Regulation of the Catalase Gene Promoter by Sp1, CCAAT-recognizing Factors, and a WT1/Egr-related Factor in Hydrogen Peroxide-resistant HP100 Cells

There has been a significant accumulation of experimental data showing that ROS² such as the superoxide anion, hydrogen peroxide, the hydroxyl radical, and singlet oxygen can play critical roles as physiological mediators in the onset of apoptosis that occurs in response to various extracellular stimuli (reviewed in Refs. 1 and 2). ROS are thought to be involved in transition from the inducer phase to the effector phase during the progression of apoptosis, where the initial diverse signaling pathways converge into universal regulatory events, including the mitochondrial permeability transition (1). Due to subsequent generation in mitochondria, ROS are also considered to function in triggering transition from the effector phase to the degradation phase by the release of cytochrome c and the activation of caspases. Among the various ROS, the importance of H₂O₂ has been emphasized. For example, the inhibitory capacity of catalase in stress-induced apoptosis has been demonstrated in mammalian lymphocytes (3), fibroblasts (4), smooth muscle cells (5), neurons (6, 7, 8), and leukemia cell lines (9, 10). In addition, several cell lines with acquired resistance to apoptosis and an elevated activity of catalase have been isolated after long-term treatment with various oxidative stresses (11, 12, 13, 14, 15); at the same time, the activities of other ROS-metabolizing enzymes such as superoxide dismutase and GSHPx vary depending on the cell line (see Table 1). These observations suggest the existence of an antiapoptotic signaling pathway in which catalase functions in the down-regulation of cellular levels of H₂O₂ produced in response to proapoptotic stimuli. Thus, it is important to identify signaling molecules that cause the elevated expression of catalase in these cell lines.

Because one way that ionizing radiation kills cancer cells is by inducing apoptosis, elucidation and potential control of the pathway by which it elicits apoptosis are important goals for radiotherapy. The tumor suppressor protein p53 has been shown to serve as a pivotal component in radiation-induced apoptosis (16, 17). However, signaling pathways independent of p53 have also been revealed, including those involving ceramide, which is generated at the plasma membrane by sphingomyelinase (18, 19, 20). There is evidence that H₂O₂ plays a role in such ionizing radiation-induced apoptosis. For example, the transcription factor Egr-1 functions in p53-independent radiation-induced apoptosis (21), and its transcriptional induction by X-rays requires ROS (22). Furthermore, both X-ray and H₂O₂-dependent induction of the Egr-1 gene have been shown to be mediated through a CC(A+T-rich)₆GG element situated in its 5′-flanking region (23). The involvement of catalase in DNA damage-induced apoptosis has recently been demonstrated through the use of a H₂O₂-resistant cell line, HP100. This cell line, which was isolated from HL60 by repeated exposure to H₂O₂, overexpresses catalase and displays significant resistance to H₂O₂-induced apoptosis (11, 12). Production of H₂O₂ in both HL60 and HP100 cells has been observed after treatment with a DNA-damaging agent, but its generation and the subsequent activation of caspase 3, loss of mitochondrial transmembrane potential (ΔΨ_m), and DNA ladder formation were delayed in HP100 cells as compared with HL60 cells (24).

In considering a critical function for H₂O₂ in ionizing radiation-induced apoptosis, regulation of the activity of H₂O₂-metabolizing enzymes, such as catalase and GSHPx, is likely to be required. Thus, it is interesting to note that catalase mRNA levels are decreased after X-ray irradiation in HP100 cells.³ A reduction in catalase mRNA has also been observed in mouse splenocytes after γ-ray irradiation (25). These observations strongly suggest that ionizing radiation-inducible apoptosis is mediated, at least in part, by a signaling pathway involving the stabilization of H₂O₂ via a down-regulation of catalase.

With this in mind, we have investigated the molecular mechanism for elevated expression of the catalase gene in HP100 cells, as well as its down-regulation by ionizing radiation. We found that the elevated catalase expression is strongly regulated at the transcriptional level by both Sp1 and CCAAT-recognizing factors and that much higher levels of nuclear Sp1 and NF-Y are expressed in HP100 nuclei as compared with HL60 nuclei. In addition, we also demonstrated that association of a WT1/Egr-related factor with the overlapping Sp1/Egr-1 recognition sequence at the core promoter of the catalase gene is induced by X-rays; this association may lead to inactivation of the core promoter by disturbing or competing with the transactivating ability of Sp1.

Human HL60 promyelocytic leukemia cells and their H₂O₂-resistant variant cell line, HP100, were a gift from Dr. M. Akashi. The cells were cultured in RPMI 1640 (Life Technologies, Inc.) supplemented with 10% fetal bovine serum (JRH), 100 units/ml penicillin, and 100 μg/ml streptomycin (Life Technologies, Inc.) at 37°C under 5% CO₂ in a humidified atmosphere. HP100 cells were routinely maintained in the presence of 100 μm H₂O₂, although it was removed from the medium 7 days before the experiments described here. When 3AT, a specific inhibitor of catalase, was used, cells were incubated for 2 h in the presence of 25 mm 3AT (Sigma Chemical Co.) before treatment with H₂O₂. 3AT has been used to modulate catalase activity in a wide variety of cell types, including HL60 cells (26).

