GADD153-mediated anticancer effects of N-(4-hydroxyphenyl)retinamide on human hepatoma cells.

The anticancer effects of N-(4-hydroxyphenyl)retinamide (4HPR), a potential chemopreventive or chemotherapeutic retinamide, are thought to be derived from its ability to induce apoptosis. However, the mechanism of apoptosis induced by 4HPR remains unclear. Thus, this study was designed to identify the gene(s) responsible for induction of apoptosis by 4HPR. Apoptosis was effectively induced by 4HPR in human hepatoma cells. Using the differential display-PCR method, a gene involved in the response to 4HPR was identified, and cells in which the expression of that gene was modulated were analyzed for survival, induction of apoptosis, and cell cycle. GADD153, a gene involved in growth arrest and apoptosis, was preferentially expressed in human hepatoma cells as well as in other cancer cells during 4HPR-induced apoptosis. 4HPR regulates GADD153 expression at the post-transcriptional level in Hep 3B cells and at the transcriptional and post-transcriptional levels in SK-HEP-1 cells, when assayed by in vitro transfection and mRNA stability experiments. To determine the role of the GADD153 protein overexpression that is induced by 4HPR, Hep 3B cells with ectopic overexpression of GADD153 were found to be growth-arrested (at G(1)) and readily underwent apoptosis following treatment with 4HPR or even when they reached confluence. N-Acetyl-l-cysteine or GADD153 antisense significantly protected the cells from 4HPR-induced apoptosis, accompanying by the inhibition of GADD153 overexpression. Parthenolide-mediated overexpression of GADD153 resulted in enhanced 4HPR-induced apoptosis. These results suggest that GADD153 overexpression induced by 4HPR may contribute to the anticancer effects (induction of apoptosis and growth arrest) of 4HPR on cancer cells.

GADD153, a member of the CCAAT/enhancer-binding protein family of transcription factors, is transcriptionally activated and is highly expressed following treatment of cells with a variety of growth arrest and/or DNA-damaging factors (1,2), such as calcium ionophore (3), glucose deprivation (4), oxidative stress (5), reductive stress (6), endoplasmic reticulum stress (7), or activation of acute phase response (8). GADD153 has been implicated in the commitment to growth arrest or cell death. Microinjection of GADD153 induces 3T3 cells to arrest at the G 1 /S boundary (9), while ectopic expression of GADD153 causes M1 myeloblast leukemia cells to undergo apoptosis (10). Chemotherapeutic drugs, including paclitaxel, cisplatin, and etoposide, were reported to induce GADD153 overexpression, which was associated with cellular injury/apoptosis (11)(12)(13). More recently, GADD153 was reported to sensitize cells to endoplasmic reticulum stress through mechanisms that involve down-regulation of Bcl-2 and enhanced oxidant injury (14). However, studies to date are not sufficient to explain how the anticancer effects of GADD153 overexpression function in drug-induced apoptosis.
N-(4-Hydroxyphenyl)retinamide (4HPR), 1 a potential chemotherapeutic or chemopreventive retinamide, seems promising in terms of its effectiveness relative to toxicity. 4HPR also induces apoptosis of human neuroblastoma cells (15,16), cervical (17), breast (18), human ovarian carcinoma cells (19), head and neck (20), lung (21), and human malignant hematopoietic cells. 4HPR-sensitive cells include those carrying mutations of the retinoic acid receptor (RAR) and which are thus unresponsive to all-trans-retinoic acid (22). Furthermore, 4HPR was found to activate RARs and several RAR-specific antagonists partially inhibited 4HPR-induced apoptosis (23). Thus, both RAR-independent and -dependent pathways are involved in 4HPR-mediated apoptosis. 4HPR stimulates the generation of intracellular free radicals, which appear to play a causative role in the induction of apoptosis in vitro (24,25). Increased ceramide levels resulting from 4HPR treatment induce cell death in a p53-and caspase-independent manner through mixed apoptosis and necrosis pathways (26). In addition, 4HPR modulates the expression levels of some apoptosisrelated genes, such as p21, c-myc, and c-jun (27), and activates c-Jun N-terminal kinase-mediated apoptotic signaling (28). However, the underlying mechanism(s) by which 4HPR induces apoptosis remains unclear. Quite recently, we observed that 4HPR effectively induces apoptosis in Fas-defective hepatoma cells through caspase-8 activation (29), and 4HPR-induced apoptosis was effectively modulated by alteration of mitochondrial membrane potential (30). Thus, to elucidate the molecular and cellular mechanism(s) by which 4HPR induces apoptosis, we undertook to identify the gene(s) involved in this drug-induced apoptosis. Using the differential-display PCR (DD-PCR) method, we identified genes that were differentially expressed between untreated cells and 4HPR-treated cells. One of them, GADD153, which was preferentially expressed in 4HPR-treated hepatoma cells, seemed to be responsible for the induction of apoptosis and growth arrest. Thus, in this study, we address the role of GADD153 overexpression in the antitumor effects of 4HPR.

