 18, and 24 h. Total RNA was isolated and then subjected to RT-PCR analysis. ML-1 (B) and MCF-7 cells (C) were treated with tunicamycin (TM) or brefeldin A (BFA) for various times, and cell lysates were analyzed by Western blotting with antibodies for COX-2, GRP78, and ␣-tubulin. GAPDH, glyceraldehyde-3-phos-phate dehydrogenase
at National Taiwan University on June 18, 2009
www.jbc.org
treated with the NF-B inhibitor PDTC. PDTC blocked the tunicamycin induction of COX-2 expression in a dose-depend-ent fashion (Fig. 4A). The time course for inhibition of COX-2 induction was measured in the presence of 50MPDTC (Fig. 4A). Another NF-B inhibitor Bay 11-7082, an IB kinase inhibitor, inhibited the COX-2 induction in a similar fashion. The degradation of IB␣ was attenuated by Bay 11-7082 (Fig. 4B). Furthermore, the induction of COX-2 was inhibited by the expression of the dominant negative mutant of IB kinase in ML-1 cells (Fig. 4C). PDTC, Bay 11-7082, or IB kinase domi-nant negative mutant did not alter the expression of GRP78, suggesting that NF-B is not involved in that branch of UPR pathway. These results altogether indicated that ER stress induced COX-2 in an NF-B-dependent pathway.
p38 MAPK Is Required for COX-2 Induction—As the
expres-sion of COX-2 is regulated through MAP kinases too, we inves-tigated whether the MAPKs played a role in the induction of COX-2. ML-1 cells were treated with 2.5g/ml tunicamycin, and the phosphorylated active form or total extracellular sig-nal-regulated kinase (ERK), p38, and JNK were examined with Western blotting. All three MAPKs were activated but with different kinetic fashion (Fig. 5A). Phosphorylated ERK (PERK) appeared at 2 h and phosphorylated JNK (pJNK) ap-peared at 6 h after tunicamycin treatment. On the other hand, phosphorylated pp38 occurred at an early time (30 min to 3 h) and stayed for 24 h (Fig. 5A). The requirement of various forms of MAPK was determined with the addition of 1–10MMAPK inhibitors, SB203580, PD98059, and JNK inhibitor II, respec-tively, and COX-2 levels were determined from Western blots. p38 MAPK inhibitor SB203580 as low as 1Mcan inhibit the induction of COX-2 (Fig. 5B). OneMERK inhibitor PD98059 had no effect, and higher doses (5 and 10M) did inhibit the induction (Fig. 5C). In contrast, JNK was not involved in the induction of COX-2 as the JNK inhibitor had no effect (Fig. 5D).
Induction of NF-B Is Dependent on pp38 MAPK and
Phos-phorylation of eIF2␣—Phosphorylation of the ␣ subunit of
eIF2␣ is required for the activation of NF-B in response to diverse stresses (31); therefore, we examined whether inhibi-tion of eIF2␣-phosphorylation by 2-aminopurine (2-AP) affects the expression of COX-2. 2-AP alone did not alter the expres-sion of COX-2 or GRP78 (data not shown). Both phosphoryla-tion of eIF2␣ and the expression of COX-2 was attenuated by 10 mM2-AP (Fig. 6A). 2-AP delayed and inhibited the nuclear translocation of p65 and p50 subunits of NF-B (Fig. 6B).
EMSA analysis further confirmed significant inhibition of NF-B DNA binding activity by 2-AP at 1.5 and 3 h after ER stress (Fig. 6C). On the other hand, pp38 kinase inhibitor SB203580 only mildly delayed the nuclear translocation of p50 and p65 subunits of NF-B but significantly inhibited NF-B DNA binding activity at 1.5 and 3 h after ER stress (Fig. 6, B and C). The JNK inhibitor had no effect on either COX-2 expression or NF-B DNA binding activity (Figs. 5D and 6C). Therefore, activation of NF-B DNA binding activity is medi-ated through eIF2␣ and pp38 MAPK.
Hepatitis B Virus Mutant Large Surface Protein Can Induce COX-2 in Vitro—ER stress can be induced by either drugs such
as tunicamycin or by overexpression of mutant proteins includ-ing virus gene products. In hepatitis B virus-associated tocarcinoma, deletions in the Pre-S1 or Pre-S2 region of hepa-titis B virus large surface proteins have been detected (49). These hepatitis B virus mutant large surface proteins were
FIG. 2. Transcription and nuclear export are required for COX-2
induction. ML-1 cells were incubated with tunicamycin and/or
transcrip-tional inhibitor actinomycin D (act D) (A) and the nuclear export inhibitor leptomycin B (Lep B) (B). The expressions of COX-2 and GRP78 were analyzed by Western blotting.␣-Tubulin is an internal control.