The induction of apoptosis was examined by DNA fragmentation (27). Cells were treated with 100 μm H₂O₂ in the presence or absence of 3AT, followed by DNA extraction using a kit supplied by Takara (ApopLadder EX Kit) that can efficiently remove DNA fragments of high molecular weight.

X-rays were generated from a Pantak unit operating at 200 kilovolt peak (kVp) and 20 mA, with a 0.5 mm copper plus 0.5 mm aluminum filter. A dose rate of 1.4 Gy/min was used.

cDNAs for human catalase, GAPDH, and GSHPx were amplified by reverse transcription-PCR and cloned into pCR2.1 (Invitrogen). Next, 5 μg of linearized and denatured plasmids were spotted and fixed onto a GeneScreen membrane (New England Nuclear Life Science Products).

Cells (5 × 10⁷) were irradiated with 20 Gy of X-rays and then incubated for various time intervals at 37°C. After washing with ice-cold PBS, cells were suspended in 4.5 ml of ice-cold hypotonic buffer [10 mm Tris-HCl (pH 7.4), 10 mm KCl, and 3 mm MgCl₂], left on ice for 30 min, and lysed by adding 0.5 ml of 5% NP40, followed by incubation on ice for 5 min (28). Subsequently, nuclei were pelleted by centrifugation at 110 × g for 5 min at 4°C, washed with 5 ml of 0.5% NP40, and resuspended in 200 μl of storage buffer [40% glycerol, 50 mm Tris-HCl (pH 8.3), 5 mm MgCl₂, and 0.1 mm EDTA]. mRNA elongation was performed by mixing the nuclear suspension with 100 μl of reaction buffer containing 340 mm KCl; 5 mm MgCl₂; 10 mm Tris-HCl (pH, 8.0); 4.5 mm DTT; 0.75 mm each of ATP, CTP, and GTP; 36 units of RNase inhibitor (Takara), and 200 μCi of [α-³²P]UTP (3000 Ci/mmol; ICN Biomedicals), followed by incubation at 30°C for 30 min. After the DNA was digested with 210 units of RNase-free DNase I (Takara), elongated mRNA was isolated following the standard protocol of phenol/chloroform extraction.

Denatured RNA was hybridized with the membrane filter in hybridization buffer [50% formamide, 3× SSC, 10× Denhardt’s solution, 10 mm phosphate buffer (pH 7.0), 0.1 mg/ml salmon sperm DNA, 0.2 mg/ml yeast tRNA, and 0.1% SDS] for 72 h at 42°C. The membrane was washed sequentially with 2× SSC at room temperature for 30 min, 2× SSC at 42°C for 30 min, and 0.1× SSC at 42°C for 30 min.

The 5′-flanking region of the human catalase gene had been previously sequenced up to −1527 bp with respect to the major mRNA initiation site (29). Based on the nucleotide sequence reported there, a further upstream region was cloned and sequenced in chromosome-walking fashion using a kit supplied by Clontech (Genome Walker kit). The sequenced data were deposited in the DNA Data Bank of Japan, European Molecular Biology Laboratory, and GenBank nucleotide sequence databases under accession number AB034940.

Regions containing the catalase gene promoter spanning nucleotides −4526/+16, −3113/+16, −2379/+16, −1518/+16, −1226/+16, −935/+16, −733/+16, −404/+16, and −176/+16 were PCR-amplified using the 5′ primers 5′-CAATGGTACCGTCTATGTCCACGTCCTTTGCTGC-3′, 5′-CAATGGTACCAAACACCAGATCAGTAGCGTGGC-3′, 5′-CAATGGTACCGAGTTCTGAAAATTGACTTCAGAGAACAGC-3′, 5′-CAATGGTACCTGTGGACTTTGGAGATGAACAGCTG-3′, 5′-CAATGGTACCGACACCAAATTACACAG CCAACAGCATC-3′, 5′-CAATGGTACC-AATCCTAGCACCTGAGGAGGTGTAG-3′, 5′-CAATGGTACCGAA GCCAATTTGGCAGTGTACCAGAG-3′, 5′-CAATGGTACCGCTGAGAAAGCATAGCTATG GAGCG-3′, and 5′-CAATGGTACCTATCTCCGGTCTTCAGGCCTCCTTC-3′, respectively, and the common 3′ primer 5′-GTCAGATCTCAGCAGGCAAATCTGCCTGTTGC-3′; the underlined bases were introduced to facilitate cloning with KpnI for 5′ primers and BglII for 3′ primers. LA Taq DNA polymerase (Takara) was used to ensure high-fidelity amplification. The PCR fragments were ligated into pGL3-Basic Photinus pyralis luciferase reporter vector (Promega). Recombinant plasmids for each construct were extracted and purified by alkaline lysis and ion-exchange chromatography (Genomed) following the manufacturer’s instructions. Site-directed mutagenesis was performed on the CAT-404 construct containing the region −404/+16 by use of a kit supplied by Promega (GeneEditor in vitro Site-Directed Mutagenesis System).

