Upregulation of GADD153 expression in the apoptotic signaling of N-(4-hydroxyphenyl)retinamide (4HPR)


  • Yuhe Xia,

    1. Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, The University of Hong Kong, Hong Kong, People's Republic of China
    Search for more papers by this author
  • Nai-Sum Wong,

    Corresponding author
    1. Department of Biochemistry, Faculty of Medicine, The University of Hong Kong, Hong Kong, People's Republic of China
    • Department of Biochemistry, Faculty of Medicine, The University of Hong Kong, 3/F, Laboratory Block, Faculty of Medicine Building, 21, Sassoon Road, Hong Kong
    Search for more papers by this author
    • Fax: +852-2855-1254

  • Wang-Fun Fong,

    1. Department of Biology and Chemistry, The City University of Hong Kong, Hong Kong, People's Republic of China
    Search for more papers by this author
  • Henk Tideman

    1. Department of Oral and Maxillofacial Surgery, Faculty of Dentistry, The University of Hong Kong, Hong Kong, People's Republic of China
    Search for more papers by this author


The molecular basis for the pharmacologic effects of N-(4-hydroxyphenyl)retinamide (4HPR) was investigated by studying the gene(s) that this compound may upregulate in cultured human epithelial tumor cells. Treatment of the cultured human nasopharyngeal carcinoma-derived cells (CNE3) with 4HPR caused modest cell-cycle arrest at G1 and apoptosis. The mRNA levels of a total of 20 genes were downregulated with the majority of them involved in cell cycle-related functions. Only the mRNA level of the growth arrest and DNA-damage inducible gene (gadd153) was upregulated by approximately 7-fold, with a concomitant increase in intracellular protein level. Similar upregulation of gadd153 by 4HPR was observed in HeLa and 2 other tumor cell lines. The 4HPR-induced apoptosis was markedly enhanced in the CNE3 cells that transiently overexpressed the gadd153 protein. Unlike 4HPR, all-trans-retinoic acid (ATRA) had no effect on the mRNA or protein level of gadd153. The ability of 4HPR and ATRA to stimulate the promoter activity of gadd153 was then examined. In the HeLa cells, both 4HPR and ATRA caused a 2- to 4-fold stimulation of the promoter activity of gadd153, but similar to the CNE3 cells, ATRA was incapable of upregulating the protein level of gadd153. This is the first demonstration that gadd153 is a 4HPR-responsive gene in tumor cells and may have a functional role to play in 4HPR-induced apoptosis. Furthermore, our data suggest that the expression of gadd153 can be regulated by 4HPR at the transcriptional level. © 2002 Wiley-Liss, Inc.

N-(4-hydroxyphenyl)retinamide (4HPR) is a retinoid compound that is able to inhibit cell proliferation and to induce cell death in a variety of cancer cell lines, such as those derived from breast,1 prostate cancers,2 leukemia3 and head and neck cancers.4 4HPR has low toxicity compared to natural retinoids, and recent clinical trials demonstrate that it may have a preventive action against the occurrence of a second breast cancer.5 Although the anticancer activities of 4HPR have been attributed to the induction of cell death, the intracellular pathways that are involved are far from being understood. The activation of retinoic acid receptors (RARs), thereby mediating the effects of 4HPR, has been previously reported.6, 7 However, a retinoid receptor-independent mechanism cannot be excluded since the induction of apoptosis by 4HPR even in all-trans-retinoic acid (ATRA)-resistant cells has been observed.8, 9 Free radicals are also known to be generated in response to treatment with 4HPR in several cellular systems,10 possibly due to an elevated production of ceramide11 or from the respiratory activity of the mitochondria.10, 12 Apart from being involved in caspase-dependent apoptosis,3, 10 the production of ROS is also implicated in caspase-independent11 cell death, culminating in the cytotoxicity of 4HPR.

4HPR is also known to cause changes in the expression of genes in the target cells. Thus in human prostate cancer cell lines, the expression levels of p21, c-myc and c-jun are modulated by 4HPR.2 The identification of genes that might be modulated by 4HPR and the mechanistic pathways by which such modulations occur represents a more definitive way of studying the mode of action of this prospective anticancer agent. In our study, we have adopted the macroarray cDNA-profiling approach to acquire the identities of the set of genes that might be modulated by 4HPR. Using a human epithelial cell line (CNE3) established from a poorly differentiated nasopharyngeal carcinoma,13 we identified more than 20 genes whose expressions were downregulated in the presence of 4HPR, and only one gene, the growth arrest and DNA-damage inducible gene (gadd153) was upregulated. gadd153 encodes a 28 kDa protein that belongs to the C/EBP family of transcription factors. It is induced by a variety of cellular stresses, such as nutrient deprivation, DNA damage and other forms of cellular injuries.14, 15 Our experiments demonstrated that gadd153 was a 4HPR-responsive gene in several human epithelial tumor cell lines, in addition to the CNE3 cells. Significantly, ATRA was unable to upregulate the mRNA and protein levels of gadd153 although it shared with 4HPR the ability to stimulate the gadd153 promoter activity. Our data suggest that 4HPR may regulate the expression of one of its target genes through transcriptional and possibly posttranscriptional means.


