The possible role of heat shock factor-1 in the negative
regulation of heme oxygenase-1
Yenn-Hwei Chou
a, Feng-Ming Ho
f, Der-Zen Liu
c, Shyr-Yi Lin
d, Li-Hsueh Tsai
e,
Chien-Ho Chen
b, Yuan-Soon Ho
b, Ling-Fang Hung
b, Yu-Chih Liang
b,∗aDepartment of Surgery, Shin Kong Wu Ho-Su Memorial Hospital, Taipei, Taiwan bGraduate Institute of Biomedical Technology, College of Medicine, Taipei Medical University,
No. 250 Wu-Hsing Street, Taipei 11014, Taiwan, R.O.C.
cGraduate Institutes of Biomedical Materials, College of Oral Medicine, Taipei Medical University, No. 250 Wu-Hsing Street, Taipei 11014, Taiwan, R.O.C.
dDepartments of Internal Medicine, College of Medicine, Taipei Medical University, Taipei, Taiwan eDepartments of Physiology, College of Medicine, Taipei Medical University, Taipei, Taiwan
fTao-Yuan General Hospital, Department of Health Taiwan, Taoyuan, Taiwan Received 23 February 2004; received in revised form 10 August 2004; accepted 20 August 2004
Abstract
We examined a possible role for heat shock factor-1 (HSF-1) in the negative regulation of HO-1 gene expression in human Hep3B hepatoma cells responding to stimulation with 15-deoxy-12,14-prostaglandin J2(15d-PGJ2) and arsenite. Overexpression of HSF-1 and heat-shock experiments indicated that HSF-1 repressed the 15d-PGJ2-and arsenite-induced HO-1 gene expression through directly binding to the consensus heat shock element (HSE) of the HO-1 gene promoter. In addition, point mutations at specific HSE sequences of the HO-1 promoter-driven luciferase plasmid (pGL2/hHO3.2-Luc) abolished the heat shock- and HSF-1-mediated repression of reporter activity. Overall, it is possible that HSF-1 negatively regulates HO-1 gene expression, and that the HSE present in the−389 to −362 region mediates HSF-1-induced repression of human HO-1 gene expression. © 2004 Elsevier Ltd. All rights reserved.
Keywords: Heme oxygenase-1; Heat shock factor-1; Hepatoma; Arsenite; 15-Deoxy-12,14-prostaglandin J2
1. Introduction
Heme oxygenase (HO) is a microsomal enzyme that degrades protoheme IX by cleaving its␣-methene
∗Corresponding author. Tel.: +886 2 27361661x3313–104;
fax: +886 2 27393447.
E-mail address: ycliang@tmu.edu.tw (Y.-C. Liang).
bridge into carbon monoxide (CO), free divalent iron, and biliverdin-IX␣ (Maines, 1997; Ponka, 1999). Three isoforms transcribed from separate genes have been characterized. HO-2 is a constitutively active form found mainly in the brain and testes (McCoubrey & Maines, 1994), and HO-3 has only recently been described, with properties similar to those of HO-2 (McCoubrey, Huang, & Maines, 1997). HO-1, known 1357-2725/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
as heat shock protein 32, is induced by stressors, in-cluding its substrate heme, cytokines, heavy metals, and oxygen free radicals (Elbirt, Whitmarsh, Davis, & Bonkosky, 1998;Keyse & Tyrrell, 1989;Lee & Chan, 2002; Vile, Basu-Modak, Waltner, & Tyrrell, 1994). Much evidence has suggested that up-regulation of HO-1 plays an important role in the cytoprotective de-fense response against oxidative stress and inflamma-tory stimuli (Choi & Alam, 1996; Rizzardini, Terao, Falciani, & Cantoni, 1993). High expression of HO-1 in the liver suggests that the HO/CO system might serve as a modulator of hepatobiliary function. For example, induction of HO-1 by heme resulting in in-creased delivery of heme to the liver occurred as a result of several processes including a splenectomy, hemol-ysis, and rhabdomyolysis (Rizzardini, Carelli, Cabello Porras, & Cantoni, 1994;Suematsu & Ishimura, 2000). The products of CO may function as a gaseous regula-tor of cytochrome P450-dependent biotransformation, such as bile acid synthesis and xenobiotic catabolism (Sano et al., 1997; Shinoda et al., 1998). In contrast to the beneficial roles, induction of HO-1 expression may be associated with endotoxic shock in vascular smooth muscle cells (Yet et al., 1997), and repression of HO-1 expression may represent a defense strategy de-veloped in humans (Shibahara, Nakayama, Kitamuro, Udono-Fujimori, & Takahashi, 2003).
Regulation of HO-1 expression has been exten-sively studied, and several cis-acting promoter ele-ments involved in its expression have also been elu-cidated (Lavrovsky, Schwartzman, Levere, Kappas, & Abraham, 1994;Lu et al., 1998; Takahashi et al., 1999). Specific sequences of a putative heat shock element (HSE) have been identified to be present in the HO-1 promoter. However, HSE seems to play a discrepant role in different species and cell types. A difference in the heat-mediated induction of HO-1 exists between rats and humans. In the rat, exposure of cells to el-evated temperatures causes a rapid increase in HO-1 expression, indicating that the HSE is a functional ele-ment in response to heat shock (Raju & Maines, 1994; Shibahara, Muller, & Taguchi, 1987). In contrast to the rat gene, the heat-mediated induction of human HO-1 seems to be observed only in certain cell lines (Keyse & Tyrrell, 1989;Mitani, Fujita, Sassa, & Kappas, 1990). The different expression patterns in various human cell lines caused by heat shock suggest that induction of HO-1 occurs in a cell line-dependent manner, and the
HSE seems to be a functional element of the human HO-1 gene.