Transfection was carried out by electroporation. Cells grown to a density of approximately 10⁶ cells/ml were washed three times with serum-free RPMI 1640. Cells (2 × 10⁷) were mixed with 35 μg of test plasmid DNA and 1–2 μg of pRL-SV40 (Renilla reniformis luciferase under the control of SV40 early enhancer/promoter; Promega) as a transfection efficiency control in 240 μl of serum-free RPMI 1640 and electroporated with the Gene Pulser (Bio-Rad) in a 0.4-cm electroporation cuvette at 250 V with a capacitance of 960 microfarads. Cell lysates were prepared 48 h after transfection by adding 100 μl of Passive Lysis Buffer (Dual-Luciferase Reporter Assay System; Promega) to the cell pellet. The luciferase activity was measured with an analytical luminometer (model LB9506; Berthold), and the promoter activity was evaluated as relative light units, which is defined as the ratio of the light intensity produced by Photinus luciferase (test plasmid) to that produced by Renilla luciferase (control plasmid).

HL60 and HP100 cells grown to a density of approximately 10⁶ cells/ml were irradiated with various doses of X-rays and then incubated for 1 h at 37°C. After washing with ice-cold PBS, 5 × 10⁶ cells were permeabilized by treatment with lysolecithin (Sigma Chemical Co.; 0.002% for HL60 and 0.001% for HP100) in ice-cold permeabilization buffer [35 mm HEPES (pH 7.4), 5 mm potassium phosphate (pH 7.4), 80 mm KCl, 5 mm MgCl₂, 0.5 mm CaCl₂, and 150 mm sucrose] for 2 min. After washing with digestion buffer [35 mm HEPES (pH 7.4), 5 mm potassium phosphate (pH 7.4), 80 mm NaCl, 5 mm MgCl₂, 2 mm CaCl₂, and 150 mm sucrose], the permeabilized cells were treated with DNase I (Takara; 4.0 units/ml for HL60 and 11 units/ml for HP100) in digestion buffer at 25°C for 5 min to partially digest the DNA within chromatin. The reaction was stopped by adding 0.1 volume of 6% SDS and 250 mm EDTA (pH 8.0). DNA was purified by a standard phenol/chloroform extraction method. A naked DNA control was prepared by cleaving high molecular weight genomic DNA isolated previously from HL60 and HP100 cells with 0.09 unit/ml DNase I at 25°C for 2 min in digestion buffer. The DNase I cleavage conditions described above resulted in a broad distribution of fragment sizes ranging from shorter than 100 bp to longer than 10 kb as determined by agarose gel electrophoresis.

The DNase I digestion pattern in the promoter region of the catalase gene was assessed by a LM-PCR method (30, 31, 32, 33) with the addition of an extension product capture step (34). The unidirectional double-stranded linker DNA, which was blunted at one end, was prepared by annealing a pair of complementary oligonucleotides, L25 (5′-GCGGTGACCCGGGAGATCTGAATTC-3′) and L11 (5′-GAATTCAGATC-3′). 5′-Biotinylated primers (primer 1, 5′-ATAGCGTGCGGTTTGCTGTGCAGAAC-3′) complementary to the upper strand at site +75/+50 were annealed to heat-denatured DNase I-treated genomic DNA (5 min at 95°C, 30 min at 60°C) and extended using Vent DNA polymerase (New England Biolabs) for 10 min at 76°C in 30 μl of the first-strand mixture containing 40 mm NaCl, 10 mm Tris-HCl (pH 8.9), 5 mm MgSO₄, 0.01% gelatin, 0.2 mm each deoxynucleotide triphosphate, 0.3 pmol of primer, 2 μg of DNase I-treated DNA, and 1.2 units of Vent polymerase. After diluting the reaction mixture with 20 μl of Vent dilution buffer [110 mm Tris-HCl (pH 7.5), 18 mm MgCl₂, 50 mm DTT, and 0.125 mg/ml DNase-free BSA], the linker DNA was ligated to DNase I-cleaved sites (where ligatable blunt ends were created by the primer extension) by the addition of 25 μl of ligation solution (10 mm MgCl₂, 20 mm DTT, 3.75 mm ATP, 0.05 mg/ml DNase-free BSA, 4 pmol of linker DNA, and 16 units/ml T4 DNA ligase). The ligation reaction was carried out overnight at 16°C, followed by precipitation with ethanol. A 175-μg portion of Dynabeads M-280 streptavidin (Dynal) was washed twice and incubated with ligation products in 70 μl of binding buffer [10 mm Tris-HCl (pH 7.7), 1 mm EDTA, and 1 m NaCl] for 15 min at room temperature. After washing the beads with 10 mm Tris-HCl (pH 7.7), 1 mm EDTA, and 2 m NaCl, the nonbiotinylated template strand was eluted from the beads by incubation at 37°C for 10 min in 35 μl of 0.15 n NaOH. The eluted DNA was neutralized and precipitated with ethanol. The fragments of interest, which contained the promoter region of the catalase gene, were PCR-amplified using linker primer (L25) and a second site-specific primer (primer 2, 5′-TGTGCAGAACACTGCAGGAGGCCTC-3′), which anneals at +59/+35 just upstream of the primer 1 annealing site. The reaction was carried out for 18 cycles (1 min at 95°C, 2 min at 68°C, and 3 min at 76°C) in an amplification mixture containing 40 mm NaCl, 20 mm Tris-HCl (pH 8.9), 5.2 mm MgSO₄, 0.009% gelatin, 0.09% Triton X-100, 0.2 mm each deoxynucleotide triphosphate, 10 pmol of primer 2, 10 pmol of L25, and 3 units of Vent polymerase. After amplification was completed, 2.3 pmol of ³²P-labeled primer (primer 3, 5′-TGTGCAGAACACTGCAGGAGGCCTCGGC-3′), whose annealing site overlaps with that of primer 2 with an additional three bases extending to the 3′ end) were added, and two cycles of primer extension were carried out (1 min at 95°C, 2 min at 71°C, and 10 min at 76°C). The reaction products were purified by phenol/chloroform extraction, precipitated with ethanol, and subjected to 6% PAGE. The gels were dried, and the bands were visualized by autoradiography. The site specificity of the primers used in this experiment was confirmed in advance by observing a single specific band when LM-PCR was carried out with genomic DNA digested with a restriction enzyme as the template.