EXPERIMENTAL PROCEDURES
Cell Lines and Reagents-The hepatoma cell lines, including Hep 3B (deletion of p53) and SK-HEP-1 (wild-type p53) (31), and other cancer cell lines were obtained from the American Tissue Culture Collection (ATCC, Manassas, VA). 4HPR was kindly supplied by the Johnson Pharmaceutical Research Institute (Spring House, PA) and was dissolved in absolute ethanol at a concentration of 10 Ϫ2 M and was stored in aliquots at Ϫ20°C for a maximum of 2 weeks. 4Ј,6Ј-diamidino-zphenylindole (DAPI), N-acetyl-L-cysteine (NAC), and propidium iodide (PI) were purchased from Sigma. Parthenolide was obtained from Calbiochem (San Diego, CA).
Cell Culture and Detection of Apoptotic Cells-Cells were cultured in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum in air containing 5% CO 2 . Viable cells were determined by trypan blue dye exclusion. The percentage of apoptotic cells was evaluated by DAPI staining. Briefly, Hep 3B and SK-HEP-1 cells were treated with the indicated concentrations of 4HPR, or with equitoxic concentrations of 10 M 4HPR or 15 M 4HPR, which caused about 40% cell death, respectively, for 3 days in 6-well dishes. Cells were fixed in ice-cold methanol for 10 min and were then stained with the DAPI reagent (1 g/ml). Fluorescence microscopy was performed using a Nikon diaphot microscope. Apoptotic cells were recognized as condensed, fragmented, degraded nuclei, or ghosts that stained only faintly. At least 200 cells were counted for each time point, and all counting was done in a blinded fashion.
Detection of DNA Fragmentation-Cells (1 ϫ 10 6 ) were seeded in 6-cm Petri dishes and allowed to settle and attach. Cells were treated with 10 M 4HPR for 48 h or 72 h. For analysis of genomic DNA, cells were harvested and combined with nonattached cells in the supernatant. Cells were resuspended in 0.5 ml of lysis buffer (50 mM Tris-HCl, 100 mM EDTA, 0.5% SDS, pH 8.0) containing 0.1 mg/ml RNase A. After incubation at 37°C for 30 min, extracts were treated with 1 mg/ml proteinase K for an additional 16 h at 37°C. DNA was extracted with phenol/chloroform, then with chloroform, and was finally precipitated with ethanol and sodium acetate. 20 l of each extract (dissolved in 50 l of H 2 O) was then loaded on a 1.5% agarose gel and separated in the presence of 0.5 g/ml ethidium bromide.
Flow Cytometric Analysis of Apoptotic Quantitation and Cell Cycle-Measurements of apoptotic cells and cell cycle distribution were performed using a modification of the technique described previously (32). Apoptotic cells were quantified by staining with FITC-conjugated Annexin V (Clontech, Palo Alto, CA). Cells (1 ϫ 10 6 ) were collected at 72 h for flow cytometric (FCM) measurement and stained with FITC-conjugated annexin V and PI as instructed by the manufacturer, and then analyzed by flow cytometry using a FACScan (BD Biosciences, San Jose, CA) with an argon laser set to excite at 488 nm. In addition, cells were harvested at each time point and were fixed in 70% ethanol and stored at 4°C. Cell cycle distribution was performed at the indicated intervals following 4HPR treatment. PI (40 g/100 l PBS) was added to 1 ϫ10 6 cells suspended in 800 l of PBS together with 100 l of RNase A (1 g/ml), and was incubated at 37°C for 30 min before FCM analysis of 2 ϫ 10 4 cells. Red fluorescence due to propidium-bound DNA was measured using a 630 nm-long bandpass filter. Data were analyzed as single-parameter frequency histograms in an SFIT model.
DD-PCR-Modified DD-PCR was performed using a GeneHunter RNA image kit (GeneHunter Corp, Nashville, TN), based on an improved method described previously (33). 4 g of total RNAs from treated or untreated cells were reverse-transcribed with 200 units of SuperScript II RT enzyme (Invitrogen) in the presence of 1 mol/liter 1-base anchored oligo(dT) primers for 1 h at 42°C in a total volume of 10 l. The reaction was terminated by incubation at 75°C for 10 min. 2 l of the reaction mixture was PCR-amplified with Dynazyme (Finnzyme OY, Epsoo, Finland) in 1 mol/liter H-AP, a 13-mer (5Ј-end primers) and oligo(dT) 15 primers (3Ј-end primers), using a GeneAmp PCR system 9600 (PerkinElmer, Norwalk, CT). PCR reactions and reamplification of cDNAs of interest were carried out as previously described (34). After cloning into the pGEM-T vector using the TA cloning system (Promega, Madison, WI), the sequenced cDNAs were analyzed via the BLAST program for matches in the GenBank TM data base and were compared with each other via FASTA analysis.