Chemicals and reagents

N-(4-hydroxyphenyl)retinamide (Fenretinide, 4HPR) was obtained from Calbiochem (La Jolla, CA). Polyclonal rabbit anti-human gadd153 antibody was from Santa Cruz Biotechnology (Santa Cruz, CA). Polyclonal rabbit antibody against human PARP was from Upstate Biotechnology (Lake Placid, NY). Monoclonal mouse anti-human caspase-3 antibody was from Transduction Laboratories (Lexington, KY). All other reagents are of the highest grade obtained from Sigma Chemical (St. Louis, MO). The gadd153-promoter reporter plasmid was a kind gift of Dr. R. Chang of the Department of Anatomy, the University of Hong Kong, and Dr. P. Fafournoux, of nutrition Cellulaire et Moleculaire, INRA de Theix, France.

Cells and cell culture

The human epithelial tumor cell lines (CNE2 and CNE3) were originally isolated and established from poorly differentiated nasopharyngeal carcinomas13 and were obtained from Prof. W.F. Fong (Department of Biology and Chemistry, the City University of Hong Kong). The AC3 cell line was derived from a head and neck cancer and was kindly provided by Prof. L. Young (University of Birmingham, Birmingham, UK). The CNE3 (and HeLa), CNE2, AC3 cells are cultured, respectively, in minimal essential medium (MEM), RPMI-1640, Dulbecco minimum essential medium (DMEM) supplemented with 10% fetal bovine serum, penicillin (100 unit/ml) and streptomycin (100 μg/l) (Life Technologies, Rockville, MD). The cells were grown in 75 cm2 plastic culture flasks (Corning, Corning, NY) in an incubator at 37°C under a humidified atmosphere of 95% air and 5% CO2.

Fluorescence microscopy

Apoptotic and necrotic cells were detected by dual fluorescence microscopy analysis of samples stained with Acridine Orange and ethidium bromide (4 μg/ml for each dye). Apoptotic cells can be distinguished from necroctic cells as follows. Acridine Orange is a cell-permeable dye that binds DNA, but ethidium bromide is excluded from cells with intact plasma membranes. During apoptosis, the damage of plasma membrane integrity is a very late event, which is different from necrosis. Therefore, all of the viable and early apoptotic cells have a green appearance due to Acridine Orange staining, instead of an orange fluorescence caused by ethidium bromide staining. Cell morphology was examined by fluorescence microscopy using a Zeiss LSM 510 microscope (Carl Zeiss, Jena, Germany). Apoptotic cells were identified as those containing densely fluorescing bodies due to nuclear fragmentation and chromatin condensation.16

DNA fragmentation assay

After treatment with 4HPR, detached and adherent cells were separately collected. Cells were lysed in a hypotonic buffer (10 mM Tris-HCl, pH 7.5–8.0, 1 mM EDTA, 0.2% Triton X-100) at 4°C for 30 min. The lysate was centrifuged at 15,000 g to separate the fragmented (soluble) from the intact (insoluble) form of DNA. Following treatment of the soluble fraction with DNase-free RNase A (300 μg/ml) and proteinase K (200 μg/ml), the DNA was extracted and electrophoresed on 1% agarose gel and then visualized by ethidium bromide staining.

Flow-cytometric analysis of DNA content

This was performed as previously described.17 After treatment with 4HPR, the adherent cells were obtained by trypsinization and washed with ice-cold PBS. These cells were combined with the detached cells, pelleted and fixed with 70% ethanol and stained with a propidium iodide solution (50 μg/ml propidium iodide, 0.1% Triton X-100 and 100 μg/ml RNase A). The samples were analyzed on an EPICS XL flow cytometer (Beckman-Coulter, Fullerton, CA). At least 10,000 events were counted per sample. The results were analyzed using the Coulter Cytological Program that sorted cells into various phases of the cell cycle according to their DNA contents.