Cells (10⁸) grown to a density of approximately 10⁶ cells/ml were irradiated with 20 Gy of X-rays and then incubated for 1 h. After sequential washing with ice-cold PBS and hypotonic buffer [10 mm HEPES (pH 7.9), 1.5 mm MgCl₂, 10 mm KCl, 0.2 mm PMSF, and 0.5 mm DTT], cells were incubated in 320 μl of hypotonic buffer for 10 min on ice. Cells were then homogenized with a Dounce homogenizer. Subsequently, nuclei were collected by centrifugation and suspended in an equal volume of extraction buffer [20 mm HEPES (pH 7.9), 25% (v/v) glycerol, 1.5 mm MgCl₂, 0.6 m KCl, 0.2 mm EDTA, 0.2 mm PMSF, and 0.5 mm DTT]. After incubation for 30 min on ice, nuclear extracts were dialyzed against a dialysis buffer [20 mm HEPES (pH 7.9), 20% (v/v) glycerol, 100 mm KCl, 0.2 mm EDTA, 0.2 mm PMSF, and 0.5 mm DTT] for 2 h. After centrifugation at 25,000 × g for 20 min, supernatant was stored at −80°C. Protein concentrations were measured by the Bradford method using a kit supplied by Pierce (Coomassie Protein Assay Reagent Kit).

Double-stranded DNA probes were prepared by annealing oligonucleotides spanning −113/−80 (5′-GGCCTGCCTAGCGCCGAGCAGCCAATCAGAAGGC-3′) and −78/−45 (5′-GTCCTCC CGAGGGGGCGGGACGAGGGGGTGGTGC-3′), and they were end-labeled with [γ-³²P]ATP (4,000 Ci/mmol; ICN Biomedicals) and T4 polynucleotide kinase (Takara). Nuclear extracts, prepared as described above, were mixed with 10,000 cpm (0.5 ng) of the probe in 20 mm HEPES (pH 7.9), 70 mm KCl, 5 mm MgCl₂, 0.05% NP40, 1 mg/ml BSA, 0.5 mm DTT, 0.1 mg/ml poly(deoxyinosinic-deoxycytidylic acid), and 12% glycerol and incubated for 20 min at 25°C. The reaction mixture was subjected to electrophoresis on a 6% polyacrylamide gel with 0.25× Tris-borate EDTA running buffer. The gels were dried, and the bands were visualized by autoradiography. In competition studies, either double-stranded Egr-1 consensus oligonucleotide (5′-GGATCCAGCGGGGGCGAGCGGGGGCGA-3′) or Egr-1 mutant oligonucleotide (5′-GGATCCAGCTAGGGCGAGCTAGGGCGA-3′; Santa Cruz Biotechnology) was added at a 100-fold excess over the radiolabeled probe. In supershift studies, polyclonal antibodies against either NF-Y (AHP298; Oxford Biomarketing) or Sp1 (PEP2; Santa Cruz Biotechnology) were used. When either competitor oligonucleotides or antibodies were used, they were preincubated with the nuclear extracts at 4°C for 25 min before the addition of the radiolabeled probe DNA.

A 4-μg sample of nuclear proteins prepared as described above was separated on a 10% SDS polyacrylamide gel and electrotransferred to a Hybond-P polyvinylidene difluoride membrane (Amersham Life Science). The membrane was then incubated sequentially with 2% BSA in TBS-T for 1 h, with either 0.067 μg/ml anti-Sp1 antibody (PEP2) or 1 μg/ml anti-NF-Y antibody (AHP298) in TBS-T for 1 h, and with a 1:1000 dilution of horseradish peroxidase-conjugated secondary antibody (Amersham Life Science) in TBS-T for 1 h. The bands were visualized by enhanced chemiluminescence using reagents supplied by Amersham Life Science (enhanced chemiluminescence Western blotting detection reagent).

The HP100 cell line is a stable variant isolated from HL60 by repetitive treatment with 100 μm H₂O₂ for 5 months. The effective dose of H₂O₂ for killing 50% of the cell population of HP100 (2.4 mm) was reported to be much higher than that of HL60 (7 μm; Ref. 11). Because HP100 overexpresses catalase, with the enzymatic activity being 18-fold higher than that of HL60 (12), it appears likely that the elevated catalase inhibits the induction of apoptosis via its enhanced decomposition of H₂O₂. However, the possibility that other apoptotic pathway signaling components may be disrupted in HP100 cannot be excluded. We therefore examined the effects of a catalase-specific inhibitor, 3AT, on the sensitivity of HP100 to H₂O₂. Fig. 1,A shows that HP100 cells were resistant to 100 μm H₂O₂ and grew with a doubling time similar to that of untreated cells, but that the growth of HP100 cells was completely inhibited if 25 mm 3AT was added 2 h before the addition of H₂O₂; treatment with 3AT alone did not affect the growth of these cells. The observed growth inhibition resulted from apoptosis because DNA fragmentation was only observed when the cells were treated with H₂O₂ in the presence of 3AT (Fig. 1 B). This is in contrast to HL60 cells, which readily die by apoptosis in 2 min after treatment with 20 μm H₂O₂ alone (35). These results suggest that the H₂O₂-induced apoptotic signaling pathway is intact in HP100 cells and that the overexpressed catalase plays a decisive role in the resistance of HP100 cells to H₂O₂. Regulation of catalase expression is therefore critical for induction of apoptosis in HP100 cells.