Northern Blot Analysis-Cells were cultured in Dulbecco's modified Eagle's medium with 10% fetal bovine serum until they reached 60% confluence and were then treated with 10 M or 15 M 4HPR for times as indicated in the text. Total RNAs were extracted from treated or untreated cells using phenol and guanidine thiocyanate solution (Tri Reagent; Molecular Research Center, Inc. Cincinnati, OH). RNAs were then fractionated by electrophoresis on 1.0% agarose gels containing formaldehyde and were transferred to membranes. Blots were hybridized overnight in 2 ϫ 10 6 cpm/ml cDNA probe, labeled with [P 32 ]dCTP (PerkinElmer Life Sciences) by random-priming, washed, then exposed to X-Omat AR film (Kodak) at Ϫ70°C, as described previously (35). As a loading control, glyceraldehyde-3-phosphate dehydrogenase (GAPDH) mRNA was detected with a GAPDH cDNA probe.
Luciferase Assay-pGADD-LUC, a hamster GADD153 promoterdriven luciferase reporter construct was a gift from Dr. Nikki J. Holbrook (National Institute on Aging, Baltimore, MD) (36). The cells were transfected with the pGADD-LUC construct using Lipofectin (Invitrogen). Cells were plated at 2 ؋ 10 4 cells per well in 24-well plates, and then 18 h later the cells were incubated at 37°C for 16 h with 500 ng of pGADD153-LUC plasmid and 50 ng of pRL-TK plasmid (Promega, Madison, WI) as well as Lipofectin. Following the transfection, the cells were replenished with complete medium and were treated with equitoxic levels of 4HPR. The cells were lysed in 120 l of lysis buffer at the indicated time intervals and were stored at Ϫ20°C until assayed. Luciferase activity was measured using the Dual-Luciferase Reporter Assay System (Promega) as per the manufacturer's instructions and was normalized by Renilla luciferase activity.
Transfections-Transfection of the GADD153 gene into Hep 3B cells was performed using an expression plasmid vector encoding human GADD153 cDNA or control pcDNA3. The construct of the GADD153 expression vector was made by ligating human GADD153 (a gift from Dr. Nikki J. Holbrook) with pcDNA3 in each BamHI/XhoI site in the sense orientation. For the generation of the antisense GADD153 expression vector, human GADD153 was PCR-amplified with the forward primer containing the XbaI restriction enzyme site (5Ј-GCTCTA-GAGGGCTGCAGAGATGGC-3Ј) and the reverse primer containing the EcoRI restriction enzyme site (5Ј-GGAATTCGGGACTGATGCTCCCA-3Ј). The PCR-amplified GADD153 double-strand DNA was ligated to pcDNA3 in XbaI/EcoRI site in the antisense orientation. Transfections were carried out by Lipofectin (Invitrogen) according to the manufacturer's protocol. Sequence-verified GADD153-transfected and neotransfected cells were selected in the presence of 600 g/ml G418 for 2-3 weeks. Finally, individual colonies were isolated using cloning rings, expanded, and assayed for expression of the transfected gene by Northern analysis and Western analysis.
Western Immunoblotting-Cells were washed twice with cold PBS on ice and were harvested by scraping with a rubber policeman. Cells were sedimented by centrifugation at 4°C and were resuspended directly into Laemmli sample buffer containing 62.5 mM Tris-HCl, pH 6.8, 2% SDS (w/v), 12% glycerol (w/v), and 5% ␤Ϫmercaptoethanol (v/v). Extracted proteins were resolved by 13% SDS-PAGE and were transferred to nitrocellulose membranes. Membranes were incubated overnight at 4°C in the primary antibody and were then incubated for 45 min in the secondary antibody. Following incubation in the secondary antibody, blots were washed three times with PBS/0.1% Tween, then developed using a commercial chemiluminescence detection kit (Amersham Biosciences). GADD153 polyclonal antibody (F168) was purchased from Santa Cruz Biotechnology (Santa Cruz, CA).
Statistical Analysis-All data were entered into Microsoft Excel 5.0, and GraphPad Software was used to perform two-tailed Student's t tests. All p values of less than 0.05 were considered to be statistically significant.