Heat shock factor-1 (HSF-1) is known to have the unique ability to bind to the HSE in a heat shock-dependent manner. Upon treatment with stress inducers, such as heat shock, activated HSF-1 translo-cates to the nucleus, binds as trimers to multiple arrays of the HSE, which are located in the promoter region of genes, and then regulates gene expression (Mosser, Theodorakis, & Morimoto, 1988). The HSE has been found to be present in the promoter of Hsp genes and several non-Hsp genes, such as TNFα and IL-1β (Cahill, Waterman, Xie, Auron, & Calderwood, 1996; Singh, He, Calderwood, & Hasday, 2002). Recent studies indicated that HSF-1 negatively regulates TNF␣ and IL-1 expression for which the HSE might serve as a negative regulatory element. In human monocytic cells, HSF-1 bound to the IL-1 promoter and repressed its activity in a manner dependent upon the presence of an intact HSE (Cahill et al., 1996). Sev-eral studies have reported a negative regulatory region in human HO-1 (Deramaudt, da Silva, Remy, Kappas, & Abraham, 1999; Lu, Pepe, Gildemeister, Tyrrell, & Bonkovsky, 1997), and the HSE possibly playing a role as a negative regulatory element in the human HO-1 gene has not been well investigated as yet.
In this study, we investigated the possible role of HSF-1 in repression of HO-1 expression in human Hep3B hepatoma cells. Due to the discrepant roles of the HSE in the human HO-1 gene and HSE’s abil-ity to act as a negative regulatory element, we ex-amined the possibility that HSF-1 negatively regulates HO-1 expression through binding to the HSE. Our re-sults demonstrated that 15-deoxy-12,14-prostaglandin J2 (15d-PGJ2)- and arsenite-induced upregulation of human HO-1 expression could be repressed by heat shock or overexpression of HSF-1. In another exper-iment, we used a mutated HO-1 promoter construct with specific point mutations in the HSE consensus se-quences, which resisted repression by heat shock or the overexpression of HSF-1.
2. Materials and methods 2.1. Cell culture and treatments
Sodium arsenite and 15d-PGJ2were purchased from Biomol Research Laboratories (Plymouth Meeting,
PA) and Merck Taiwan Branch (Taipei, Taiwan), re-spectively. Human Hep3B hepatoma cells were cul-tured in MEM containing 10% heat-inactivated fetal bovine serum (Invitrogen, Carlsbad, CA) and 1 mM sodium pyruvate. For heat shock (HS) treatments, cells were placed in a bath for 1 h at 42.5◦C and then allowed to recover for different times at 37◦C as described for individual examinations.
2.2. Plasmids and transfection assays
The pGL2/hHO3.2-Luc reporter plasmid, con-taining a 3292-bp fragment,− 3106 to + 186 relative to the transcription start site of the human HO-1 gene, was amplified from the human BAC clone, CTA-286B10 (Kim et al., 1996), using synthesized primers (Table 1). The PCR products were inserted into the pCR2.1-TOPO cloning vector (Invitrogen), and cut with KpnI/XbaI such that the resulting PCR products contained the KpnI/XbaI sites; these were ligated in frame into the unique KpnI/NheI sites present within the pGL2 plasmid (Promega, Madison, WI). Therefore, we obtained a pGL2/hHO3.2-Luc reporter construct containing an approximately 3.2 kb region of the human HO-1 promoter driving luciferase gene expression. Site-directed mutagenesis of the consensus HSE sequence (−389 to −362) of the human HO-1 promoter was made by sequential poly-merase chain reaction steps from pGL2/hHO3.2-Luc. First, complementary oligonucleotides I and II were designed, and oligonucleotides III and IV (Table 1) were also synthesized for use as primers. Individual PCRs with primers I and IV as well as with primers II and III were performed using the pGL2/hHO3.2-Luc plasmid as a template. The amplification products were annealed with each other and extended by mutually primed synthesis. Fragments were then amplified by a second PCR step in the presence of primers III and IV. The products were then cut with ApaI and inserted into pGL2/hHO3.2-Luc, which had been cut with ApaI to eliminate the wild-type fragments.
The pcDNA3/HSF-1 expression plasmid, con-taining approximately 1.6 kb of full-length cDNA of human HSF-1, was amplified from the pHUHSF1 plasmid (Rabindran, Giorgi, Clos, & Wu, 1991) using the synthesized primers (Table 1). The PCR products were inserted into the pTARGET vector (Promega), cut with NheI/KpnI such that the resulting
PCR products contained the NheI/KpnI sites, and ligated in frame into the unique NheI/KpnI sites present within the pcDNA3.1 vector (Invitrogen). All sequences of wild-type or mutant pGL2/hHO3.2-Luc and pcDNA3/HSF-1 were confirmed and verified by DNA sequencing. For establishment of stable expression of HSF-1 cell lines, Hep3B cells were transfected with pcDNA3/HSF-1 and selected with G418 for 4 weeks. One HSF-1-overexpressing clone (H8) was chosen to examine HO-1’s expression.
For transfection, cells were seeded in 6 cm dishes (Western blot and RT-PCR) or 6-well plates (lu-ciferase assay) at a density of 6× 105and 2.5× 105, respectively. The next day, cells were transfected with either the wild-type or mutant pGL2/hHO3.2-Luc reporter plasmid, the pcDNA3/HSF-1 plasmid (when required), and phRL-TK (Promega) as the internal control plasmid using LipofetAMINE 2000TM (In-vitrogen). After transfection (9 h), the medium was replaced with complete medium, and incubated was continued for another 39 h. Transfected cells were then directly treated with drugs, subjected to heat shock, or subjected to heat shock prior to drug treatment for Western blot, RT-PCR, or luciferase assay. Luciferase assays were performed using a FireLiteTM dual luciferase assay kit (Packard Instrument Company, Meriden, CT) and a TopCount microplate scintillation and luminescence counter (Packard 9912v1) according to the manufacturer’s instructions. Luciferase activities of the reported plasmid were normalized to luciferase activities of the internal control plasmid (Liang, Tsai, Lin-Shiau, Chen, & Lin, 2001).