By Northern analyses, it has been determined that the level of catalase mRNA is 16-fold higher in HP100 cells than in HL60 cells (12) and that this level decreases after X-ray irradiation.³ A possible mechanism for this phenomenon involves regulation of the catalase gene at the transcriptional level. However, the 3′-untranslated region of the human catalase gene is highly A/T rich, and the presence of four ATTTA sequences (36) suggests the possibility of a regulated degradation of its transcripts. Actually, the half-life of catalase mRNA has been shown to be variable in the lung of rats when exposed to an oxidative stress (37). In addition, it has been reported that the catalase gene is amplified in HP100 cells (12), implicating a mRNA elevation caused by the increased gene dosage. Therefore, we performed a transient expression analysis of luciferase gene constructs and a nuclear run-on analysis to determine whether catalase gene expression is regulated at the transcription level.

For the transient expression analysis, we examined the region up to −4526 bp from a major catalase transcription start site located −73 bp upstream of the initiation codon (29). As shown in Fig. 2, an enhanced promoter activity (3.4–8.3-fold) was observed in HP100 cells compared with that in HL60 cells for every construct containing the 5′-flanking region (anywhere from −176 to −4526). We conclude that elevated catalase mRNA in HP100 cells is due to an enhanced transcription of individual genes and that the majority of cis-regulatory elements causing its elevation are located in the region from −176 to +16 bp.

Next we performed transient transfections followed by irradiation of transfectants with X-rays. However, no convincing down-regulation of the catalase gene promoter in HP100 cells was seen. Because the luciferase protein encoded by pGL3-Basic vector is artificially modified to be stabilized in mammalian cells, we speculate that any transcriptional down-regulation may have been masked by long-lived luciferase proteins that had accumulated before X-ray irradiation. In a further attempt to address this issue, we performed a nuclear run-on experiment. As shown in Fig. 3,A, a drastic and transient decrease in transcription of the catalase gene in HP100 cells after irradiation with 20 Gy of X-rays was revealed. When normalized to the transcriptional activity of the GAPDH gene, transcription of the catalase gene after 1 h of incubation was reduced to one-tenth of that in unirradiated HP100 cells (Fig. 3 B). It can therefore be concluded that the reduction in catalase mRNA levels after X-ray irradiation in HP100 cells is regulated primarily at the transcriptional level.

The transcriptional activity of the catalase gene in unirradiated HP100 cells was 20-fold higher than that in HL60 cells (Fig. 3,B), a figure that was similar to the previously reported fold increase in mRNA (12). However, it was larger than that observed in the transient expression analysis (5.6-fold for the CAT-4526 construct in Fig. 2). This apparent discrepancy may be attributable to the amplification of the catalase gene that occurs in HP100 cells.

In Fig. 3, no difference in the transcriptional activity of the GSHPx gene, which encodes another H₂O₂-decomposing enzyme, can be observed between HL60 and HP100 cells, which is consistent with the observation by Kasugai and Yamada (38) that both cell lines show comparable GSHPx enzymatic activity. Furthermore, no change in the transcriptional activity of the GSHPx gene was observed after X-ray irradiation.

To elucidate the molecular mechanisms behind the transcriptional regulation of the catalase gene, we attempted to identify the cis-regulatory elements that caused its elevated transcription in HP100. For this purpose, we analyzed DNA-protein interactions in the upstream region of the catalase gene by DNase I genomic footprinting. Fig. 4,A shows the DNase I cleavage pattern on the upper strand in a region encompassing the transcription start site to approximately −200 bp, which fully covers the region −176 to +16 where the cis-regulatory elements causing elevated transcriptional activity in HP100 cells had been determined to reside based on the transient expression analysis (Fig. 2). Throughout the region examined, the DNase I cleavage pattern was very similar between HL60 and HP100 cells as well as between irradiated HP100 cells and unirradiated HP100 cells. However, in the vicinity of the transcription start site, strongly increased sensitivity to DNase I in HP100 cells was observed (HS1 and HS2), which suggests that this region has an open chromatin conformation in this cell type. In addition, clear footprints (FP1, FP2, and FP3) were observed in both HL60 and HP100 cells in the region from −100 bp to −50 bp, where the band intensities were obviously lower compared with that in the naked DNA lane (Lanes N). Fig. 4 B shows a magnified portion of the region around these observed footprints. Three bands at approximately −99, −98, and −69 bp (arrows) that are more intense in HP100 cells as compared with HL60 cells can be clearly observed, as well as two bands at approximately −101 and −75 bp (arrowheads designated X) that show an X-ray-dependent reduction in intensity in HP100 cells. In the naked DNA, differential DNase I sensitivity was observed for bases at approximately −101 and −75 bp when comparing HL60 and HP100 cells. It appears that these discrepancies are not due to a point mutation in HP100 DNA because we were able to confirm the identity of the nucleotide sequence in HL60 and HP100 cells across the whole region investigated. It is supposed that when the naked DNA was isolated, removal of nuclear proteins may have been insufficient at these locations, potentially causing the differential base sensitivity to DNase I attack.