Complementary-DNA array screening

32P]-labeled cDNA was synthesized from total RNA extracted from cells using the RNeasy Midi Kit according to manufacturer's instructions (Qiagen, Hilden, Germany). The 32P-cDNA probes were used to hybridize with either the CLONTECH Atlas human cDNA array membrane for cell cycle (cat. no. 7748-1) or for apoptosis (cat. no. 7743-1). A full list of the gene probes contained in these arrays can be found at the following website: http://www.atlas.clontech.com. After hybridization, the arrays were exposed to X-ray films in a Hypercassette™ with Hyperscreen™ intensifying screen (Amersham Pharmacia Biotech, Buckinghamshire, UK) at −70°C. Quantification and analysis of the hybridization signal intensity of each dot doublet on the array images were performed using the AtlasImage software.

RT-PCR for detection of gadd153 MRNA in the CNE3 cells

Reverse transcription of RNA isolated from the 4HPR-treated CNE3 cells (adherent and detached) was performed in a final reaction volume of 20 μl containing total RNA (1.5 μg) in First Strand Buffer with DTT (10 mM), dNTP (0.5 mM), Oligo(dT)12-18 primer (0.5 μg) and 200 units/μl Moloney murine leukemia virus reverse transcriptase (200 units). The reaction was carried out at 37°C for 60 min. It was then terminated by heating at 70°C for 15 min. One microlitre of the reaction mixture was then amplified by PCR using either of the following pairs of primers: 5′-GAAACGGAAACAGAGTGGTCATTCCCC-3′ (sense) and 5′-GTGGGATTGAGGGTCACATCATTGGCA-3′ (antisense), which will produce a 309 bp fragment of the gadd153 gene, or 5′-CTCAGACACCATGGGGAAGGTGA-3′ (sense) and 5′-ATGATCTTGAGGCTGTTGTCATA-3′ (antisense) to produce a 450 bp fragment of the gapdh gene.18 The PCR amplification was conducted in a reaction buffer containing 20 mM Tris-HCl (pH 8.4), 50 mM KCl, 1.5 mM MgCl2, 167 μM dNTPs, 2.5 units of Taq DNA polymerase and 0.1 μM primers. The reactions were performed in a Perkin-Elmer Thermal Cycler 9700 (PE Applied Biosystems, Foster City, CA) using the following program: denaturing for 1 min at 94°C, annealing for 1 min at 60°C and elongating for 1 min at 70°C for a total of 20 cycles; the final extension took place at 72°C for 5 min. Equal volumes of each PCR sample were subjected to electrophoresis on a 1.2% agarose gel, which was then stained with ethidium bromide and photographed under UV illumination.

Western blot assays

Adherent, 4HPR-treated cells were lysed by incubating on ice for 30 min in a buffer containing 9.1 mM Na2HPO4, 1.7 mM NaH2PO4, 150 mM NaCl, pH 7.4, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 100 μg/ml PMSF, 20 μg/ml aprotinin, 5 μg/ml leupeptin, 1 mM sodium orthovanadate. The lysate was cleared by centrifugation and the protein content in the supernatant was determined by Bradford protein assay (BioRad, Hercules, CA). One hundred micrograms of the cell extract was separated on 10% or 12% SDS-PAGE. The proteins were transferred electrophoretically to nitrocellulose membranes. After blocking with a buffer containing 5% nonfat dry milk and 2% BSA in 5 mM Tris-HCl (pH 7.5), 150 mM NaCl, 0.1% Tween 20, the membrane was incubated with the primary antibody for 1 hr at room temperature and then with the second anti-IgG antibody conjugated to horseradish peroxidase. The membrane was developed with the ECL™ (enhanced chemiluminescence) reagent (Amersham Parmacia) and visualized by autoradiography.

Luciferase reporter assays

Cells were seeded in 24-well culture plates 24 hr before transfection with the pGL3-luciferase reporter plasmid containing the gadd153-promoter. The plasmid DNA (0.2 μg) was mixed with a FuGENE-6 (0.6 μl)/serum-free medium (18 μl) mixture and this was then added to 1 well of cells. The cells were incubated for 24 hr after transfection. 4HPR or ATRA was then added to the cells and after further incubation for 24 hr, luciferase activity in the adherent cells was measured in accordance with manufacturer instructions (Promega, Madison, WI). Luciferase activity was expressed as arbitrary fluorescence units per unit amount of cell lysate.

Plasmid and transient transfection

pCMVgadd153, a human gadd153 cDNA clone in pCMV.3 vector, was generously provided by Dr. A.J. Fornace (Gene Response Section, NCI, Bethesda, MD).19 Transient transfection assays were conducted using FuGENE-6 Transfection reagent (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions, using 2 μg of plasmid DNA (pCMVgadd153 or the empty vector pCMVneo). Eight hours after transfection, the cells were cultured in normal complete medium for another 16 hr. The cells were then treated with 4HPR for the designed time intervals before harvest for analysis.