FIG. 3. Activation of NF-B during ER stress. A, ML-1 cells were
incubated with tunicamycin for 0, 1.5, 3, 6, 12, 18, and 24 h, and NF-B subunits in the cytoplasmic and nuclear fractions were analyzed by Western blotting with antibodies against subunits p65, p50, Rel-B, and IB␣. Cyclin-dependent kinase 4 (CDK4) and ␣-tubulin were used as internal markers for nuclear and cytoplasmic proteins. B, ML-1 cells were treated with tunicamycin for 6 h and probed with p65 anti-body and FITC-conjugated secondary antianti-body. Hoechst staining re-vealed the nucleus. C, nuclear NF-B DNA binding activity was meas-ured by EMSA using a probe corresponding to the NF-B-binding site of the murine cox-2 promoter. The arrows indicated the two NF-B com-plexes, and the arrowhead indicates the supershifted complexes with anti-p65 subunit or anti-p50 subunit antibody (Ab). Anti-c-Jun antibody serves as a control for supershift control.
at National Taiwan University on June 18, 2009
www.jbc.org
shown to induce ER stress (11, 12). Therefore, it is important to investigate whether the mutant large surface protein can in-duce the expression of COX-2. The expressions of COX-2 and GRP78 were examined in the following cell lines: ML-1 cells, ML-1 transfectants expressing two types of Pre-S mutant sur-face proteins (49), and an ML-1 control transfectant. The ex-pression of COX-2 was elevated in ML-1 cells expressing hep-atitis mutant surface proteins but not in control transfectants (Fig. 7A). One of the major products of COX-2, prostaglandin E2, was measured in ML-1 cells expressing Pre-S mutant sur-face proteins. Prostaglandin was elevated 4 –5-fold by the ex-pression of Pre-S mutant surface proteins (Fig. 7B). That the expression of COX-2 induced by hepatitis B mutant large sur-face proteins requires the activation of pp38 MAPK and
tran-FIG. 5. Activation of p38 MAPK is essential for COX-2
induc-tion during ER stress. A, ML-1 cells were treated with TM for the
times indicated, and the activation of MAPK pathways was determined by Western blot using antibodies specific for active forms of MAPK or total MAPK. COX-2, GRP78, and␣-tubulin were indicators for endo-plasmic reticulum stress and internal control. B–D, ML-1 cells were treated with tunicamycin as follows: B, pp38 inhibitor SB203580 (SB);
C, ERK inhibitor PD98059 (PD); D, JNK inhibitor, and the expression
of COX-2 and GRP78 was determined with Western blotting analysis. The amount of COX-2 induction and attenuation by pp38 inhibitor SB203580 was semi-quantitated by densitometry scanning in three separate experiments. *, p⬍ 0.05; **, p ⬍ 0.01.
FIG. 4. Attenuation of COX-2 expression by NF-B inhibitors
and IB kinase dominant negative mutant. A, ML-1 cells were
treated with tunicamycin and increasing doses of PDTC or 50MPDTC
for various times. B, ML-1 cells were treated with tunicamycin and increasing doses of Bay 11-7082 or 50MBay 11-7082 for various times, and the expressions of COX-2, IB␣, and GRP78 were analyzed by Western blotting. C, ML-1 cells were transfected with IB kinase dom-inant negative mutant IKK2M or pCDNA3.1 vector control, and the expressions of COX-2 and GRP78 were measured.
at National Taiwan University on June 18, 2009
www.jbc.org
scription factor NF-B was demonstrated by the inhibition of COX-2 expression by p38 inhibitor SB203580 and NF-B in-hibitor Bay 11-7082 (Fig. 8A). The production of PGE2was also
significantly attenuated (Fig. 8B). To demonstrate further the role of the HBV mutant large surface proteins in the induction of COX-2 in vitro, mutant HBV large surface proteins were expressed in the presence of pon-A, an inducible promoter in HuH-7 cell lines. Stable transfectants and control transfectants were induced by the addition of pon-A, and the expressions of COX-2 and GRP78 were measured (Fig. 9A). Inducible expres-sion of mutant large surface proteins was quantitated by EIA against HBsAg (Fig. 9B). Mutant HBV large surface proteins can enhance the expression of COX-2 and GRP78.