Because the FP1, FP2, and FP3 footprints were not altered between HL60 and HP100 cells or by X-ray irradiation, it can be concluded that nuclear proteins are stably bound to these sites. However, we observed that the cleavage intensities of several bases at the edges of the footprints were altered between HL60 and HP100 cells (indicated by arrows in Fig. 4 B), which indicates that some change in the fine structure of the DNA-protein interaction occurs between these two cell types. Such changes may be caused by an altered interaction between the bound proteins, possibly mediated by a non-DNA binding coactivator or alternatively by a substitution of the bound proteins with some others. This result strongly suggests that the cis-regulatory elements that cause elevated transcription of the catalase gene in HP100 cells are located either at the observed footprints or in their immediate flanking regions.

The cis-regulatory elements responsible for down-regulation of the catalase gene in HP100 cells after X-ray irradiation were not mapped directly by transient transfection analysis in this study (see above). However, Fig. 4 B shows an X-ray-dependent change in the DNase I sensitivity of bases (indicated by arrowheads designated X) that are only a few nucleotides removed from those bases whose in vivo DNase I sensitivity differed between HL60 and HP100 cells. These results suggest that specific DNA-protein interactions thought to play a role in the elevated catalase transcription of HP100 cells are further modified by X-rays and strongly imply that a common cis-regulatory element functions in both the elevated transcription in HP100 and its down-regulation by X-rays.

Fig. 4 C depicts the nucleotide sequence around the observed footprints. There are numerous motif sequences of transcription factors. Among these, clustered CCAAT boxes (CCAAT-42, CCAAT-92, CCAAT-121, and CCAAT-130) and tandemly repeated Egr-1 recognition sequences (Egr-1–57 and Egr-1–70; Ref. 39), one of which is overlapped by a Sp1 recognition sequence (Sp1–67; Ref. 40), characterize this region. These elements are important in regulation of redox-sensitive genes. The CCAAT-92 sequence is located at the FP1 footprint, and the Sp1–67, Egr-1–70, and Egr-1–57 sequences are colocated at the FP3 footprint. These motifs are certainly candidates for the functional elements that cause the elevated transcription of the catalase gene in HP100 cells as well as its reduced transcription after X-ray irradiation.

To identify the functional cis-regulatory element(s) responsible for elevated transcription of the catalase gene in HP100 cells, mutations were introduced into the luciferase construct CAT-404 (Fig. 2) by site-directed mutagenesis. As shown in Fig. 5,A, the CCAAT sequences (CCAAT-42, CCAAT-92, CCAAT-121, and CCAAT-130) and the overlapping recognition sequence for Egr-1 and Sp1 (EgrSp-70) were substituted with sequences containing recognition sites for restriction enzymes (Mut-42, Mut-92, Mut-121, Mut-130, and Mut-70, respectively) to facilitate screening of the mutated constructs. These constructs were transfected into HL60 and HP100 cells, and the luciferase activity was measured 48 h after transfection. As shown in Fig. 5 B, a mutation at EgrSp-70 resulted in loss of most of the basal activity of the promoter, clearly demonstrating that EgrSp-70 plays a critical role in transcription of the catalase gene. The GC box in EgrSp-70 may therefore function as the core promoter in the TATA-less catalase gene. A mutation at CCAAT-92 had no influence on promoter activity in HL60 cells, although it drastically reduced promoter activity in HP100 cells to a level almost identical to that in HL60 cells. These results suggest that CCAAT-92 functions as a transcriptional enhancer in HP100 cells. In contrast, a mutation at CCAAT-130 did not significantly affect promoter activity in either HL60 or HP100 cells, indicating a nonessential role for CCAAT-130. Interestingly, a mutation at either CCAAT-42 or CCAAT-121 led to increased promoter activity in HL60 cells but to decreased promoter activity in HP100 cells. This opposing function may be due to recognition by different factors in HL60 and HP100 cells or, alternatively, to binding factors being modified differently in these cells. In summary, it can be concluded that the GC box in EgrSp-70 functions as a core promoter for the catalase gene and that CCAAT-42, CCAAT-92, and CCAAT-121 function as modulators of transcriptional activity. CCAAT-92 functions as an enhancer element in HP100 cells, whereas both CCAAT-42 and CCAAT-121 function as either negative or positive modulators in HL60 and HP100 cells, respectively. All these elements are necessary for the elevated transcription of the catalase gene in HP100 cells.