2.3. Western blot analysis
Equal amounts of total cellular proteins or nuclear proteins (50g) were resolved by SDS–polyacrylam-ide gel electrophoresis (PAGE), transferred onto an Im-mobilon-P membrane (Millipore, Bedford, MA), and blotted with HO-1, HO-2, Hsp70, or anti-HSF-1 (Santa Cruz Biotechnology, Santa Cruz, CA) antiserum as described previously (Liu et al., 2003).
2.4. RT-PCR and mRNA stability
Total RNA was isolated from both control and test cultured cells, and the mRNA level was
de-Table 1
Primer sequencesaand their applications in this study Construction of plasmids
Human HO-1 promoter reporter plasmid
Forward: 5-AGAGAACAGTTAGAAAAGAAAG-3 Reverse: 5-TACGGGCACAGGCAGGATCAGAA-3
Human HO-1 promoter reporter plasmid with a mutant HSE (−389 to −362) Oligonucleotide I: 5-GCCTCCCAGCTTTCTGGCACCTTATGGGACG-3 Oligonucleotide II: 5-CAGGCGTCCCATAAGGTGCCAGAAAGCTGGGA-3
Oligonucleotide III: 5-AATCACAGTATTGGGAAAGGACTGTATGA-3 Oligonucleotide IV: 5-CTTTATGTTTTTGGCGTCTTCCA-3
Human HSF-1 cDNA
Sense: 5-TCCTTGCTCGAGATGGATCTG-3 Antisense: 5-CTAGGAGACAGTGGGGTCCTT-3 RT-PCR
HO-1 (nucleotides 676–1097 of GenBank accession nos. XM009946) Sense: 5-TGATAGAAGAGGCCAAGA-3
Antisense: 5-TTTCCAGAGAGAGGGACA-3
HO-2 (nucleotides 338 to 693 of GenBank accession nos. NM002134) Sense: 5-TGGAGCGCAACAAGGACCAT-3
Antisense: 5-CCGGTAGAGCTGCTTGAACT-3 EMSA
hHO-1/HSEf: HSE (−1482 to −1455) of the human HO-1 gene promoter Forward: 5-TGGTTTTCGGGGAACCTTCAGAGGAAGAA-3 Reverse: 5-AGATTCTTCCTCTGAAGGTTCCCCGAAAA-3
Wt hHO-1/HSEn: wild HSE (−389 to −362) of the human HO-1 gene promoter Forward: 5-CAGCTTTCTGGAACCTTCTGGGACGCCT-3
Reverse: 5-AGAAGGCGTCCCAGAAGGTTCCAGAAAG-3,
Mut hHO-1/HSEn: mutant HSE (-389 to -362) of the human HO-1 gene promoter Forward: 5-CAGCTTTCTGGCACCTTATGGGACGCCT-3
Reverse: 5-AGAAGGCGTCCCATAAGGTGCCAGAAAG-3 Hsp70/HSE: HSE (−114 to −87) of the human Hsp70 gene promoter
Forward: 5-GAAACCCCTGGAATATTCCCGACCTGGC-3 Reverse: 5-AGAGCCAGGTCGGGAATATTCCAGGGGT-3
aThe underlining indicates HSE inverted repeats (CnnGAAnnTTCnnG). The italicized bases indicate a mutation to generate the mutated HSE probe or reporter.
tected by RT-PCR as described previously (Liang, Lin-Shiau, Chen, & Lin, 1999a). Two sets of primers were used to amplify the cDNA of 1 and HO-2 (Table 1). For measurement of mRNA stability, cells were treated with or without HS and then stim-ulated with 15d-PGJ2 for 6 h, followed by treat-ment with actinomycin D (4g/ml). Total RNA was prepared, and mRNA was measured as described above.
2.5. Electrophoretic mobility shift assay (EMSA) Nuclear extracts were prepared as described in our previous report (Liang et al., 1999b). The partially
com-plementary oligonucleotides were prepared (Table 1), and when required, filled in with [␣-32P] dCTP by the Klenow enzyme. For binding reactions, 10g of nuclear proteins was mixed with 3l of the labeled probe (approximately 5000 cpm) and 2g of poly(dI-dC)· poly(dI-dC) in a reaction buffer (10 mM Tris, pH 7.4, 50 mM NaCl, 5% glycerol, and 1 mM DTT) for 20 min at room temperature. To prove the specificity of binding, it was required that the reactions be carried out in the presence of a 50-fold excess of unlabeled wild-type or mutant HSE probe or in the presence of the anti-HSF-1 antibody. Complexes were separated by 4.5% native polyacrylamide gels and exposed to autoradiography.
3. Results
3.1. 15d-PGJ2and arsenite induce HO-1 expression
Our previous study found that 15d-PGJ2was a more effective activator than its precursor, 12-PGJ2 (Liu et al., 2004) in the induction of HO-1 expression. Therefore, 15d-PGJ2and another identified activator, arsenite (Keyse & Tyrrell, 1989), were used in this study. We examined the effects of 15d-PGJ2 and ar-senite on HO-1 expression in human Hep3B cells by Western blot analysis and RT-PCR. 15d-PGJ2and ar-senite increased the protein and mRNA expressions of HO-1 in concentration-dependent manners (Fig. 1). On the other hand, the protein and mRNA levels of HO-2 did not change. These results suggest that 15d-PGJ2 and arsenite induce HO-1 expression at the transcrip-tion levels.