Hepatitis B Virus Mutant Large Surface Protein Can Induce COX-2 in Vivo—To confirm further that deletion forms of
mu-tant HBV large surface proteins can induce COX-2 in vivo, we created transgenic mice that express the Pre-S2 deletion form of HBV large surface protein under the control of its native promoter. The expression of HBV large surface protein was detected in the liver and kidney (Fig. 9). Elevated expression of COX-2 was observed in liver and kidney tissue in male or female mice (Fig. 10). In addition, we used laser capture dis-section to isolate type II hepatoma cells expressing the Pre-S2 deletion form of HBV large surface proteins from human tumor tissue (48), and RT-PCR to quantify the COX-2 mRNA. High expression of COX-2 mRNA was observed in hepatocytes that express the deletion forms of HBV large surface proteins (Fig. 10B). These results altogether demonstrate that Pre-S mutant surface proteins can induce COX-2, and the induction is possi-bly mediated through ER stress, at least partly.
Hepatitis B Virus Mutant Large Surface Proteins Enhance Anchorage-independent Growth in Soft Agar—The expression
of COX-2 was associated carcinogenesis; therefore, the expres-sion of Pre-S mutant HBV large surface proteins may enhance cellular transformation through the generation of COX-2. The expression of Pre-S mutant surface proteins containing dele-tions in either the Pre-S1 region or Pre-S2 region enhanced the growth in soft agar (Fig. 11A), which is an indication for an-chorage-independent growth. Addition of COX-2 inhibitor
FIG. 6. Activation of NF-B is mediated through ␣ subunit of
eIF2␣ phosphorylation and p38 MAPK. ML-1 cells were treated
with TM and eIF2␣ phosphorylation inhibitor 2-AP or pp38 inhibitor SB203580 for the indicated times. A, the expression of phosphorylated form of eIF2␣, total eIF2␣, COX-2, IB␣, and GRP78 in total cell lysates was determined with Western blotting. B, NF-B complex subunits in the nuclear fractions were analyzed by Western blotting with antibod-ies against subunits p65, p50, Rel-B, and IB␣. Cyclin-dependent ki-nase 4 (CDK4) and␣-tubulin were used as internal markers for nuclear and cytoplasmic proteins respectively. C, nuclear NF-B DNA binding activity was measured with EMSA using a probe corresponding to the NF-B-binding site of the murine cox-2 promoter. The arrows indicated the two NF-B complexes. JNK inhibitor did not affect NF-B DNA binding activity, serving as a control.
FIG. 7. Increase of GRP78, COX-2 protein, and PGE2in ML-1
cells expressing del-PreS1 (Pre-S1⌬) and del-PreS2 (Pre-S2⌬)
mutant HBV large surface protein. A, the upper panel, Western blot
analysis of COX-2 protein and GRP78 expression in 1st lane, ML-1; 2nd
lane, ML-1-⌬1 (⌬1); 3rd lane, ML-1-⌬2 (⌬2); 4th lane, ML-1-neo (V), and 5th lane, ML-1 cells treated with tunicamycin; the lower panel shows
HBV large surface proteins expression as determined by HBsAg ELISA.
B, the level of COX-2 product PGE2in culture medium was determined by EIA. at National Taiwan University on June 18, 2009
www.jbc.org
etodolac to ML-1 cells expressing mutant surface proteins sig-nificantly inhibited their growth on soft agar (Fig. 11A). Induc-ible expression of mutant large surface proteins in HuH-7 cells also enhanced growth in soft agar, and the addition of etodolac completely abolished this growth (Fig. 11B).