To identify the factors binding to the cis-regulatory elements, we performed EMSA analyses. Because we observed that CCAAT-92 functions as a transcriptional enhancer in HP100 cells (Fig. 5), EMSA analysis was first performed with a probe containing the CCAAT-92 element (Fig. 6,A). As shown in Fig. 6 B, multiple complexes were formed with nuclear extracts from both HL60 and HP100 cells. However three additional shifted bands (S1, S2, and S3) were observed when nuclear extracts from HP100 cells were used. Incorporation of anti-NF-Y antibodies caused a disappearance of S3, demonstrating that S3 represents a complex containing NF-Y. Because these complexes were not observed when nuclear extracts from HL60 cells were used, at least some of the factors composing these complexes are likely to be responsible for the elevated transcription of the catalase gene in HP100 cells. However, because the band pattern was not changed when HP100 cells were irradiated with X-rays, these factors are probably not involved in down-regulation of the catalase promoter by X-rays.

The results of our genomic footprinting analysis suggest that a common cis-regulatory element functions in both the elevated transcription of the catalase gene in HP100 cells and its down-regulation by X-rays. In this regard, it is interesting to note that Egr-1 has been shown to down-regulate the rat Pgp2/mdr1b gene (41) and the mouse adenosine deaminase gene (42) promoters by competing with Sp1 for binding to an overlapping sequence motif. In addition, Egr-1 is an immediate early gene responding to ionizing radiation and is inducible less than 1 h after irradiation (21, 22). Accordingly, we hypothesize that Egr-1 is induced by X-rays in HP100 cells and that competition with Sp1 for binding to the core promoter in EgrSp-70 results in reduced promoter activity. To test this hypothesis, we performed EMSA analyses with the probe containing EgrSp-70 sequence (Fig. 7,A). As shown in Fig. 7 B, we were able to observe three shifted bands (S1–S3) when 3.3 μg of nuclear protein per reaction were used. We found, however, that an additional band (S4) appeared when a low concentration of HP100 nuclear proteins (0.10–0.41 μg/reaction) was used. This observation can be explained by assuming that the nuclear factor that forms the S4 complex acquires the ability to interact with itself or other proteins after binding to DNA and that when a high concentration of nuclear proteins was used, the binding of multiple proteins to the S4 complex retarded its migration on the gel, giving rise to an accidental comigration with one of the S1–S3 bands. To avoid complications, we carried out the EMSA experiments described below under conditions in which S4 was observable (0.22 μg nuclear protein/reaction).

When anti-Sp1 antibodies (PEP2) were incorporated, the intensity of S1 obviously decreased, and S2 and S3 disappeared (Fig. 7 C); this disappearance was accompanied by the appearance of the supershifted bands SS1 and SS2. These results demonstrate that S1, S2, and S3 represent complexes containing Sp1. Combining this result with the finding that the GC box in EgrSp-70 functions as a core promoter and the previously proposed model that Sp1 functions in recruiting RNA polymerase II by tethering the preinitiation complex to TATA-less promoters (43, 44), it appears that Sp1 plays an essential role in transcription of the catalase gene by binding to the GC box in EgrSp-70.

On the other hand, the addition of an Egr-1 consensus competitor oligonucleotide, but not an Egr-1 mutant oligonucleotide, at a 100-fold molar excess obviously reduced the S4 band intensity (Fig. 7,D), demonstrating that S4 contains a member of the WT1/Egr protein family. In addition, Fig. 7, B–D, reveals that the S4 band intensity was increased after X-ray irradiation, whereas a significant change was not observed for S1, S2, and S3. Therefore, we conclude that the WT1/Egr-related factor is X-ray inducible.

When EMSA was carried out with nuclear extracts of HL60 cells at the same concentration (0.22 μg/reaction), bands S1 and S4 were observed only at very low intensity (Fig. 7,C). This demonstrates an extremely low binding activity of both Sp1 and the WT1/Egr-related factor in HL60 cells, similar to the observation with NF-Y made in Fig. 6. We further investigated the nuclear content of Sp1 and NF-Y by Western blotting analysis. Fig. 8,A shows Coomassie Blue staining of nuclear proteins separated by SDS-PAGE, revealing similar amounts of proteins loaded on the gel. When probed with anti-Sp1 antibodies (PEP2), two specific protein species were observed, as shown in Fig. 8 B, the lower of which is considered to be a degradation product of Sp1. It is evident that both Sp1 and NF-Y are quite abundant in the nuclei of HP100 cells as compared with the nuclei of HL60 cells.

In summary, it seems that remarkably high levels of Sp1 are expressed in HP100 nuclei, which leads to activation of the core catalase promoter in cooperation with CCAAT-recognizing factors. Moreover, by analogy with the observation that members of the WT1/Egr family are capable of repressing a wide range of mammalian promoters (45, 46, 47, 48), an X-ray-inducible WT1/Egr-related factor, which forms the S4 complex, appears to function by inactivating the core promoter in competition with Sp1. However, because the binding activity of Sp1 does not appear to change after X-ray irradiation, the WT1/Egr-related factor may also act by modulating the transactivation ability of Sp1.