4HPR induces growth arrest and apoptosis in human epithelial tumor cells

To investigate the molecular action of 4HPR, we first examined the biologic effects it exerted on the human epithelial tumor cell line CNE3. In preliminary experiments, treatment of CNE3 cells with 4HPR (2.5–20 μM) for 72 hr resulted in a decrease of cellular proteins in the 4HPR-treated cultures (less than 50% of the control at 10 μM of 4HPR), as measured by the sulforhodamine-based cytotoxicity assay (data not shown).20 One likely explanation is that cell proliferation is inhibited by 4HPR, consequently resulting in reduction of cell numbers. The effects of 4HPR (7.5 μM) on the proliferation of CNE3 cells were therefore studied by cell-cycle analysis. The percentages of cells in the G1, S and G2/M phases at various time intervals (24, 48, 72 hr) after treatment with 4HPR were compared to the untreated cells (control). Arrest at G1 was evident but modest (69.8 ± 3.1 vs. 57.3 ± 0.3 in the control) after 48 hr of incubation. Concomitantly, there was a reduction of cells in the S-phase (8.3 ± 0.8 vs. 13.1 ± 2.4 in the control), as well as in the G2/M-phase (9.1 ± 1.6 vs. 21.3 ± 0.4 in the control). More important, substantial cells with sub-G1 DNA was seen after 72 hr of incubation (34.8 ± 12.0 vs. 4.1 ± 0.5 in the control). Similar changes were seen at 5 μM of 4HPR.

When the CNE3 cells treated with 4HPR (7.5 μM) were examined by Acridine Orange/ethidium bromide dual fluorescent staining, cell shrinkage was observed with chromosomal condensation and nucleus fragmentation (Fig. 1a, right panel). Such changes in cellular ultrastructure were absent in the control cells (Fig. 1a, left panel). The treatment of cells with 4HPR resulted in some detachment of cells from the culture dishes. When DNA was extracted from the detached, 4HPR-treated cells and then analyzed by agarose gel electrophoresis, a DNA-ladder pattern typical of internucleosomal degradation was readily seen. Minimal DNA-ladder pattern was seen in the adherent, 4HPR-tretaed cells (Fig. 1b). To see if the DNA cleavage was associated with the activation of caspase(s), the level of the 32 kDa form of caspase-3 in 4HPR-treated cells (detached and adherent together) was examined by Western blot analysis. There was little change in the level of caspase-3 even after 48 hr of incubation with 4HPR. However, a small reduction of its protein level was observed after 72 hr of incubation. Consistently, not much PARP was degraded in the first 48 hr. But an 86 kDa fragment of PARP was detected at 72 hr, possibly due to the proteolytic activity of active caspase-3 (Fig. 1c). Such biochemical and morphological features suggested that 4HPR induced apoptosis in the CNE3 cells. The induction of apoptosis by 4HPR was also examined in 2 other epithelial tumor cell lines, CNE2 and AC3. As shown in Figure 1d, 4HPR produced a dose-dependent increase in the proportion of cells with sub-G1 DNA content in CNE2 cells. By contrast, AC3 cells were significantly less sensitive to 4HPR, as demonstrated by the relatively small proportion of cells with sub-G1 DNA content.

Figure 1.

Induction of apoptosis by 4HPR in human epithelial tumor cells. (a) After treatment with 4HPR (7.5 μM) for 72 hr, the CNE3 cells were stained with Acridine Orange and ethidium bromide. The stained cells were examined by fluorescence microscopy. The fluorescent bodies in 4HPR-treated cells were due to condensed chromatin. Left panel, control cells; right panel, 4HPR-treated cells. Magnification 1,000×. (b) After incubating CNE3 cells with 4HPR (5 or 7.5 μM) or with an equal volume of vehicle (DMSO) for 60 hr, the detached and adherent cells were collected separately. Cellular DNA was prepared from these cells and subject to agarose gel electrophoresis. Lane 1, control; lane 2, 4HPR 7.5 μM (adherent cells); lane 3, 4HPR 7.5 μM (detached cells); lane 4, 4HPR 5 μM (adherent cells); lane 5, 4HPR 5 μM (detached cells); lane 6, 100 bp DNA marker. (c) The CNE3 cells were incubated with 4HPR (5 μM) for the time periods indicated and then harvested for the detection of Caspase-3 and PARP by Western blot analysis. (d) The production of sub-G1 DNA in the CNE3, CNE2 and AC3 cells after treatment with 4HPR. The cells were treated with or without 4HPR at the indicated concentrations for 72 hr, harvested, stained with propidium iodide and analyzed for changes in DNA content by flow cytometry. Data are plotted as mean ± SD. The results are representative of 3 independent experiments.