4HPR-induced Apoptosis in Hepatoma
Cells-We observed that 4HPR effectively induced cell death in a concentration-dependent manner after 3 days in culture. Hep 3B cells, which have a defect in p53 function, rounded up, shrank, and detached during 3 days of exposure to 10 M 4HPR. In contrast, SK-HEP-1 cells, which have normal p53 function, were less susceptible to 4HPR-induced cell death. The IC 50 for Hep 3B and for SK-HEP-1 cells was 12.5 and 17.5 M, respectively (Fig.  1A). Thus, we used equitoxic concentrations, 10 and 15 M, that caused 40% cell death of Hep 3B and SK-HEP-1 cells, respec-tively. Cells exposed to those equitoxic concentrations of 4HPR for 3 days were stained with Annexin V and PI to prove apoptotic cell death. 4HPR effectively induced apoptosis in up to 40% of those hepatoma cells, as shown by their positive staining for Annexin V as assessed by FCM (Fig. 1B). The percent of cells in the sub-G 1 fraction that were stained with PI showed similar results (data not shown). In support of those findings, staining with the DNA-binding dye DAPI revealed condensed chromatin and fragmented nuclear morphologies characteristic of apoptosis in Hep 3B and SK-HEP-1 hepatoma cells. The percentage of apoptotic cells measured using DAPI was the same as the apoptotic population measured by FCM in each cell line (data not shown).
Induction of Endogenous GADD153 by 4HPR-We identified the GADD153 gene, which is preferentially expressed in apoptotic Hep 3B cells, by the DD-PCR method (Fig. 1C) and confirmed that equitoxic levels of 4HPR caused the prominent induction of GADD153 mRNA in Hep 3B and SK-HEP1 cells in a time-dependent manner (Fig. 1D, upper panels). Next we determined whether the GADD153 protein expression correlates with GADD153 mRNA induction during 4HPR-induced apoptosis. 4HPR induced the expression of GADD 153 protein, which relates to GADD153 mRNA expression. (Fig. 1D, lower panels). In addition, to check whether GADD153 is also induced in other types of cancer cells by 4HPR, we choose colon (HCT116 and HT29), lung (A549 and NCI-H522), and breast (MCF7 and MDA-MB-231) cancer cell lines, according to p53 function (wild-type and mutant type, respectively), and kidney cancer cells with wild-type p53 (A498 and ACHN) (37). Each cell line was treated with 10 M 4HPR for 48 h. The apoptotic cell death was counted by staining with DAPI and ranged from 21 to 51% ( Fig. 2A) while less than 7% of untreated control cells were apoptotic. 4HPR also induced GADD153 overexpression in all cell lines, independent of their functional p53 status (Fig.  2B). Therefore, GADD153 overexpression by 4HPR seems to be general, not cell specific.
4HPR Stabilizes GADD153 mRNA-To determine the effect of 4HPR on activation of the GADD153 promoter, Hep 3B and SK-HEP-1 cells were transiently transfected with pGADD-LUC, which contains the hamster GADD153 promoter coupled to the luciferase reporter gene (11). Those transfected cells were exposed to 10 or 15 M 4HPR, respectively, for 48 h (Fig.  3A). The lack of change in luciferase activity (relative to the level in untreated control cells) indicated that 4HPR did not increase GADD153 promoter activation in transfected Hep 3B cells. In contrast, in SK-HEP-1 cells, 4HPR increased GADD153 promoter activation in a time-dependent manner that corresponded well with its effect on the level of endogenous GADD153 mRNA (over 4-fold induction of GADD153 mRNA after 48 h treatment with 4HPR). These results suggest that 4HPR increases GADD153 mRNA levels at the post-transcriptional level in Hep 3B cells, whereas it occurs at the transcriptional level in SK-HEP-1 cells. To further investigate the post-transcriptional regulation by 4HPR, we investigated whether 4HPR might affect the stability of GADD153 mRNA. The decline of GADD153 mRNA levels in 4HPR-treated cells was examined following 4HPR withdrawal and/or the addition of a inhibitor of transcription, actinomycin D. Subconfluent The arrows indicate bands of GADD153 cDNA specifically expressed in 4HPR-treated cells. D, total RNAs from Hep 3B and SK-HEP-1 cells treated with 10 and 15 M 4HPR, respectively, were fractionated by electrophoresis on 1.0% agarose gels containing formaldehyde and were then transferred to membranes. Blots were hybridized overnight with 2 ϫ 10 6 cpm/ml GADD153 cDNA probe labeled with [P 32 ]dCTP by randompriming, washed, then exposed to X-Omat AR film (Kodak) at Ϫ70°C. The blots were stripped and sequentially hybridized with a probe for GAPDH cDNA as a loading control (upper panels). For immunoblotting of GADD153, 30 g of extracted proteins were resolved by 13% SDS-PAGE and were transferred to the membrane. The blot was probed with a polyclonal antibody to GADD153 (F168) and was then stripped and reprobed with a monoclonal antibody to actin as a loading control (lower panels).