3.2. Activation of HSF-1 down-regulates HO-1 expression
To investigate whether heat shock influences HO-1 expression induced by HO-15d-PGJ2and arsenite, cells were subjected to heat shock before being treated with 15d-PGJ2and arsenite. 15d-PGJ2and arsenite-induced
Fig. 1. 15d-PGJ2and arsenite induction of HO-1 expression. Cells were treated with 15d-PGJ2(5, 10, or 15M), or arsenite (1, 5, or 10M) for 14 (A) or 8 h (B), as indicated. Expressions of HO-1 and HO-2 were determined by Western blot (A) or by RT-PCR (B).
HO-1 protein expression; however, elevated tempera-tures resulted in divergent changes in the expression of HO-1 and Hsp70 proteins (Fig. 2A). Heat shock did induce Hsp70 expression, but down-regulated the induction of HO-1. In addition, 15d-PGJ2 alone in-creased Hsp70 protein expression. The time-course ex-periment indicated that both 15d-PGJ2 and arsenite-induced HO-1 expression during 8–20 h, with a peak at 14 h. However, the pre-addition of HS down-regulated the induction of HO-1, and the maximal expression also was at 14 h after treatment with 15d-PGJ2 and arsenite (Fig. 2B). Several lines of evidence have in-dicated that HSF-1 has the unique ability to bind to the HSE after heat shock. To provide direct evidence of whether HSF-1 is involved, we examined the effect of heat-independent HSF-1 overexpression on HO-1 expression. Cells were transfected with an HSF-1 ex-pression plasmid and then treated with 15d-PGJ2 or arsenite in the absence of heat shock. The induction of HO-1 by 15d-PGJ2and arsenite was down-regulated in cells transfected with the HSF-1 expression plas-mid by 40 and 30%, respectively (Fig. 2C). Transfec-tion of the HSF-1 expression plasmid alone resulted in slight activation of the Hsp70 protein, indicating that an observable amount of HSF-1 had been acti-vated. Transfection with the HSF-1 expression plas-mid or heat shock also down-regulated HO-1 mRNA expression induced by 15d-PGJ2or arsenite (Fig. 2D) similarly to the inhibition of HO-1 protein expres-sion (Fig. 2A and C). We, next investigated whether HS affected HO-1 mRNA transcript stability. Cells were pre-treated with or without HS and with 15d-PGJ2 to induce HO-1 mRNA expression. Transcrip-tion was then blocked with actinomycin D, and the half-lives of the HO-1 transcripts were assessed. The results showed that HS did not change the rate of HO-1 mRNA degradation (Fig. 2E). In additional experi-ments, we selected stable cell lines expressing HSF-1 to further examine the effect of HSF-1 on the down-regulation of HO-1 expression. Cells overexpressing HSF-1 (H8) had lower HO-1 expression than the par-ent cells in response to 15d-PGJ2(Fig. 2F, lanes 3 and 4). In addition, HO-1 expression was further decreased when HSF-1-overexpressing cells were subjected to heat shock (Fig. 2F, lanes 5 and 6). It is worthwhile noting that heat shock alone was insufficient to in-duce HO-1 expression under these experimental con-ditions.
Fig. 2. Down-regulation of HO-1 expression by activation of HSF-1. (A & B) Cells were undisturbed (37◦C) or subjected to heat shock (HS) before addition of 15d-PGJ2(10M) or arsenite (5 M). After 14 h (A) or indicated times (B), expressions of HO-1, HO-2, and/or Hsp70 were determined by Western blots. (C & D) Cells were transiently transfected with a mock or HSF-1 expression plasmid. Following transfection, cells were either directly treated with 15d-PGJ2(10M) or arsenite (5 M), or subjected to HS before the addition of 15d-PGJ2(10M) or arsenite (5M), as indicated. After 14 (C) or 8 h (D), expressions of HO-1, HO-2, or Hsp70 were determined by Western blots (C) or by RT-PCR (D). (E) Cells were undisturbed (37◦C) or subjected to heat shock (HS) before addition of 15d-PGJ2(10 M); after 6 h, cells were treated with actinomycin D (4g/ml) for 0–4 h. Cells were harvested, and total RNA was prepared. HO-1 mRNA levels were determined by RT-PCR normalized to GAPDH mRNA levels. Results are expressed relative to mRNA levels at 0 h, which were arbitrarily set to 100%. Values are expressed as the mean± S.E. of triplicate tests. (F) Stable cell lines expressing HSF-1 (H8) and mock expression cells (mock) were undisturbed (37◦C) or subjected to HS before the addition of 15d-PGJ2(10M). After 14 h, expressions of HO-1 and HSF-1 were determined by Western blots.
