Regulation of Expression of the Rat Orthologue of Mouse Double Minute 2 (MDM2) by H₂O₂-induced Oxidative Stress in Neonatal Rat Cardiac Myocytes

Abstract

The Mdm2 ubiquitin ligase is an important regulator of p53 abundance and p53-dependent apoptosis. Mdm2 expression is frequently regulated by a p53 Mdm2 autoregulatory loop whereby p53 stimulates Mdm2 expression and hence its own degradation. Although extensively studied in cell lines, relatively little is known about Mdm2 expression in heart where oxidative stress (exacerbated during ischemia-reperfusion) is an important pro-apoptotic stimulus. We demonstrate that Mdm2 transcript and protein expression are induced by oxidative stress (0.2 mm H₂O₂) in neonatal rat cardiac myocytes. In other cells, constitutive Mdm2 expression is regulated by the P1 promoter (5′ to exon 1), with inducible expression regulated by the P2 promoter (in intron 1). In myocytes, H₂O₂ increased Mdm2 expression from the P2 promoter, which contains two p53-response elements (REs), one AP-1 RE, and two Ets REs. H₂O₂ did not detectably increase expression of p53 mRNA or protein but did increase expression of several AP-1 transcription factors. H₂O₂ increased binding of AP-1 proteins (c-Jun, JunB, JunD, c-Fos, FosB, and Fra-1) to an Mdm2 AP-1 oligodeoxynucleotide probe, and chromatin immunoprecipitation assays showed it increased binding of c-Jun or JunB to the P2 AP-1 RE. Finally, antisense oligonucleotide-mediated reduction of H₂O₂-induced Mdm2 expression increased caspase 3 activation. Thus, increased Mdm2 expression is associated with transactivation at the P2 AP-1 RE (rather than the p53 or Ets REs), and Mdm2 induction potentially represents a cardioprotective response to oxidative stress.

Exposure of cardiac myocytes to sufficiently high levels of reactive oxygen species (ROS)⁶ such as H₂O₂ leads to their death (1–3), and this probably involves a continuum from apoptosis to necrosis, depending on the severity of the oxidative stress (4). In aerobic tissues such as the heart, the mitochondria probably represent a significant source of ROS, and increased ROS production by these organelles during hypoxia and ischemia-reperfusion injury may be particularly important in myocardial injury (5, 6). However, at lower concentrations, ROS have been reported to promote either growth of the cardiac myocyte (7) or to induce “preconditioning” (8), either of which potentially increases the ability of the cardiac myocyte to survive cytotoxic stresses. H₂O₂-induced oxidative stress simultaneously stimulates a number of potentially pro-apoptotic and cytoprotective signaling pathways in the whole heart or cardiac myocytes (9), and the final outcome (cell death or survival) could depend on which signaling pathway(s) predominates and endures.

As shown by our microarray studies, H₂O₂ can positively and negatively regulate global gene expression in cardiac myocytes (3, 10). One gene consistently up-regulated by H₂O₂ in rat cardiac myocytes at toxic and nontoxic concentrations is the orthologue of transformed mouse 3T3 cell double minute 2 (Mdm2) (3, 10), a proto-oncogene (11, 12), to which we will refer as Rdm2.⁷ The human orthologue will be abbreviated as HDM2. The Mdm2 protein binds to the pro-apoptotic p53 tumor suppressor transcription factor to inhibit its transactivating activity (13). Perhaps more importantly, Mdm2 is a ubiquitin-protein isopeptide ligase that ubiquitinates p53 and other proteins (13), thus promoting their proteasomal degradation. Indeed, the stability of p53 protein appears to be of major importance in controlling its abundance (13). In addition, Mdm2 may autoubiquitinate to promote its own degradation (14, 15).