HP100 cells are remarkably resistant to H₂O₂-inducible apoptosis. In this study, we have demonstrated that the overexpression of catalase plays a decisive role in this phenotype (Fig. 1). We then investigated the molecular mechanisms responsible for the elevated expression of catalase in HP100 cells and for its down-regulation by X-rays, and we found that: (a) the expression of catalase is regulated primarily at the transcriptional level; (b) the GC box located −70 bp from the major transcriptional start site of the catalase gene functions as the core promoter element; (c) much higher levels of Sp1 are expressed in HP100 cells than in HL60 cells, associating with the overlapping Sp1/Egr-1 recognition sequence at −70 bp; (d) a WT1/Egr-related factor is induced in response to 20 Gy of X-ray irradiation and associates with the overlapping Sp1/Egr-1 recognition sequence at −70 bp; (e) the X-ray-inducible WT1/Egr-related factor does not seem to compete with Sp1 for binding to their overlapping recognition sequence; (f) the CCAAT element at −92 bp strongly enhances transcription in HP100 cells; (g) higher levels of NF-Y are expressed in HP100 cells than in HL60 cells, associating with the CCAAT element at −92 bp; and (h) both the inverted CCAAT sequence at −42 bp and the CCAAT sequence at −121 bp regulate transcription negatively in HL60 cells and positively in HP100 cells.

Based on these findings, we suggest that a mechanism such as that shown in Fig. 9 could be the means by which promoter regulation of the catalase gene occurs. Fig. 9,A illustrates the catalase gene promoter in HL60 cells, where the association of factors/complexes (F2 and F3) with the CCAAT-92 and EgrSp-70 sequences is revealed by genomic footprinting (Fig. 4). In addition, the involvement of factors (F1 and F4) recognizing the CCAAT-121 and CCAAT-42 sequences is also expected because these CCAAT elements work to repress the promoter activity (Fig. 5,B). The presence of DNase I-sensitive sites observed in the genomic footprinting (Fig. 4) suggests an interaction among factors. Because of the low nuclear content of Sp1 and NF-Y in HL60 cells, the transcription initiation complex containing RNA polymerase II may not be very effectively recruited to the transcriptional start site. In HP100 cells, on the other hand, displacement of CCAAT-92-bound factors (F2) by a complex containing NF-Y (F2′/NF-Y) may have occurred (Fig. 6), resulting in the appearance of DNase I-sensitive sites at approximately −98 and −99 bp (Fig. 4,B). Modification or replacement of the F1 and F4 factors may also be important because the CCAAT-42 and CCAAT-121 elements function in enhancing the promoter activity in this cell. Binding of Sp1 to the EgrSp-70 element might now result in an assemblage of factors that allows efficient entry of the initiation complex at the transcriptional start site (Fig. 9,B). NF-Y bound to the CCAAT-92 sequence may cooperatively stimulate this process, as has been reported for the transcription of the type A natriuretic peptide receptor gene (49) and the rat fatty acid synthase gene (50). When HP100 is irradiated with X-rays, induction of the WT1/Egr-related factor may somehow lead to complex disassembly, resulting in the initiation complex being inhibited from entry (Fig. 9,C). The identification and precise characterization of factors postulated in Fig. 9 remain the subject of future study. In particular, identification of the WT1/Egr-related factor is of great interest in relation to the potential involvement of catalase in ionizing radiation-inducible apoptosis.

It is natural to believe that the overexpression of catalase confers resistance of HP100 cells to H₂O₂-inducible apoptosis because H₂O₂ is thought to be rapidly metabolized in these cells. This was confirmed in the present study by showing that resistance to H₂O₂-inducible apoptosis is lost when an inhibitor of catalase, 3AT, is added to the HP100 culture. However, in cell lines established by long-term treatment with oxidative stresses other than H₂O₂, elevated catalase activity is also consistently observed (Table 1). This suggests that H₂O₂ may play a critical role in the apoptotic pathway inducible by a wide range of oxidative stresses. Thus, our finding that Sp1 and CCAAT-recognizing factors participate in the activation of catalase in HP100 cells marks these transcription factors as key molecules in the regulation of oxidative stress-inducible apoptosis.

Catalase has been thought to play a central role in the protection of cells against oxidative stress by converting H₂O₂ to O₂ and H₂O. The observation that catalase is induced by H₂O₂ stimulation in bacterial cells (51) and in Schizosaccharomyces pombe (52) provides strong evidence for this hypothesis. However, in mammals, an objection can be raised based on observations that catalase is down-regulated by H₂O₂ (29), lipopolysaccharides [Ref. 37; which cause an oxidative stress by generating O₂⁻ and H₂O₂ (53)], and ionizing radiation (25), the cytotoxic effects of which are primarily mediated by the production of ROS. It may be hypothesized that the predominant role of catalase in multicellular organisms is not to protect cells against H₂O₂ but to regulate apoptosis by controlling cellular levels of H₂O₂. This hypothesis is supported by a large body of data demonstrating the antiapoptotic effects of catalase (3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15). Furthermore, the promoter structure of the mammalian catalase gene is quite distinct from that of yeast catalase in that it has no TATA box, multiple CCAAT boxes, GC boxes, and multiple transcription start sites (29, 54, 55, 56, 57). In this context, the down-regulation of catalase gene transcription in HP100 cells by X-rays provides a model system to study the regulation of apoptosis by antioxidant enzymes. Whether the onset of apoptosis can be modulated by artificially inhibiting the pathway for induction of the WT1/Egr-related factor after X-ray irradiation remains to be determined.

We thank Drs. M. Akashi and M. Hachiya (National Institute of Radiological Sciences, Chiba, Japan) for providing us with the HL60 and HP100 cell lines and for helpful suggestions on this work. We also thank Drs. H. Ishihara and T. Nakajima (National Institute of Radiological Sciences) for valuable advice and stimulating discussion.