Screening of 4HPR-responsive genes

We next examined if there were any changes in the expression of genes in the 4HPR-treated cells. Using the macroarray cDNA screen specific for detecting genes involved in apoptosis and cell-cycle control, we looked for genes whose expressions in the CNE3 cells might be affected by the treatment of 4HPR (5 μM) for 60 hr. Of a total of 225 different genes screened, 20 genes had their mRNA level downregulated over a wide range (1.6–14-fold). Only the mRNA level of gadd153 was upregulated by 7-fold (Table I). Fifteen other genes involved in various intracellular functions did not show any response to 4HPR (Table II). Each of the 20 genes that were downregulated is involved in a particular cellular function, notably in the regulation of G2-M progression or in mitosis. Thus, cyclin-B1, Kinase-Polo and Cdc2 exhibited the greatest reduction in their mRNA levels (10–14-fold). The Cdc25B mRNA level was reduced 6.5-fold and that of the CAS mRNA-level by 9-fold. The rest of the genes affected by 4HPR are variously involved in functions such as nuclear transport, DNA repair, antioxidation and tumor suppression and suffer a rather mild degree of downregulation (1.6–3.3-fold).

Table I. Alteration of Gene Expression in CNE3 Cells Treated with 4HPR
Cellular functionProtein/gene nameFold induced (+)/reduced (−)
  1. CNE3 cells were treated with 4HPR (5 μM) for 60 hr and total RNA was extracted (from the adherent and detached cells) for the synthesis of 32P-labeled cDNA probes. The labeled probes were used to hybridize with the cDNA macroarrays. The autoradiographs were visualized in a phosphoimager and the signals were analysed using the AtlasImage software.

Stress responseGadd153/DDIT3/CHOP+6.9
Regulation of G2-M transitionCDC25B−6.5
CytokinesisCDC10 protein homolog−3.7
Cell adhesion and motility, synaptic plasticityCDK5−2.8
DNA replicationPCNA−2.2
Regulation of G1-S transitionCDK4−1.7
Regulation of G2-M transitionCyclin B1/CCNB1−14.0
Activation of cyclin B1, maturation of centrosome, formation of bipolar spindle, mitotic exitSerine/threonine-protein kinase PLK1/STPK13−12.6
Regulation of G2-M transitionCDC2/CDK1−10.3
Phosphorylation of cdkWee1Hu CDK tyrosine 15-kinase−6.0
Inhibition of cdkCDKN1C/p57-KIP2−4.9
Regulation of metaphase to anaphase transitionCDC27HS protein−2.0
Nuclear transport factorCellular apoptosis susceptibility protein (CAS)−9.0
DNA repairHHR23B−3.3
Apoptosis inhibitorBAG-1−2.7
Detoxification, against oxidative stressGlutathione-S-transferase (GST) homolog−2.7
Tumor suppressorp33ING1−2.6
DNA replicationChromatin assembly factor (CAF) 1 p48 subunit−2.5
Adaptor proteinSH2/SH3 adaptor GRB2−2.0
Transformation proteinTransforming protein rhoA H12 (RHO12)−1.6
Table II. Genes that are not Affected by 4HPR
Gene namePutative cellular function
  1. CNE3 cells were treated with 4HPR (5 μM) for 60 hr and total RNA was extracted (from the adherent and detached cells) for the synthesis of 32P-labeled cDNA probes. The labeled probes were used to hybridize with the cDNA macroarrays. The autoradiographs were visualized in a phosphoimager and the signals were analysed using the AtlasImage software.

CDKN2B/p14-INK4BInhibition of cdk
p35 cyclin-like CAK1-associated proteinRegulation of cdk activity
CDC16HSMetaphase to anaphase transition
CDC37 homologTargeting of cdk4,Raf-1 to Hsp90
CDC6-related proteinRegulation of DNA-replication in relation to check point control
MAPKK5Oxidative stress response
MAPKK1/MKK1ERK activation
c-myc binding protein MM-1Putative tumor suppressor
c-myc purine-binding transcription factor (PUF)Putative suppressor of tumor metastasis
PIG7P53-inducibe protein
BAKProapoptotic Bcl-protein
Caspase-4 precursorApoptosis
GSK3 alphaCellular energy metabolism and apoptosis
Ubiquitin-conjugating enzyme E2 (UBE2A)Intracellular protein degradation
UV excision repair protein (RAD23A)DNA-repair