Hep 3B cells or SK-HEP-1 cells were treated with 10 or 15 M 4HPR, respectively, for 48 h to bring about the accumulation of GADD153 mRNA levels. Cells treated with 4HPR were then fed with 4HPR-free medium and to which 4HPR alone, actinomycin D (Act D, 5 g/ml)ϩ4HPR, or Act Dϩvehicle was added, respectively (13). This time is considered 0 h. RNA was extracted at subsequent time points from each group, and relative mRNA levels were calculated by comparison of GAPDH-normalized values with the level observed in cells at time 0 (Fig.  3B). 4HPR sustained the increase of GADD153 mRNA following treatment for 12 h in both Hep 3B cells and SK-HEP-1 cells.
In the presence of actinomycin D, 4HPR did not change the GADD153 mRNA level in Hep 3B cells, whereas 4HPR slowly decreased the GADD153 mRNA level in SK-HEP-1 cells. The half-life (t1 ⁄2 ) of GADD153 mRNA is ϳ5 h. The treatment with vehicle progressively decreased the GADD153 mRNA level in both cell lines. However, the rate of decrease of GADD153 mRNA was higher in SK-HEP-1 cells (t1 ⁄2 ϳ1 h) than in Hep 3B cells (t1 ⁄2 ϳ2 h). These results imply that 4HPR regulates GADD153 mRNA levels post-transcriptionally in Hep 3B cells and at the transcriptional level in SK-HEP 1 cells, in accordance with results from the in vitro transfection experiment. In contrast, in the presence of actinomycin D, the rate of decrease of GADD153 mRNA was higher during vehicle treatment than during 4HPR treatment in SK-HEP-1 cells, suggesting that some post-transcriptional regulation of the GADD153 mRNA concurrently exists. Thus, 4HPR may regulate GADD153 mRNA expression post-transcriptionally and/or transcriptionally in a cell specific manner.
Effect of Ectopic GADD153 Overexpression on Apoptosis-To determine the role of GADD153 protein overexpression, we introduced the GADD153 expression system into Hep 3B cells because these cells expressed less constitutive GADD153. Two types of Hep 3B cells (3B-G1 and 3B-G5), which stably expressed different levels of human GADD153 protein, were isolated (Fig. 4A) and then treated with 10 M 4HPR. In the cells transfected with GADD153, the level of expression was compared with that observed in 4HPR-treated cells. Northern blot analysis revealed that the expression of GADD153 mRNA in the transfectants was lower than that seen in the treated cells, thus it does not seem to be sufficient to induce apoptosis on its own. Compared with control cells transfected with the empty vector alone (3B-C1 and 3B-C2), the 3B-G1 and the 3B-G5 cells became readily apoptotic within 48 h of treatment (p Ͻ 0.01). 3B-G1 cells, which express less ectopic GADD153, were up to respectively, for 48 h, and were then fed with 4HPR-free medium to which 4HPR alone, actinomycin D (Act D, 5 g/ml)ϩ4HPR, or Act Dϩvehicle was added, respectively. Total RNA was extracted from the three sets at subsequent time point (upper). Autoradiograms were read using densitometric scanning and were normalized against GAPDH mRNA levels (lower). 70% apoptotic while 3B-G5 cells, which express more ectopic GADD153, were up to 80% apoptotic (Fig. 4B). The proportion of apoptotic cells seems to be related to the expression level of the GADD153 protein. Furthermore, both types of cells were apoptotic (27.0 or 52.5%, respectively, p Ͻ 0.01) at confluence after 72 h of culture even without any drug treatment (Fig. 4C), suggesting that confluence or nutrient depletion also sensitizes both types of cells to apoptosis. This was confirmed since serum deprivation resulted in similar proportions of apoptotic cells, which displayed disintegrated nuclei and nonrandom DNA fragmentation, as assessed by agarose gel electrophoresis of genomic DNA (Fig. 4C, lower left). Similarly, the apoptotic sub-G 1 fraction of the cells was increased by confluence using FCM analysis (Fig. 4C, lower right). These results suggest that apoptotic cell execution induced by 4HPR or by confluence/ nutrient depletion is higher in GADD153 transfectants than in vector control cells.