3.3. HSF-1 binds to HSE consensus sequences of the human HO-1 promoter
We attempted to confirm whether activated HSF-1 is translocated to the nucleus in HSF-15d-PGJ2- or heat shock-treated cells, and whether activated HSF-1 can bind to HSE located in the human HO-HSF-1 pro-moter. The first investigation was to detect if HSF-1 is present in the nucleus. 15d-PGJ2and heat shock in-creased HSF-1 translocation to the nucleus in a time-dependent manner, with the beginning increase seen at 45 min (Fig. 3A). Second, computer-assisted iden-tification of the putative HSE (set at a cutoff score of > 85%) was performed using MOTIF (available athttp://motif.genome.ad.jp), and we found the puta-tive HSE in the region of−1482 to −1455 and −389 to −362 of the 5-flanking region of human HO-1. Two putative HSE sequences (1/HSEf and hHO-1/HSEn) from the HO-1promoter and HSE sequences (Hsp70/HSE, as the positive control) from the Hsp70 promoter were selected as the EMSA probes (Table 1). No observable binding band was observed for nuclear extracts from 15d-PGJ2- or heat shock-treated cells us-ing the hHO-1/HSEf probe, indicatus-ing that the HSE se-quence present in the−1482 to −1455 region of the human HO-1promoter seemed to be a pseudo-HSE for HSF-1 binding (Fig. 3B). On the other hand, nuclear ex-tracts from 15d-PGJ2- or heat shock-treated cells bound the hHO-1/HSEn probe with a similar pattern as in the case of the Hsp70/HSE probe, indicating that the HSE sequence present in the−389 to −362 region of the hu-man HO-1promoter is a definite HSE for HSF-1 bind-ing (Fig. 3B). In a further examination to determine if binding to the hHO-1/HSEn probe was specific, it was not detected when the incubation was carried out in the presence of the anti-HSF-1 antibody or an excess of the wild-type HSE probe (Fig. 3C). On the contrary, an
Fig. 3. HSF-1 binding to HSE consensus sequences in the human HO-1 promoter. (A) Cells were subjected to heat shock (HS) or treated with 15d-PGJ2(10M) for the indicated times. Nuclear proteins were used to detect the level of HSF-1 by Western blots. (B) Gel shift assays used different HSE probes, as well as nuclear extracts from control cells, HS-, or 15d-PGJ2(10M)-treated cells. (C) To examine the specificity of binding to the HSE sequence of the human HO-1 promoter, the reaction was carried out in the absence of a competitor (lane 3), in the presence of anti-HSF-1 antibody (lane 4), or in the presence of a 50-fold excess of an unlabeled mutant HSEn (mut HSEn, lane 5) or a wild-type HSEn (wt HSEn, lane 6) probe. Nuclear extracts from cells treated with 15d-PGJ2(10M, lane 2) or exposed to HS (lanes 3–6) were used as a source of HSF-1. Positions of the complexes of HSF-1 and32P-labeled HSE probes are marked by arrows. The open arrowhead shows additional factors, but not HSF-1, bound to HSE. The sequences of HSE probes are described in Table 1. hHO-1/HSEf, HSE sequence (−1482 to −1455) far away from the transcription start site of human HO-1 gene; hHO-1/HSEn, HSE sequence (−389 to −362) near the transcription start site of the human HO-1 gene; Hsp70/HSE, HSE sequence (−114 to −87) from the human Hsp70 promoter.
excess of the mutant hHO-1/HSEn probe did not affect it.
3.4. Mutation of the HSE consensus sequences in the human HO-1promoter eliminated
HSF-1-mediated repression
To investigate the functional role of the HSE site as a transcriptional element which may repress HO-1 gene expression in response to HO-15d-PGJ2or arsen-ite, we chose the reporter construct, pGL2/hHO3.2-Luc, which contains a definite consensus HSE element located in the region from−389 to −362, upstream of the transcription start site. The mutant HSE of the pGL2/hHO3.2-Luc construct was also used to study HSE’s role in regulating the HO-1 gene, by making site-directed mutations in two of eight base pairs, as illustrated inFig. 4A.Figure 4Bshows the concentra-tion dependence of the inducconcentra-tion effect of 15d-PGJ2or arsenite on the reporter activity of pGL2/hHO3.2-Luc, with the maximum being reached for both at 10M. We next examined down-regulation of promoter activ-ity by overexpression of HSF-1 or by heat shock. Both heat shock and overexpression of HSF-1 significantly down-regulated the induction of wild-type reporter ac-tivity induced by 15d-PGJ2or arsenite (Fig. 4C). On the other hand, the reporter construct containing the mutated HSE, although strongly induced by 15d-PGJ2 or arsenite, was significantly resistant to repression by additional treatments of heat shock or overexpression of HSF-1 (Fig. 4D). These results demonstrate that re-pression of the HO-1promoter by heat shock or overex-pression of HSF-1 is mediated through the HSE. More-over, both 15d-PGJ2 and arsenite treatment resulted in similar levels of induction of wild-type and mutant pGL2/hHO3.2-Luc reporter activities (Fig. 4E). Treat-ment with 15d-PGJ2 resulted in 3.52- and 4.08-fold
Fig. 4. Elimination of HSF-1-mediated repression by mutation of HSE consensus sequences in the human HO-1 promoter. (A) The human HO-1 promoter-luciferase reporter construct (pGL2/hHO3.2-Luc) contains an HSE site in the region from−369 to −382. The site-directed mutant contains a difference of two-base changes indicated by underlining. Cells were transiently co-transfected with wild-type (B, C, E) or mutant (D, E) pGL2/hHO3.2-Luc reporter and mock expression (C, D) or the HSF-1 expression (C, D) plasmid. Following transfection, cells were undisturbed (37◦C) or subjected to heat shock (HS) before the addition of 15d-PGJ2(10M) or arsenite (5 M), as indicated. Total cell lysates were collected, and luciferase activities were determined. Values are expressed as the mean± S.E. of triplicate tests.*p < 0.05, compared with the individual test of mock expression alone (Student’s t-test). The indicated numbers were obtained from#the induction fold of 15d-PGJ2or arsenite treatment to their control, and##the induction fold of mutant to wild-type in control cells.
activations of the wild-type and mutant reporter ac-tivities, respectively. Arsenite also produced 4.65- and 5.11-fold activations in the wild-type and mutant re-porter, respectively. However, the mutant reporter had a high basal activity in the controls, and showed a 6.83-fold increase in comparison with the wild-type reporter (Fig. 4E, columns 1 and 2). These results suggest that basal expression of the HO-1 gene may be limited by an existing HSE complex, and therefore it is expressed at a low level.