DISCUSSION
ER stress can induce multiple signal pathways including the ATF6/IRE1 UPR pathway, eIF2␣ pathway, and NF-B EOR pathway (19 –28). These pathways may cross-talk with each other or converge on common downstream effectors (50). In this report, we have demonstrated that ER stress can induce the expression of COX-2, and the induction is dependent on the transcription factors NF-B and p38 MAPK. The enhancement of NF-B DNA binding activity also requires p38 MAPK acti-vation and eIF2␣ phosphorylation (Fig. 12). IRE1 or eIF2␣ is involved in activation of transcription factor NF-B during
endoplasmic reticulum stress (30, 31). Phosphorylation of eIF2␣ is also essential for activation of NF-B in our experi-mental system. NF-B and p38 MAPK may not be the up-stream signal for the ATF6 pathway, because the induction of GRP78 by ER stress was not affected by the inhibitors for p38 MAPK and NF-B. Recently, pp38 activation was reported to be mediated through IRE1 because ER stress-induced pp38 activation was attenuated in IRE1-deficient cells (51). There-fore, IRE1-dependent activation of NF-B may also partly me-diated through the p38 MAPK pathway (Fig. 12). Endoplasmic reticulum stress can be induced by N-glycosylation-inhibition through tunicamycin or by expression of mutant viral surface proteins such as hepatitis B virus large surface proteins. The induction of COX-2 by drugs or overexpression of mutant pro-teins was similarly mediated through p38 MAPK and NF-B. However, the details of regulation may not be similar (i.e. the pathway to NF-B activation may be different in tunicamycin-treated ML-1 cells and ML-1 cells expressing mutant HBV surface proteins). Preliminary data suggest that calcium ion is required for tunicamycin-induced NF-B but is not essential for the induction of NF-B in cells expressing HBV large sur-face proteins (data not shown).
Nuclear translocation of NF-B subunit p65 occurred ⬃1.5 h after ER stress, and the p50 subunit translocated later (Fig. 3A). NF-B DNA binding activity increased significantly at
FIG. 8. Attenuation of COX-2 expression and PGE2production
by p38 MAPK inhibitor (SB203580) and NF-B inhibitor (Bay 11-7082). A, ML-1 cells expressing PreS2⌬ were incubated with
SB203580 (SB) or Bay 11-7082 (Bay) for 24 h. The expression of COX-2 was determined by Western blotting. B, the PGE2in culture medium was determined by EIA. *, p⬍ 0.05.
FIG. 9. Inducible expression of hepatitis B large surface
pro-teins enhanced the expression of COX-2 in HuH-7 cells. PreS1⌬ and PreS2⌬ HBV large surface proteins were induced by addition of ponasterone A, and the expression of COX-2 and GRP78 (A) and HBV large surface proteins (B) was determined by Western blotting and EIA, respectively.
FIG. 10. Elevated COX-2 is associated with PreS2⌬ expression
in vivo. A, COX-2 expression was elevated in transgenic mice
express-ing PreS2 deletion (⌬2) HBV large surface protein. The expression of COX-2 in liver, kidney, and muscle tissues (C, control;⌬2, PreS2⌬) was determined by Western blotting. B, COX-2 mRNA expression was ele-vated in human HBsAg type II hepatoma cells expressing mutant HBV large surface protein. The normal and HBsAg type II hepatocytes were isolated by laser capture microdissection from hepatoma tissue. Total RNA was isolated, and RT-PCR was used to measure COX-2 and -actin.
at National Taiwan University on June 18, 2009
www.jbc.org
early times during ER stress and was mainly composed of p65/p65 and p65/p50 complexes (Fig. 3C). Degradation of IB was observed 3– 6 h after ER stress, which is a little later than the nuclear translocation of NF-B. This phenomenon may be because of the possibility that the phosphorylation of IB and release of p65 subunit precedes the degradation of IB. Alter-natively, other non-IKK kinases may be partly involve in phos-phorylation of IB and release of p65 (52). In our ER stress
model, activation of p38 MAPK may be involved in enhancing nuclear NF-B DNA binding activity because inhibition of p38 MAPK decreases the DNA binding activity of NF-B with mi-nor effects on nuclear translocation of NF-B subunits. Re-cently, p38 MAPK was shown to activate the NF-B transcrip-tional activity without affecting its nuclear translocation (53, 54), which may be similar to the role of p38 MAPK in this report. The appearance of NF-B in the nucleus began 1.5 h after ER stress which is consistent with the induction of COX-2 mRNA. COX-2 protein expression increased significantly 12–18 h after ER stress. The NF-B appears to be essential for the induction of COX-2 mRNA; however, NF-B alone may not be sufficient to fully induce the expression of COX-2 protein. P38 MAPK may not only act on transcriptional level of COX-2 mRNA but may also regulate the stability of COX-2 mRNA (55, 56). In addition, activation of p38 MAPK may further enhance the activation of the NF-B-dependent gene (57). Therefore, coordination of NF-B and p38 MAPK may be required for the full induction of COX-2 protein.