Effect of Ectopic GADD153 Overexpression on the Cell Cycle-Hep 3B cells that stably overexpress GADD153 protein showed lower cell numbers than the vector control cells during culture. Thus, we investigated whether GADD153 overexpression is associated with inhibition of cell growth. Thymidine uptake assays for DNA synthesis were significantly inhibited i.e. 70 and 53% of the control in 3B-G1 and 3B-G5 cells, respectively (p Ͻ 0.01) (Fig. 5A). We next analyzed the impact of GADD153 overexpression on the cell cycle of Hep 3B cells. In asynchronized cells, GADD153 overexpression increased the G 1 fraction of cells from 45.9 to 59.7%, compared with vector control cells after 12 h in culture (Fig. 5B). Thus, we further studied the cell cycle inhibition in synchronized cells. For synchronization, Hep 3B cells were arrested in M phase with 0.5 g/ml nocodazole. Cells were then analyzed at 3-h intervals following release from the M block. Cells that stably overexpressed GADD153 clearly showed delayed cell cycling and G 1 arrest (disappearance of G 2 /M), and thus, GADD153 overexpression seems to be responsible for the G 1 growth arrest.
Inhibition of GADD153 Overexpression Relates to Inhibition of Apoptosis-We examined whether prevention of GADD153 overexpression might lead to the inhibition of 4HPR-induced apoptosis. The generation of reactive oxygen species (ROS) is responsible for 4HPR-induced apoptosis in some tumor cells (24,25). Accordingly, we observed that an antioxidant NAC effectively inhibited 4HPR-induced apoptosis in hepatoma cells, although not completely (29). NAC (2 mM) inhibited 4HPR-induced apoptosis by about 41% (23.5 versus 39.7%) (p Ͻ 0.05) (Fig. 6A). Thus we examined whether this inhibition of apoptosis by NAC correlates with the inhibition of GADD153 overexpression. Interestingly, NAC inhibited the overexpression of GADD153 that is induced by 4HPR, which is proportional to inhibition of apoptosis (Fig. 6B). For another approach to examine the down-regulation of expression of GADD153, we used transfection of an antisense expression plasmid of GADD153. During 4HPR-induced apoptosis, the GADD153 protein overexpression was inhibited in Hep3B cells transfected with the antisense GADD153 cDNA (3B-AS3 and 3B-AS14) (Fig. 6C). 3B-AS3 and 3B-AS14 cells were significantly resistant to 4HPR, e.g. about 27 and 22.2% of transfectants were apoptotic (p Ͻ 0.05 and p Ͻ 0.01, respectively), compared with 44.3 and 39.3% of vector control cells (3B-C3 and 3B-C4) (Fig. 6D). These results suggest that the inhibition of GADD153 expression correlates positively with cell survival in hepatoma cells.
Parthenolide-mediated GADD153 Overexpression Enhances 4HPR-induced Apoptosis-Recently, we observed that an NF-B inhibitor, parthenolide, enhanced the GADD153 mRNA overexpression induced by 4HPR as described previously (39). Therefore, we determined the role of parthenolide on the apoptotic cell death induced by 4HPR. Non-cytotoxic dose of parthenolide (4 M) effectively enhanced 4HPR-induced GADD153 overexpression by about 2-fold in Hep 3B and in SK-HEP-1 cells (Fig. 7A), which in turn enhanced 4HPR-induced apoptosis by about 2-fold in Hep 3B and in SK-HEP-1 cells (55.5 and 53.0%, respectively) (p Ͻ 0.01), compared with cells treated with 4HPR alone (25.5 and 29.0%, respectively) (Fig. 7B). Similarly, FCM analysis revealed that parthenolide enhanced 4HPR-induced apoptosis by 2-fold (p Ͻ 0.01) (Fig. 7C). The degree of induction of GADD153 seemed to correlate with the increase of apoptotic cell death. These results suggest that GADD153 overexpression modulates the cell execution response to 4HPR in cancer cells. DISCUSSION Previously we observed that 4HPR effectively induced apoptosis and/or inhibited cell growth (G 1 arrest) in hepatoma cells in a cell-specific manner (30). However, little is known about the molecular and cellular mechanism(s) by which treatment with 4HPR has these effects. Thus, the aims of this study were to identify the gene(s) responsible for 4HPR-induced growth inhibition and apoptosis and to clarify its functional role(s). Human hepatomas are resistant to chemotherapy and radiation and are often associated with mutations of p53 (40). p53independent cell growth arrest or apoptosis pathways are known to respond to varying chemotherapeutic drugs and stress in tumor cells, and it is also known that 4HPR induces p53-independent apoptosis (41). Using DD-PCR we found that FIG. 5. Inhibition of the cell cycle by ectopic GADD153 overexpression. A, GADD153 overexpressing (3B-G1, 3B-G5) and control (3B-C2) cells were plated in 96-well microtiter plates at a density of 1 ϫ 10 3 cells/well and were incubated for 72 h. 24 h prior to harvest, 1 Ci of [ 3 H]thymidine was added to each well, and the incubation was continued for an additional 24 h. Cellular uptake of [ 3 H]thymidine was determined by liquid scintillation counting. Assays were performed in triplicate, and the experiment was repeated at least two times with similar results. Vertical bars represent the means Ϯ S.E. of three experiments performed in duplicate (**, significantly different from vector control at p Ͻ 0.01). B, in asynchronized cells, the cell cycle distribution of the cells was performed 12 h after seeding using a FACScan (as described under "Experimental Procedures"). For synchronization, 3B-G5 and 3B-C2 cells were arrested in M phase by treatment with 0.5 g/ml nocodazole. Cells were analyzed at 3-h intervals following release from the M block. Data were analyzed as single-parameter frequency histograms in the SFIT model. M 1 represents a fraction of apoptotic sub-G 1 (Asyn, asynchronized; Syn, synchronized).