4. Discussion
In this study, we demonstrate that the negative tran-scriptional regulation of the HO-1 gene involves an HSE conserved cis-acting sequence in human Hep3B hepatoma cells. Under normal conditions, low expres-sion of HO-1 was found in adult human liver cells. It is therefore possible that (1) no significant transcriptional activator induces its expression; or (2) certain constitu-tive HSE-binding factors (Liu, Kim, Yang, & Li, 1993) bind with HSE, and limit HO-1 expression at low levels. We found that mutations in the specific sequences of HSE resulted in high basal activity of the HO-1 reporter (Fig. 4E). In addition, gel shift assays indicated that control cells contain a HSE sequence-specific binding band (Fig. 3B, C, open arrowhead), which was com-peted for by excess unlabeled HSE oligonucleotides (Fig. 3C, lane 6). These results suggest that certain con-stitutive HSE-binding factors really exist and bind with the HSE site. As mention earlier, some human cell lines can increase HO-1 expression after heat shock, while other human cell lines cannot. It is possible that compo-nents of HSE complexes, including constitutive HSE-binding factors, HSF-1, and other unknown factors, de-termine whether HO-1 is expressed after heat shock. It is known that there are a number of genes, apart from the classical heat shock genes, whose expression can be down-regulated by HSF-1 through interacting with another transcription factor, such as NF-IL6 or STAT-1 (Stephanou, Isenberg, Nakajima, & Latchman, 1999; Xie, Chen, Stevenson, Auron, & Calderwood, 2002). The activated transcription factors in response to 15d-PGJ2or arsenite may interact with HSF-1 in combina-tion with heat shock, and ultimately repress the HO-1 expression. Therefore, our results suggest the possibil-ity that HSF-1’s partner determines the transactivation
potential, but determining which of the transcription factor joins with HSF-1 in regulating the HO-1 pro-moter requires further study.
Various extracellular stresses, such as oxidants and heavy metals, induce HO-1 expression, and also simultaneously activate HSF-1 (Ozaki, Deshpande, Angkeow, Suzuki, & Irani, 2000;Vilaboa et al., 1995). In this study, we found that 15d-PGJ2induced both HO-1 and Hsp70 expressions (Fig. 2), and also activated HSF-1 (Fig. 3). The activation of HSF-1 may play an inhibitory feedback role on excess HO-1 expression. In liver cells, repression of excess HO-1 expression may reduce energy expenditure consumed for heme metabolism and prevent the local accumulation of car-bon oxide, iron, and bilirubin. Shibahara et al. hypothe-sized that repression of HO-1 expression may represent a potential defense strategy in humans (Shibahara et al., 2003). The repression of HO-1 by HSF-1 may play an important role in the physiology of the liver, but this requires further study.
The ability of the precursor of 15d-PGJ2,12-PGJ2, to induce HO-1 expression in the rat has been exam-ined (Koizumi, Odani, Okuyama, Ichikawa, & Negishi, 1995). In human hepatoma cells, we screened a large number of lipophilic chemicals, including arachidonic acid and its metabolites, and found that 15d-PGJ2 in-creased HO-1 expression by about 5-fold higher than that of cells treated with12-PGJ2 at the same con-centration (Liu et al., 2004). Recently, 15d-PGJ2has been identified as a high-affinity ligand for peroxisome proliferator-activated receptor-␥ (PPAR␥). Activation of PPAR␥ by 15d-PGJ2has been considered to have anti-inflammatory actions, and the possible mechanism is linked to inhibition of multiple NFB activation steps (Straus et al., 2000). However, the repression is partly dependent on PPAR␥ expression (Chawla et al., 2001), and some of the receptor-independent biological ac-tion of 15d-PGJ2and other cyclopentenone PGs have been proposed (Rossi et al., 2000). We hypothesize that the anti-inflammatory effects of 15d-PGJ2in response to inflammatory stimuli may in part be mediated by the induction of an anti-inflammatory enzyme, HO-1. Lee and Chau (2002) demonstrated that the anti-inflammatory action of interleukin-10 was mediated by the induction of HO-1. In this study, we found that 15d-PGJ2strongly increased HO-1 expression in hu-man hepatoma cells. Although, the molecular mecha-nisms underlying the 15d-PGJ2-induced HO-1
expres-sion were not investigated in this study, the unpub-lished data suggest that induction of HO-1 expression by 15d-PGJ2did not mediate the activation of PPAR␣ or PPAR␥.
In conclusion, we provide evidence of a unique HSE sequence in the human HO-1 promoter, which binds HSF-1 and is required for HSF-1-mediated transcrip-tional repression. This is further proof of the concept that HSF-1 cannot repress gene transcription by using the mutant HSE sequence in the HO-1 reporter plasmid.
Acknowledgments
This work was supported by the National Science Council, Republic of China (NSC 93-2311-B-038-003-), Shin Kong Wu Ho-Su Memorial Hospital (SKH-TMU-92-05), and Seed funding from Taipei Medical University, Republic of China (TMU 91-Y05-A138). We are greatly indebted to Dr. H. Shizuya (Divi-sion of Biology, California Institute of Technology, Pasadena, CA) for providing the human BAC clone, CTA-286B10, and Dr. C. Wu (Department of Health and Human Services, Public Health Service, National Institutes of Health, Bethesda, MD) for providing the pHUHSF1 plasmid.
References
Cahill, C. M., Waterman, W. R., Xie, Y., Auron, P. E., & Calder-wood, S. K. (1996). Transcriptional repression of the prointer-leukin 1beta gene by heat shock factor 1. Journal of Biological Chemistry, 271, 24874–24879.
Chawla, A., Barak, Y., Nagy, L., Liao, D., Tontonoz, P., & Evans, R. M. (2001). PPAR-gamma dependent and independent effects on macrophage-gene expression in lipid metabolism and inflamma-tion. Nature Medicine, 7, 48–52.
Choi, A. M., & Alam, J. (1996). Heme oxygenase-1: function, reg-ulation, and implication of a novel stress-inducible protein in oxidant-induced lung injury. American Journal of Respiratory cell and Molecular Biology, 15, 9–19.
Csonka, C., Varga, E., Kovacs, P., Ferdinandy, P., Blasig, I. E., Szil-vassy, Z., et al. (1999). Heme oxygenase and cardiac function in ischemic/reperfused rat hearts. Free Radical Biology & Medicine, 27, 119–126.