Alterations in endoplasmic reticulum homeostasis trigger a complex series of events, including synthesis of chaperone, de-crease of translation, and degradation of unfolded proteins to promote cellular survival. Endoplasmic reticulum stress induced p53 cytoplasmic localization and prevented p53-dependent apo-ptosis (58). Endoplasmic reticulum stress induced by glucose depletion could enhance the expression of phospho-glycoprotein (59), which may affect cancer outcome. Breast cancer cells can secrete pro-angiogenic vascular endothelial growth factor in re-sponse to ER stress (60). In this report, we demonstrated the following: 1) the expression of COX-2 may be induced by expres-sion of surface proteins in cells with latent virus infection or other stimuli, including drugs that perturb ER function, and 2) the COX-2 pathway may be used to enhance anchorage-independent growth. This further extends the impact of endoplasmic reticu-lum stress on cellular carcinogenesis.
Drug-induced ER stress leads to degradation of IB and to nuclear translocation of NF-B in hepatocytes and breast can-cer cells. Degradation of IB was not observed in the ER stress-induced activation of NF-B in mouse embryo fibroblasts (31). This minor discrepancy may be cell type-specific. Similarly, ER stress-induced GSK3 activation plays an opposite role in different cell types (61). The importance of NF-B in the induction of COX-2 has been demonstrated in many reports (41– 44, 62, 63). Furthermore, NF-B plays an important role in liver carcinogenesis (64, 65). Activation of NF-B was observed frequently in hepatocarcinoma, and the essential role of NF-B for cancer growth was confirmed in several human cancer cell lines (66, 67). Therefore, control of NF-B activity may be an important therapeutic target for the treatment of human hepatocarcinoma (68).
Because latent hepatitis B and hepatitis C virus infection is clearly associated with hepatocarcinogenesis, and the expres-sion of viral surface proteins induces endoplasmic reticulum stress and COX-2 expression, COX-2 is therefore a rational chemopreventive target for hepatocarcinoma. COX-2 inhibitor is a chemoprevention agent for colon cancer, and its use has been proposed recently (69, 70) to decrease the incidence of other types of cancer including gastric and lung cancer. COX-2 inhibitor is used to treat hepatoma (71, 72), but its role in prevention is not clearly defined (73). Recently a COX-2 inhib-itor was shown to prevent hepatocarcinogenesis in an animal model (74). Our results strongly suggest that preclinical inves-tigation of the effect of COX-2 inhibitors on HBV carrier status is warranted.
Induction of endoplasmic reticulum stress in carcinogenesis may be mediated through latent expression of viral surface
FIG. 11. Expression of PreS1⌬ and PreS2⌬ HBV large surface
protein enhanced anchorage-independent growth, which was abolished by COX-2 inhibitor. PreS1⌬ (⌬1) or PreS2⌬ (⌬2) HBV large surface proteins were stably expressed in ML-1 cells (A) or induc-ibly expressed in HuH-7 cells (B). Ten thousand cells were cultured in 6-well plates containing 0.35% soft agar, and the foci (⬎0.5 mm) num-ber was determined 3 weeks later with or without 2.5 mM COX-2 inhibitor etodolac. Inducer ecodynosine was added in the medium of HuH-7 transfectants for the induction of HBV large surface proteins. P, parental cell; V, vector control.
FIG. 12. Signaling of endoplasmic reticulum stress to COX-2.
Accumulation of unfolded or malfolded proteins in the endoplasmic reticulum leads to activation of multiple signaling pathways. First, activation of IRE1 and ATF6 represents the standard UPR pathways and turns on the expression of many downstream genes including
GRP78. Second, activation of ER localized kinase (PERK) to decrease
the translation rate is mediated through a change in eIF2␣ phospho-rylation. Third, activation of p38 MAPK and NF-B may alter cellular homeostasis by regulating multiple genes including cox-2. This path-way (indicated by heavy arrows) is demonstrated in this report. These three pathways are not independent from each other; for example, activation of NF-B is dependent on phosphorylation of eIF2␣ subunit and activation of p38 MAPK.
at National Taiwan University on June 18, 2009
www.jbc.org
proteins or by nutrient deprivation. These latent endoplasmic reticulum stresses induce many survival pathways, including sequestering of p53 to inactivate p53-dependent pathway, ex-pression of the pro-angiogenic factor (vascular endothelial growth factor), and COX-2 (shown in this report). Our study indicated that sites of ER stress should be considered impor-tant targets of future carcinogenesis investigations.