the GADD153 gene was preferentially expressed during 4HPRinduced apoptosis in hepatoma cells. The induction of the GADD153 gene by anticancer agents has been reported to occur in a p53-independent fashion (42). We first found that GADD153 overexpression was associated with the 4HPR-me-diated apoptosis in hepatoma cells as well as in other types of cancer cells, regardless of the state of their p53 function. Furthermore, there was a good correlation between increases in GADD153 mRNA levels and the chemo-response rates of cisplatin in head and neck cancers (43). Introduction of the GADD153 gene into gastric cancer cells increased their sensitivity to anticancer drugs (44). Thus, the sum of these results suggests that the GADD153 gene may be an important target gene for apoptotic cell death induced by anticancer drugs.
The ability of 4HPR to activate the transfected GADD153 promoter is barely detectable in Hep 3B cells although it stabilized the GADD153 mRNA. Therefore, post-transcriptional stabilization may play a major role in increasing GADD153 mRNA levels in response to this drug. However, in addition to the increase in GADD153 mRNA stability induced by 4HPR, the magnitude of the activation of the GADD153 promoter substantially correlated with changes elicited in endogenous mRNA levels in response to treatment with 4HPR in SK-HEP-1 cells. Thus, 4HPR transcriptionally and post-transcriptionally regulates GADD153 mRNA levels in SK-HEP-1 cells. The mRNA levels of GADD genes have been shown to be regulated both at the transcriptional and at the post-transcriptional levels. Transcriptional mechanisms accounted for the induction of GADD153 genes in cells treated with DNA-damaging agents. In contrast, signals responsible for withdrawal from an active proliferative state have been shown to stabilize GADD153 transcripts (45,46). Treatment with methyl methanesulfonate or with UV radiation stabilizes GADD153 mRNA in actively proliferating cells, an effect that was not observed in growtharrested cells (47). Etoposide regulates GADD153 mRNA mainly at the transcriptional level (37). In contrast, 4HPR regulates GADD153 mRNA levels at the transcriptional and/or the post-transcriptional levels in a cell-specific manner.
The GADD153-transfected cells had increased sensitivity to 4HPR-mediated cytotoxicity, which was associated with cell death and the formation of internucleosomal DNA ladders. Even confluence/nutrient depletion alone readily triggered apoptosis in those cells without any drug treatment. Other groups had previously reported that induction of GADD153 correlated with the onset of apoptosis (37,48). To date, however, few studies have addressed the mechanistic link between expression of GADD153 and cell death. Recently, overexpression of GADD153 sensitized cells to ER stress through the down-regulation of Bcl-2 expression. This down-regulation of Bcl-2 expression enhanced oxidant injuries, e.g. depletion of cellular glutathione and exaggerated production of ROS (14). Our study previously revealed that 4HPR did not distinctly change Bcl-2 protein levels in SK-HEP-1 cells, and furthermore, that Hep 3B cells were defective in Bcl-2 expression (29). Thus, the down-regulation of Bcl-2 by GADD153 overexpression did not seem to be a causative role in hepatoma cells. Confluent culture or nutrient depletion readily induces apoptosis in cells stably transfected with GADD153. In support of this concept, the overexpression of GADD153 by platelet-derived growth factor-BB in vascular smooth muscle cells was reported to significantly reduce cell viability and to induce apoptosis, when 100% confluency is reached (49). GADD153 was found to be induced during serum starvation or glutamine deprivation (40,50). Thus, cell-cell contact or nutrient depletion may critically trigger apoptosis in cells overexpressing GADD153.