Characterization of the effect of 4HPR on the expression of gadd153

To study the possible significance of 4HPR-induced gadd153 expression, experiments were performed firstly to characterize the time course as well as the dose-dependent expression of this protein. In untreated CNE3 cells, the level of gadd153 transcript was very low as detected by RT-PCR experiments. However, after 12 hr incubation with 4HPR (5 μM), a substantial increase in the transcript level was seen, which continued to elevate over the next 36 hr of incubation (Fig. 2a). In Western blot analysis, the gadd153 protein became detectable after 24 hr of incubation with 4HPR (5 μM), consistent with an increase in the gadd153 mRNA level (Fig. 2b). The dose-response relationship was then examined and 2 additional cell lines, CNE2 and AC3, were included. The stimulation of gadd153 protein synthesis was detected even at a comparatively low dose of 4HPR (2.5 μM). As the concentration of 4HPR increased in the incubation, there was a marked increase in the level of gadd153 protein (Fig. 2c). The addition of 4HPR to CNE2 cells caused a dose-dependent increase in the protein level of gadd153. However, the level of gadd153 protein in AC3 cells was comparatively weak and was only elevated significantly at the maximum dose of 4HPR tested, consistent with the lower level of apoptosis that was observed in the experiment presented in Figure 1d.

Figure 2.

The upregulation of gadd153 expression by 4HPR. (a) The cells were treated with 4HPR (5 μM) for various periods of time as indicated. The presence of gadd153 mRNA was examined by RT-PCR as described in Material and Methods. gapdh served as internal control. (b) The protein level of gadd153 in the CNE3 cells were analyzed by immunoblotting after exposure to 4HPR (5 μM) for various time intervals. (c) The CNE3, CNE2 and AC3 cells were incubated with different concentrations of 4HPR for 24 hr, and then harvested for the detection of gadd153 expression by Western blot analysis. C, control cells that received an equivalent amount of DMSO instead of 4HPR.

Gadd153 was not induced by all-trans-retinoic acid (ATRA)

To see if the natural retinoid (ATRA) was also able to stimulate the expression of gadd153, ATRA at various concentrations was added to the CNE3 cells. Unlike 4HPR, ATRA even at a concentration of 20 μM was unable to upregulate the protein level of gadd153. Nor did the simultaneous presence of ATRA (5 μM) prevent the gadd153 response due to 4HPR (5 μM), since the 2 retinoids mixed together would still result in the synthesis of the gadd153 protein (Fig. 3a). Consistently, the mRNA level of gadd153 in the ATRA-treated cells was essentially the same as the control, but that in the 4HPR cells was clearly elevated (Fig. 3b)

Figure 3.

The effect of ATRA on the protein and mRNA levels of gadd153 in CNE3 cells. (a) CNE3 cells were incubated with ATRA or 4HPR or a mixture of both retinoids for 24 hr. Cell lysates were then prepared and analysed by Western blotting. The gadd153 protein was not detected in either cell line at any of the ATRA concentrations. Only cells incubated with 4HPR or 4HPR plus ATRA expressed the gadd153 protein. (b) CNE3 cells were treated with 4HPR (10 μM) or ATRA (10 μM) for 24 hr. An equal volume of DMSO was added to the cells in the control. Total RNA was extracted from the cells after incubation, and RT-PCR was performed to detect the presence of gadd153 mRNA.

4HPR and ATRA stimulate the promoter activity of the gadd153 gene

Previous studies have shown that the gadd153 gene possesses an inducible promoter whose activity can be stimulated by a variety of genotoxic cues.21 To see if 4HPR-induced gadd153 expression is due to an increase in promoter activity, a luciferase-reporter plasmid that carries the promoter region of gadd153 (DNA fragment lying within nucleotide positions −951 and +91 of the gadd153 gene)22 was transiently transfected into the CNE3 cells, and the luciferase activity was measured 24 hr later. However, the luciferase activity that could be measured in the cells was too low (in the range of several hundred units) to be interpreted. We therefore performed the experiments in the cell line HeLa. In these cells, an approximately 2- to 3-fold increase in luciferase activity was observed for 4HPR and rather unexpectedly for ATRA, suggesting that both these retinoids can stimulate the promoter activity of gadd153 (Fig. 4a). On the other hand, when the change in the protein level of gadd153 was examined in the HeLa cells, only 4HPR but not ATRA (up to 20 μM) was found to produce a detectable albeit weak increase (Fig. 4b).

Figure 4.