Deramaudt, T. B., da Silva, J. L., Remy, P., Kappas, A., & Abraham, N. G. (1999). Negative regulation of human heme oxygenase in microvessel endothelial cells by dexamethasone. Proceeding of Society for Experimental Biology and Medicine, 222, 185–193.
Elbirt, K. K., Whitmarsh, A. J., Davis, R. J., & Bonkosky, H. L. (1998). Mechanism of sodium arsenite-mediated induction of heme oxygenase-1 in hepatoma cells. Role of mitogen-activated protein kinases. Journal of Biological Chemistry, 273, 8922–8931.
Keyse, S. M., & Tyrrell, R. M. (1989). Heme oxygenase is the major 32 kDa stress protein induced in human skin fibroblasts by UVA radiation, hydrogen peroxide, and sodium arsenite. Proceedings of the National Academy of Sciences of the United States of Amer-ica, 86, 99–103.
Kim, U. J., Birren, B. W., Slepak, T., Mancino, V., Boysen, C., Kang, H. L., et al. (1996). Construction and characterization of a human bacterial artificial chromosome library. Genomics, 34, 213–218. Koizumi, T., Odani, N., Okuyama, T., Ichikawa, A., & Negishi, M. (1995). Identification of a cis-regulatory element for delta 12-prostaglandin J2-induced expression of the rat heme oxygenase gene. Journal of Biological Chemistry, 270, 21779–21784. Lavrovsky, Y., Schwartzman, M. L., Levere, R. D., Kappas, A., &
Abraham, N. G. (1994). Identification of binding sites for tran-scription factors NF-kappa B and AP-2 in the promoter region of the human heme oxygenase 1 gene. Proceedings of the Na-tional Academy of Sciences of the United States of America, 91, 5987–5991.
Lee, T. S., & Chau, L. Y. (2002). Heme oxygenase-1 mediates the anti-inflammatory effect of interleukin-10 in mice. Nature Medicine, 8, 240–246.
Liang, Y. C., Lin-Shiau, S. Y., Chen, C. F., & Lin, J. K. (1999a). Inhibition of cyclin-dependent kinases 2 and 4 activities as well as induction of Cdk inhibitors p21 and p27 during growth arrest of human breast carcinoma cells by (−)-epigallocatechin-3-gallate. Journal of Cell Biochemistry, 75, 1–12.
Liang, Y. C., Huang, Y. T., Tsai, S. H., Lin-Shiau, S. Y., Chen, C. F., Lin, J. K., et al. (1999b). Suppression of inducible cy-clooxygenase and inducible nitric oxide synthase by apigenin and related flavonoids in mouse macrophages. Carcinogenesis, 20, 1945–1952.
Liang, Y. C., Tsai, D. C., Lin-Shiau, S. Y., Chen, C. F., & Lin, J. K. (2001). Suppression of inducible cyclooxygenase and inducible nitric oxide synthase through peroxisome proliferator-activated receptor-␥ by apigenin and related-flavonoids in macrophages. FEBS Letters, 496, 12–18.
Liu, J. D., Lin, S. Y., Ho, Y. S., Pan, S., Hung, L. F., Tsai, S. H., et al. (2003). Involvement of c-Jun N-terminal kinase activation in 15-deoxy-12,14-prostaglandin J2and prostaglandin A1-induced apoptosis in AGS gastric epithelial cells. Molecular Carcinogen-esis, 37, 16–24.
Liu, R. Y., Kim, D., Yang, S. H., & Li, G. C. (1993). Dual control of heat shock response: involvement of a constitutive heat shock element-binding factor. Proceedings of the National Academy of Sciences of the United States of America, 90, 3078–3082. Liu, J. D., Tsai, S. H., Lin, S. Y., Ho, Y. S., Hung, L. F., Pan, S.,
et al. (2004). Thiol antioxidant and thiol-reducing agents attenu-ate 15-deoxy-12,14-prostaglandin J2-induced heme oxygenase-1 expression. Life Science, 74, 245oxygenase-1–2463.
Lu, T. H., Lambrecht, R. W., Pepe, J., Shan, Y., Kim, T., & Bonkovsky, H. L. (1998). Molecular cloning, characterization, and expression of the chicken heme oxygenase-1 gene in
trans-fected primary cultures of chick embryo liver cells. Gene, 207, 177–1786.
Lu, T. H., Pepe, J. A., Gildemeister, O. S., Tyrrell, R. M., & Bonkovsky, H. L. (1997). Regulation of expression of the hu-man heme oxygenase-1 gene in transfected chick embryo liver cell cultures. Biochimica et Biophysics Acta, 1352, 293–302. Maines, M. D. (1997). The heme oxygenase system: a regulator of
second messenger gases. Annual Review of Pharmacology and Toxicology, 37, 517–554.
McCoubrey, W. K., Jr., & Maines, M. D. (1994). The structure, or-ganization and differential expression of the gene encoding rat heme oxygenase-2. Gene, 139, 155–156.
McCoubrey, W. K., Jr., Huang, T. J., & Maines, M. D. (1997). Iso-lation and characterization of a cDNA from the rat brain that encodes hemoprotein heme oxygenase-3. European Journal of Biochemistry, 247, 725–732.
Mitani, K., Fujita, H., Sassa, S., & Kappas, A. (1990). Activation of heme oxygenase and heat shock protein 70 genes by stress in human hepatoma cells. Biochemical and Biophysical Research Communications, 166, 1429–1434.
Mosser, D. D., Theodorakis, N. G., & Morimoto, R. I. (1988). Coor-dinate changes in heat shock element-binding activity and HSP70 gene transcription rates in human cells. Molecular and Cellular Biology, 8, 4736–4744.