3B-G1 and 3B-G5 cells showed inhibited cell growth, particularly at G 1 arrest, which is in accordance with a previous report wherein microinjection of GADD153 expression plasmids into NIH-3T3 cells blocked the cells from progressing from G 1 to S phase. However, a mechanistic link between the FIG. 6. NAC or antisense GADD153 inhibits 4HPR-induced apoptosis. A, Hep 3B cells were treated with 10 M 4HPR in the presence or absence of 2 mM NAC for 72 h and apoptotic cell death was determined by DAPI staining. Each bar represents the mean Ϯ S.E. of quadruplicate determinations (*, significantly different from treatment with 4HPR alone at p Ͻ 0.05). B, Western blot analysis of GADD153 expression. Cell lysates were prepared from Hep 3B cells treated with 10 M 4HPR in the presence (ϩ) or absence (Ϫ) of 2 mM NAC for 72 h. 30 g of proteins were extracted and resolved by 13% SDS-PAGE and were transferred to the membrane. The blot was probed with a polyclonal antibody to GADD153 (F168) and was then stripped and reprobed with a monoclonal antibody to actin as a loading control. C, GADD153 expression in Hep 3B cells transfected with antisense GADD153 cDNA or empty vector. Cell lysates were prepared from Hep 3B cells treated with 10 M 4HPR for 72 h. 30 g of proteins were extracted and resolved by 13% SDS-PAGE and were transferred to the membrane. The blot was probed with a polyclonal antibody to GADD153 (F168) and was then stripped and reprobed with a monoclonal antibody to actin as a loading control. D, Hep 3B cells transfected with antisense GADD153 cDNA were treated with 10 M 4HPR for 72 h. Apoptotic cell death was determined by DAPI staining. Vertical bars represent the means Ϯ S.E. of quadruplicate experiments (* and **, significantly different from vector controls at p Ͻ 0.05 and p Ͻ 0.01, respectively). overexpression of GADD153 and G 1 arrest remains largely undefined. GADD153 is a transcription factor that plays an important role in regulating the expression of various genes (2). Thus, GADD153 overexpression could activate downstream pathways of signal transduction G 1 arrest in cancer cells. Previously, the tumor suppressor retinoblastoma protein pRb, a central regulator of the G 1 /S phase transition (51), was found to be highly up-regulated in 4HPR-treated breast cancer cells (52), suggesting a link between the anti-proliferative activity of the retinoid and pRb regulation. Recently, the antiproliferative activity of 4HPR was reported to arise from its capacity to maintain pRb in a de-phosphorylated growth-suppressive status in S-G 2 /M, possibly through cyclin D1 mRNA down-regulation and inhibition of pRb-targeting Cdk2 and Cdk4 kinase activities against pRb (53). Thus, it should be further elucidated whether pRb and Cdk regulation by 4HPR is mediated through the overexpression of GADD153.
These free radicals play a key role in the apoptotic response to 4HPR. Antioxidants effectively block the formation of these 4HPR-induced free radicals and markedly suppress the apoptotic effects of the retinamide. In the present study we observed that NAC inhibits the overexpression of GADD153 as much as it inhibits 4HPR-induced apoptosis, imply that ROS generation is the upstream pathway in the overexpression of GADD153. Similarly, NAC prevents induction of the GADD153 gene and accompanying apoptosis by etoposide (13). In agreement with this, in transfection experiments with an antisense expression plasmid of GADD153, the down-regulation of GADD153 increased resistance to 4HPR-induced apoptosis in cells stably transfected with antisense GADD153.
Recently, NF-B was reported to repress GADD153 activation and parthenolide was shown to sensitize breast cancer cells to ER stress (39). Furthermore, parthenolide and IB␣ increase paclitaxel-induced apoptosis through inhibition of a distinct NF-B regulated cell survival pathway (38). Similarly, we observed that NF-B was activated during 4-HPR-induced apoptosis and that its inhibitor parthenolide enhanced the overexpression of GADD153, which in turn correlated with enhanced 4HPR-induced apoptosis. However, parthenolide can have other biological functions beyond NF-B inhibition because it is a small chemical molecule. Thus, it should be further elucidated that NF-B activation suppresses 4HPR-induced apoptosis through inhibition of GADD153 expression. Thus, these results suggest that modulation of GADD153 expression can affect 4HPR-induced apoptosis.
In summary, it is not currently known how overexpression of the GADD153 gene connects to the downstream signal transduction pathway, which leads to apoptosis and cell growth arrest. However, our data reveal that the higher expression of GADD153 correlates with the greater induction of apoptosis and cell growth inhibition in 4HPR-treated cells and that modulation of GADD153 expression can change the sensitivity of cancer cells to 4HPR. Thus, oxidative stress-mediated GADD153 overexpression appears to contribute to the anticancer effects of 4HPR, and these anticancer effects may be potentiated by parthenolide through the enhancement of GADD153 overexpression. Thus, further elucidation of the downstream signaling pathway(s) involved in 4HPR-induced apoptosis and cell growth arrest will advance chemotherapeutic or chemopreventive strategies that use 4HPR to treat various cancer diseases.