The effect of 4HPR and ATRA on the expression of gadd153 in HeLa cells. (a) HeLa cells grown in 24-well plates were transfected with the luciferase reporter plasmid that was driven by the gadd153 promoter. Twenty-four hours after transfection, 4HPR or ATRA at the indicated concentration was added to the cells, which were further incubated for 24 hr. The cells were lysed after incubation and 20 μl of the lysate was taken for the measurement of luciferase activity. Triplicate measurement was made for each concentration of the retinoids. Results were calculated in mean ± SD. *p < 0.005. The results are representative of 2 experiments. (b) HeLa cells were incubated with either 4HPR or ATRA for 24 hr. Cell lysates were prepared and analysed for the presence of gadd153 protein by Western analysis.

Constitutive expression of gadd153 enhances apoptosis induced by 4HPR

To investigate the role of gadd153 in the action of 4HPR, a mammalian expression plasmid containing the full-length cDNA of gadd153 was transfected into CNE3 cells to achieve transient constitutive expression of the gadd153 protein. In cells transfected with the expression plasmid, the protein level of gadd153 was already detectable 8 hr after transfection, and a high expression of gadd153 was found at 24 hr after transfection (Fig. 5a). No detectable presence of the gadd153 protein was found in CNE3 cells transfected with the empty vector (pCMVneo) or in the wild-type CNE3 cells. In the next experiment, the expression of gadd153 in the transfected cells was examined after treatment with 4HPR (5 μM) for 24, 48 and 60 hr. As expected, the gadd153 protein in the transfected cells was constitutively expressed throughout the entire incubation period, with the highest expression occurring at the 24 hr time point. However, in the wild-type cells, the gadd153 protein was still not expressed at the 24 hr time point, and marked expression was only seen in the 48 and 60 hr time points (Fig. 5b). Thus, the wild-type CNE3 cells had a lag period between the application of 4HPR and the appearance of gadd153 protein, in contrast to the gadd153-transfected cells. When apoptosis was measured upon the addition of 4HPR for 24 to 60 hr, there was always substantially more cells with subG1-DNA content in the gadd153-transfected cells compared to that observed for cells transfected with the empty plasmid (Fig. 6a). The addition of 4HPR over the range of 1–5 μM produced more or less the same percentage of cells with sub-G1 DNA content in both the gadd153-transfected and the empty vector-transfected cells. However, at a dose of 10 μM, the percentage of sub-G1 DNA cells rose from less than 10% (in the empty vector-transfected cells) to almost 40% in the gadd153-transfected cells (Fig. 6b).

Figure 5.

The transient overexpression of gadd153 in CNE3 cells. (a) Cells were transfected with pCMVgadd153 or the control vector pCMVneo and were then incubated in complete medium for a fixed period of time as indicated in the brackets. The cells were then harvested and checked for the presence of gadd153 protein by immunoblotting. (b) Cells were transfected with pCMVgadd153 or the control vector pCMVneo and incubated in complete medium for another 16 hr. Cells were then treated with or without 4HPR (5 μM) for the indicated time intervals before harvest for Western blot analysis. The same experiments were performed twice with similar results.

Figure 6.

The enhancement of apoptosis in CNE3 cells that overexpress gadd153. After transfection with the CMVgadd153 or the CMVneo plasmid, the CNE3 cells were treated with 4HPR (5 μM) for different time intervals (a) or treated with 4HPR at 4 different doses for a fixed time period (28 hr) (b). The cells were then harvested for flow-cytometric analysis for the percentage of cells with sub-G1 DNA content. The experiment was performed twice with similar results.


The most important finding in our study is the upregulation of the gadd153 gene expression by 4HPR that also induced apoptosis in the CNE3 cells. DNA laddering was most apparent in the detached, 4HPR-treated cells, and this is interpreted as a consequence of the advancement of apoptosis. As the cells progress down the apoptotic pathway, the cells lost the ability to attach to the substratum. The gadd153 is a stress-induced, low-molecular-weight protein that belongs to the C/EBP family of transcription factors. gadd153 forms heterodimers with other members of the C/EBP family, resulting in inhibition of transcriptional activities initiated at the CCAAAT sites. The gadd153 was also shown to stimulate the expression of several genes like the carbonic anhydrase IV.23 Hence gadd153 functions as an inducer as well as a repressor of gene expressions. In our study, the stimulation of gadd153 expression by 4HPR was verified both by RT-PCR experiments and Western blot analysis. Although both the detached and adherent cells were harvested together and used in the RT-PCR experiments, only the adherent cells were used in the Western blot experiments. Together it means that gadd153 started to be expressed in the early stage of apoptosis, when the cells were still adherent. The induction of the expression of gadd153 by 4HPR in several human epithelial tumor cell lines suggests that this may be a mechanism of 4HPR in this category of cancers. An important feature of the 4HPR-induced expression of gadd153 protein is that the same response was not observed for ATRA. In the CNE3 cells, only 4HPR (but not ATRA) could stimulate an increase in gadd153-mRNA level, implying that the selectivity of action occurs at a very fundamental level of gene regulation. To investigate whether this was due to selective initiation of transcription, we studied the effect of 4HPR (and ATRA) on the promoter activity of gadd153 that was previously shown to respond to many stimuli for gadd153 upregulation.24 In the experiments with the CNE3 cells, the luciferase activity was in general much lower than that of the HeLa cells, even for the control plasmid that was driven by the SV-40 promoter. This may be due to incompatibility between the CNE3 cells and this type of reporter plasmid, resulting in low transfection efficiency. On the other hand, a more than 2-fold increase of the reporter activity in the HeLa cells was observed in the presence of either 4HPR or ATRA. Thus both 4HPR and ATRA is potentially capable of initiating the transcription of gadd153. The increase of gadd153-mRNA level in the CNE3 cells treated with 4HPR was consistent with such a prediction. Since both 4HPR and ATRA are capable of stimulating the gadd153 promoter but only 4HPR can produce an increase in gadd153 protein, the following possibilities exist. The gadd153 mRNA is constitutively degraded as rapidly as it is made and that 4HPR (but not ATRA) can suppress the degradation of the mRNA. Alternatively, ATRA (but not 4HPR) is able to accelerate the degradation of gadd153 mRNA. Further experiments will need to distinguish between these 2 possibilities.