REFERENCES 1. Pahl, H. L. (1999) Physiol. Rev. 79, 683–701
2. Patil, C., and Walter, P. (2001) Curr. Opin. Cell Biol. 13, 349 –355 3. Chevet, E., Cameron, P. H., Pelletier, M. F., Thomas, D.Y., and Bergeron, J. J.
(2001) Curr. Opin. Struct. Biol. 11, 120 –124 4. Ron, D. (2002) J. Clin. Investig. 110, 1383–1388 5. Kaufman, R. J. (2002) J. Clin. Investig. 110, 1389 –1398 6. Paschen, W., and Frandsen, A. (2001) J. Neurochem. 79, 719 –725 7. Sherman, M. Y., and Goldberg, A. L. (2001) Neuron 29, 15–32 8. Araki, E., Oyadomari, S., and Mori, M. (2003) Intern. Med. 42, 7–14 9. Shi, Y., Taylor, S. I., Tan, S. L., and Sonenberg, N. (2003) Endocr. Rev. 24,
91–101
10. Xu, Z., Jensen, G., and Yen, T. S. (1997) J. Virol. 71, 7387–7392
11. Hsieh, Y. H., Su, I. J., Wang, H. C., Chang, W. W., Lei, H. Y., Lai, M. D., Chang, W. T., and Huang, W. (2004) Carcinogenesis 25, 2023–2032
12. Wang, H. C., Wu, H. C., Chen, C. F., Fausto, N., Lei, H. Y., and Su, I. J. (2003) Am. J. Pathol. 163, 2441–2449
13. Tardif, K. D., Mori, K., and Siddiqui, A. (2002) J. Virol. 76, 7453–7459 14. Waris, G., Tardif, K. D., and Siddiqui, A. (2002) Biochem. Pharmacol. 64,
1425–1430
15. Su, H. L., Liao, C. L., and Lin, Y. L. (2002) J. Virol. 76, 4162– 4171 16. Haze, K., Yoshida, H., Yanagi, H., Yura, T., and Mori, K. (1999) Mol. Biol. Cell
10, 3787–3799
17. Yoshida, H., Okada, T., Haze, K., Yanagi, H., Yura, T., Negishi, M., and Mori, K. (2000) Mol. Cell. Biol. 20, 6755– 6767
18. Li, M., Baumeister, P., Roy, B., Phan, T., Foti, D., Luo, S., and Lee, A. S. (2000) Mol. Cell. Biol. 20, 5096 –5106
19. Yoshida, H., Matsui, T., Yamamoto, A., Okada, T., and Mori, K. (2001) Cell 107, 881– 891
20. Shen, X., Ellis, R. E., Lee, K., Liu, C. Y., Yang, K., Solomon, A., Yoshida, H., Morimoto, R., Kurnit, D. M., Mori, K., and Kaufman, R. J. (2001) Cell 107, 893–903
21. Calfon, M., Zeng, H., Urano, F., Till, J. H., Hubbard, S. R., Harding, H. P., Clark, S. G., and Ron, D. (2002) Nature 415, 92–96
22. Lee, K., Tirasophon, W., Shen, X., Michalak, M., Prywes, R., Okada, T., Yoshida, H., Mori, K., and Kaufman, R. J. (2002) Genes Dev. 16, 452– 466 23. Harding, H., Zhang, Y., and Ron, D. (1999) Nature 397, 271–274
24. Shi, Y., Vattem, K. M., Sood, R., An, J., Liang, J., Stramm, L., and Wek, R. C. (1998) Mol. Cell. Biol. 18, 7499 –7509
25. Pahl, H. L., and Baeuerle, P. A. (1997) Trends Biochem. Sci. 22, 63– 67 26. Pahl, H. L., Sester, M., Burgert, H. G., and Baeuerle, P. A. (1996) J. Cell Biol.