Nath, K. A., Balla, G., Vercellotti, G. M., Balla, J., Jacob, H. S., Levitt, M. D., et al. (1992). Induction of heme oxygenase is a rapid, protective response in rhabdomyolysis in the rat. Journal of Clinical Investigation, 90, 267–270.
Nishizawa, J., Nakai, A., Matsuda, K., Komeda, M., Ban, T., & Na-gata, K. (1999). Reactive oxygen species play an important role in the activation of heat shock factor 1 in ischemic-reperfused heart. Circulation, 99, 934–941.
Ozaki, M., Deshpande, S. S., Angkeow, P., Suzuki, S., & Irani, K. (2000). Rac1 regulates stress-induced, redox-dependent heat shock factor activation. Journal of Biological Chemistry, 275, 35377–35383.
Ponka, P. (1999). Cell biology of heme. American Journal of Medical Science, 318, 241–256.
Rabindran, S. K., Giorgi, G., Clos, J., & Wu, C. (1991). Molecu-lar cloning and expression of a human heat shock factor HSF1. Proceedings of the National Academy of Sciences of the United States of America, 88, 6906–6910.
Raju, V. S., & Maines, M. D. (1994). Coordinated expression and mechanism of induction of HSP32 (heme oxygenase-1) mRNA by hyperthermia in rat organs. Biochimica et Biophysics Acta, 1217, 273–280.
Rizzardini, M., Carelli, M., Cabello Porras, M. R., & Cantoni, L. (1994). Mechanisms of endotoxin-induced haem oxygenase mRNA accumulation in mouse liver: synergism by glutathione depletion and protection by N-acetylcysteine. The Biochemical Journal, 304, 477–483.
Rizzardini, M., Terao, M., Falciani, F., & Cantoni, L. (1993). Cy-tokine induction of haem oxygenase mRNA in mouse liver. In-terleukin 1 transcriptionally activates the haem oxygenase gene. The Biochemical Journal, 290, 343–347.
Rossi, A., Kapahi, P., Natoli, G., Takahashi, T., Chen, Y., Karin, M., et al. (2000). Anti-inflammatory cyclopentenone prostaglandins
are direct inhibitors of IkappaB kinase. Nature, 403, 103– 108.
Sano, T., Shiomi, M., Wakabayashi, Y., Shinoda, Y., Goda, N., Yam-aguchi, T., et al. (1997). Endogenous carbon monoxide suppres-sion stimulates bile acid-dependent biliary transport in perfused rat liver. American Journal of Physiology, 272, G1268–G1275. Shibahara, S., Muller, R. M., & Taguchi, H. (1987). Transcriptional
control of rat heme oxygenase by heat shock. Journal of Biolog-ical Chemistry, 262, 12889–12892.
Shibahara, S., Nakayama, M., Kitamuro, T., Udono-Fujimori, R., & Takahashi, K. (2003). Repression of heme oxygenase-1 expres-sion as a defense strategy in humans. Experimental Biology and Medicine (Maywood), 228, 472–473.
Shinoda, Y., Suematsu, M., Wakabayashi, Y., Suzuki, T., Goda, N., Saito, S., et al. (1998). Carbon monoxide as a regulator of bile canalicular contractility in cultured rat hepatocytes. Hepatology, 28, 286–295.
Singh, I. S., He, J. R., Calderwood, S., & Hasday, J. D. (2002). A high affinity HSF-1 binding site in the 5’-untranslated region of the murine tumor necrosis factor-alpha gene is a transcriptional repressor. Journal of Biological Chemistry, 277, 4981–4988. Stephanou, A., Isenberg, D. A., Nakajima, K., & Latchman, D. S.
(1999). Signal transducer and activator of transcription-1 and heat shock factor-1 interact and activate the transcription of the Hsp-70 and Hsp-90beta gene promoters. Journal of Biological Chemistry, 274, 1723–1728.
Straus, D. S., Pascual, G., Li, M., Welch, J. S., Ricote, M., Hsiang, C. H., et al. (2000). 15-deoxy-delta 12, 14-prostaglandin J2 inhibits multiple steps in the NF-kappa B signaling pathway. Proceed-ings of the National Academy of Sciences of the United States of America, 97, 4844–4849.
Suematsu, M., & Ishimura, Y. (2000). The heme oxygenase-carbon monoxide system: a regulator of hepatobiliary function. Hepa-tology, 31, 3–6.
Takahashi, S., Takahashi, Y., Ito, K., Nagano, T., Shibahara, S., & Miura, T. (1999). Positive and negative regulation of the human heme oxygenase-1 gene expression in cultured cells. Biochimica et Biophysica Acta, 1447, 231–235.
Vilaboa, N. E., Calle, C., Perez, C., de Blas, E., Garcia-Bermejo, L., & Aller, P. (1995). cAMP increasing agents prevent the stimulation of heat-shock protein 70 (HSP70) gene expression by cadmium chloride in human myeloid cell lines. Journal of Cell Science, 108, 2877–2883.
Vile, G. F., Basu-Modak, S., Waltner, C., & Tyrrell, R. M. (1994). Heme oxygenase 1 mediates an adaptive response to oxida-tive stress in human skin fibroblasts. Proceedings of the Na-tional Academy of Sciences of the United States of America, 91, 2607–2610.
Xie, Y., Chen, C., Stevenson, M. A., Auron, P. E., & Calderwood, S. K. (2002). Heat shock factor 1 represses transcription of the IL-1beta gene through physical interaction with the nuclear factor of interleukin 6. Journal of Biological Chemistry, 277, 11802–11810.
Yet, S. F., Pellacani, A., Patterson, C., Tan, L., Folta, S. C., Foster, L., et al. (1997). Induction of heme oxygenase-1 expression in vascular smooth muscle cells. A link to endotoxic shock. Journal of Biological Chemistry, 272, 4295–4301.