Exactly how the gadd153 promoter is stimulated by 4HPR and ATRA is an important question. The intracellular effects of ATRA are mediated by the intracellular RARs, which on activation bind to the retinoic acid response elements (RARE). 4HPR is also known to bind to the γ subtype of RAR and at least in some systems lead to the expression of RARβ.25, 26 Thus the stimulation of the gadd153 promoter by both ATRA and 4HPR may be mediated by RAR. However, an examination of the DNA sequence of the gadd153 promoter did not reveal any direct repeats or palindromes of the hexameric sequence (i.e., AGGTCA), which is the consensus DNA-binding motif of RAR.27 Nevertheless, a single consensus sequence was found at approximately 100 base pairs upstream of the translation start site. It will be interesting to see if this is a functional RARE. It is equally possible that the stimulation of the gadd153 promoter is mediated by a transcription factor other than the RARs. The recent finding that retinoic acid is capable of inducing gene expression through the transcription factor Sp1 supported this idea.28 The identification of the 4HPR (and ATRA) response element in the gadd153 promoter will provide important insights into the regulation of gadd153 expression by these retinoids.

What might be the significance of 4HPR-induced gadd153 expression? When the gadd153 protein was constitutively expressed in the CNE3 cells, apoptosis induced by 4HPR was enhanced. Upon the addition of 4HPR, the steady state level of the gadd153 protein found in cells transfected with the gadd153 cDNA did not differ significantly from that of the empty-vector-transfected cells. However, the accumulation of gadd153 protein occurred much faster in the cells transfected with the cDNA and this was associated with an enhancement of apoptosis. Such data suggest the idea that gadd153 has a positive role to play in initiation of apoptosis induced by 4HPR, and it might do this through its modulating function in gene transcription. A positive role of gadd153 in 4HPR-induced apoptosis is also supported by the observation that both apoptosis and gadd153 expression occurred in the same dose range of 4HPR.

A number of genes had their mRNA levels downregulated in the presence of 4HPR. Most of the 20 genes that were affected by 4HPR in this way have functions related to the later part of the cell cycle. Notably these are genes that code for CDC25B, Cyclin-B1, PLK1, CDC/cdk1, Wee1 and CDC10, which are involved in the progression of the G2 or M-phase of the cell cycle. It is uncertain in our study if these represent direct or indirect targets of 4HPR and if they could be used to explain the reduction of S and G2/M-phase cells seen after 4HPR treatment. Most of these genes have the highest expression in the G2 or M-phases. A reduction of the number of cells in the G2 and M-phases after 4HPR treatment may account for their apparent downregulation in the cDNA-profiling experiment. However, a downregulatory effect directly exerted by 4HPR cannot be ruled out at this stage in the absence of any further evidence. If the downregulation of these genes is not a cause for the reduction of S-phase and G2/M-phase cells seen after 4HPR treatment, what then is the mechanism responsible for this effect of 4HPR? A block in transition from G1 to S-phase is one possibility. In this connection, a role of gadd153 should be considered since it has previously been implicated in G1 arrest.29


We thank Mr. W. Zhuang and Ms. S. Lin for technical advice and Dr. A.J. Fornace for supplying gadd153 expression plasmid. Our study was funded by the University of Hong Kong Research Grant to H.T. and a University of Hong Kong China Medical Board grant to N.S.W.