132, 511–522
27. Pahl, H. L., and Baeuerle, P. A. (1995) EMBO J. 14, 2580 –2588 28. Pahl, H. L., and Baeuerle, P. A. (1996) FEBS Lett. 392, 129 –136
29. Leonardi, A., Vito, P., Mauro, C., Pacifico, F., Ulianich, L., Consiglio, E., Formisano, S., and Di-Jeso, B. (2002) Endocrinology 143, 2169 –2177 30. Kaneko, M., Niinuma, Y., and Nomura, Y. (2003) Biol. & Pharm. Bull. 26,
931–935
31. Jiang, H. Y., Wek, S. A., McGrath, B. C., Scheuner, D., Kaufman, R. J., Cavener, D. R., and Wek, R. C. (2003) Mol. Cell. Biol. 23, 5651–5663 32. Cao, Y., and Prescott, S. M. (2002) J. Cell. Physiol. 190, 279 –286
33. Davies, G., Salter, J., Hills, M., Martin, L. A., Sacks, N., and Dowsett, M. (2003) Clin. Cancer Res. 9, 2651–2656
34. Leng, J., Han, C., Demetris, A. J., Michalopoulos, G. K., and Wu, T. (2003) Hepatology 38, 756 –768
35. Shariat, S. F., Matsumoto, K., Kim, J., Ayala, G. E., Zhou, J. H., Jian, W., Benedict, W. F., and Lerner, S. P. (2003) J. Urol. 170, 985–989 36. Einspahr, J. G., Krouse, R. S., Yochim, J. M., Danenberg, P. V., Danenberg,
K. D., Bhattacharyya, A. K., Martinez, M. E., and Alberts, D. S. (2003) Cancer Res. 63, 3891–3893
37. Lara-Pezzi, E., Gomez-Gaviro, M. V., Galvez, B. G., Mira, E., Iniguez, M. A., Fresno, M., Martinez-A, C., Arroyo, A. G., and Lopez-Cabrera, M. (2002) J. Clin. Investig. 110, 1831–1838
38. Rahman, M. A., Dhar, D. K., Yamaguchi, E., Maruyama, S., Sato, T., Hayashi, H., Ono, T., Yamanoi, A., Kohno, H., and Nagasue, N. (2001) Clin. Cancer
Res. 7, 1325–1332
39. Liu, C. H., Chang, S. H., Narko, K., Trifan, O. C., Wu, M. T., Smith, E., Haudenschild, C., Lane, T. F., and Hla, T. (2001) J. Biol. Chem. 276, 18563–18569
40. Muller-Decker, K., Neufang, G., Berger, I., Neumann, M., Marks, F., and Furstenberger, G. (2002) Proc. Nat. Acad. Sci. U. S. A. 99, 12483–12488 41. Caivano, M., Gorgoni, B., Cohen, P., and Poli, V. (2001) J. Biol. Chem. 276,
48693– 48701
42. Mestre, J. R., Mackrell, P. J., Rivadeneira, D. E., Stapleton, P. P., Tanabe, T., and Daly, J. M. (2001) J. Biol. Chem. 276, 3977–3982
43. Wu, D., Marko, M., Claycombe, K., Paulson, K. E., and Meydani, S. N. (2003) J. Biol. Chem. 278, 10983–10992
44. Yeo, S. J., Gravis, D., Yoon, J. G., and Yi, A. K. (2003) J. Biol. Chem. 278, 22563–22573
45. Liu, W., Reinmuth, N., Stoeltzing, O., Parikh, A. A., Tellez, C., Williams, S., Jung, Y. D., Fan, F., Takeda, A., Akagi, M., Bar-Eli, M., Gallick, G. E., and Ellis, L. M. (2003) Cancer Res. 63, 3632–3636
46. Charalambous, M. P., Maihofner, C., Bhambra, U., Lightfoot, T., and Gooder-ham, N. J. (2003) Br. J. Cancer 88, 1598 –1604
47. Chen, C. C., Sun, Y. T., Chen, J. J., Chang, Y. J. (2001) Mol. Pharmacol. 59, 493–500
48. Jang, B. C., Munoz-Najar, U., Paik, J. H., Claffey, K., Yoshida, M., and Hla, T. (2003) J. Biol. Chem. 278, 2773–2776
49. Fan, Y. F., Lu, C. C., Chen, W. C., Yao, W. J., Wang, H. C., Chang, T. T., Lei, H. Y., Shiau, A. L., and Su, I. J. (2001) Hepatology 33, 277–286 50. Ma, Y., and Hendershot, L. M. (2004) J. Biol. Chem. 279, 13792–13799 51. Nguyen, D. T., Kebache, S., Fazel, A., Wong, H. N., Jenna, S., Emadali, A., Lee,
E. H., Bergeron, J. J., Kaufman, R. J., Larose, L., and Chevet, E. (2004) Mol. Biol. Cell 15, 4248 – 4260
52. Wong, B. C., Jiang, X., Fan, X. M., Lin, M. C., Jiang, S. H., Lam, S. K., and Kung, H. F. (2003) Oncogene 22, 1189 –1197
53. Watanabe, T., Jono, H., Han, J., Lim, D. J., and Li, J. D. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 3563–3568
54. Matsuyama, W., Wang, L., Farrar, W. L., Faure, M., and Yoshimura, T. (2004) J. Immunol. 172, 2332–2340
55. Faour, W. H., He, Y., He, Q. W., de-Ladurantaye, M., Quintero, M., Mancini, A., and Di-Battista, J. A. (2001) J. Biol. Chem. 276, 31720 –31731 56. Faour, W. H., Mancini, A., He, Q. W., and Di-Battista, J. A. (2003) J. Biol.
Chem. 278, 26897–26907
57. Saccani, S., Pantano, S., and Natoli, G. (2002) Nat. Immun. 3, 69 –75 58. Qu, L., Huang, S., Baltzis, D., Rivas-Estilla, A. M., Pluquet, O., Hatzoglou, M.,
Koumenis, C., Taya, Y., Yoshimura, A., and Koromilas, A. E. (2004) Genes Dev. 18, 261–277
59. Ledoux, S., Yang, R., Friedlander, G., and Laouari, D. (2003) Cancer Res. 63, 7284 –7290
60. Marjon, P. L., Bobrovnikova-Marjon, E. V., and Abcouwer, S. F. (2004) Mol. Cancer 3, 4 –16
61. Watcharasit, P., Bijur, G. N., Song, L., Zhu, J., Chen, X., and Jope, R. S. (2003) J. Biol. Chem. 278, 48872– 48879
62. Takada, Y., and Aggarwal, B. B. (2003) J. Immunol. 171, 3278 –3286 63. Shishodia, S., Majumdar, S., Banerjee, S., and Aggarwal, B. B. (2003) Cancer
Res. 63, 4375– 4383
64. Arsura, M., Panta, G. R., Bilyeu, J. D., Cavin, L. G., Sovak, M. A., Oliver, A. A., Factor, V., Heuchel, R., Mercurio, F., Thorgeirsson, S. S., and Sonenshein, G. E. (2003) Oncogene 22, 412– 425
65. Cavin, L. G., Romieu-Mourez, R., Panta, G. R., Sun, J., Factor, V. M., Thor-geirsson, S. S., Sonenshein, G. E., and Arsura, M. (2003) Hepatology 38, 1540 –1551
66. Liu, P., Kimmoun, E., Legrand, A., Sauvanet, A., Degott, C., Lardeux, B., and Bernuau, D. (2002) J. Hepatol. 37, 63–71
67. Chiao, P. J., Na, R., Niu, J., Sclabas, G. M., Dong, Q., and Curley, S. A. (2002) Cancer 95, 1696 –1705
68. Matskevich, A. A., and Strayer, D. S. (2003) Gene Ther. 10, 1861–1873 69. Gupta, R. A., and Dubois, R. N. (2001) Nat. Rev. Cancer 1, 11–21
70. Mao, J. T., Roth, M. D., Serio, K. J., Baratelli, F., Zhu, L., Holmes, E. C., Strieter, R. M., and Dubinett, S. M. (2003) Clin. Cancer Res. 9, 5835–5841 71. Kern, M. A., Schubert, D., Sahi, D., Schoneweiss, M. M., Moll, I., Haugg, A. M., Dienes, H. P., Breuhahn, K., and Schirmacher, P. (2002) Hepatology 36, 885– 894
72. Cheng, J., Imanishi, H., Amuro, Y., and Hada, T. (2002) Int. J. Cancer 99, 755–761
73. Koga, H. (2003) Cancer 98, 661– 667
74. Yamamoto, H., Kondo, M., Nakamori, S., Nagano, H., Wakasa, K., Sugita, Y., Chang-De, J., Kobayashi, S., Damdinsuren, B., Dono, K., Umeshita, K., Sekimoto, M., Sakon, M., Matsuura, N., and Monden, M. (2003) Gastroen-terology 125, 556 – 671
at National Taiwan University on June 18, 2009
www.jbc.org