Regulation of Mdm2 expression is complex and involves two alternative promoters. The P1 promoter lies 5′ to exon 1 and the P2 promoter lies within intron 1 (16). P1 primarily regulates constitutive expression of Mdm2 with P2 contributing to a lesser extent. Inducible expression of Mdm2 is regulated by P2, which contains two p53-response elements, and through these p53 itself regulates stress-induced expression of Mdm2 as part of a negative feedback loop (16). Transcripts from the P1 promoter may be translated more slowly than those from P2 because of the presence of short (but differing) upstream open reading frames (two in each) in exon 1 of both HDM2 and Mdm2 transcripts (17, 18). However, P2 contains additional response elements (AP-1, Ets) that may allow induction of Mdm2 expression in a p53-independent manner (19–21). The P1- and the P2-regulated transcripts can each give rise to two identical Mdm2 proteins with translation from the initiation codon in exon 3 encoding an Mdm2 species that migrates at about 90 kDa on SDS-PAGE. A truncated 76-kDa species can also be expressed from a second in-frame initiation codon in exon 4 either following removal of exon 3 by alternative splicing or by internal ribosomal entry (22, 23). Expression of p76 Mdm2 protein is more favored with transcription from P1 than from P2, without there being any change in absolute p90 Mdm2 protein expression (23). p76 Mdm2 cannot bind to p53 or inhibit its transactivating activity, and it may in fact inhibit the actions of p90 Mdm2 (22). Here we characterize the Rdm2 gene and the expression of Rdm2 mRNA and protein in response to exposure to ROS (H₂O₂) in neonatal rat cardiac myocytes. We demonstrate that up-regulation of Rdm2 expression by ROS is largely independent of p53 but correlates instead with transcriptional activation at the AP-1-response element. Furthermore, Rdm2 is cytoprotective against ROS-induced apoptosis (caspase 3 cleavage) in cardiac myocytes.

EXPERIMENTAL PROCEDURES

Materials

H₂O₂ was standardized volumetrically before use. It was stable at 4 °C for at least 6 months. Antibodies against Rdm2 (sc-965, a mouse monoclonal antibody raised to residues 154–167 of HDM2 corresponding to residues 169–173 in Rdm2 (supplemental Fig. 1) in which region Rdm2 and HDM2 are identical in primary sequence), c-Jun (sc-45X, for supershifting), JunB (sc-46X), JunD (sc-74X), phospho-Elk-1(Ser-383) (sc-7979X), Fli-1 (sc-356X), c-Fos (sc-52X), FosB (sc-48X), Fra-1 (sc-605X), Fra-2 (sc-171X), and ATF2 (sc-187X) were from Santa Cruz Biotechnology, Inc. Antibodies against c-Jun (9162, used for immunoblotting) and 17-kDa caspase-3 (9661) were from Cell Signaling Technology. Sarcomeric α-actin antibodies (A2172) were from Sigma, and appropriate horseradish peroxidase-linked secondary antibodies were from DAKO.

Six Santa Cruz Biotechnology, Inc., anti-p53 antibodies were tested. These were 2B2.71 (sc-71819, a mouse monoclonal antibody IgG_2b raised against residues 370–378 of human p53), DO-1 (sc-126, a mouse monoclonal antibody IgG_2a raised against residues 11–25 of human p53), FL-393 (sc-6243, a rabbit polyclonal antibody IgG raised against full-length human p53), Pab 240 (sc-99, a mouse monoclonal antibody IgG₁ raised against residues 156–214 of mammalian p53), Pab 246 (sc-100, a mouse monoclonal antibody IgG₁ raised against residues 88–93 of mouse p53), and R-19 (sc-1313, a goat polyclonal antibody IgG raised against an epitope mapping to the C terminus of rat p53). With the exception of DO-1, which is not explicitly specified by the supplier as being able to recognize rat p53, all other antibodies are supposed to recognize rat p53 on immunoblotting. As negative and positive controls, we used lysates from nontransfected 293T cells (sc-117752) and from 293T cells transfected with mouse p53 (sc-125766), respectively, and extracts of daunorubicin-treated cardiac myocytes.

The p53 electrophoretic mobility shift assay (EMSA) oligodeoxynucleotide probe was from Santa Cruz Biotechnology, Inc. (sc-2579). Other probes were synthesized by Sigma Genosys (Sigma). Phosphorothioate-protected ODNs and standard oligonucleotides were from MWG Biotec. Lipofectin was from Invitrogen. SMART RACE cDNA amplification kits were from Clontech.

Primary Culture of Neonatal Rat Cardiac Myocytes

Myocytes (yield of 3.5–4.5 million cells/heart) were dissociated from the ventricles of 1–3-day-old Sprague-Dawley rat hearts by an adaptation of the method of Iwaki et al. (24) as described previously (25). Unless stated otherwise, cells were plated in gelatin-coated Primaria culture dishes (BD Biosciences) at a density of 2 × 10⁶ cells/35-mm dish or 4 × 10⁶ cells/60-mm dish for 18 h in 15% (v/v) fetal calf serum. For most experiments, serum was withdrawn for the 24 h before experimentation, and the cells were used 2 days after their isolation.

Extraction of RNA, cDNA Synthesis, and Semiquantitative or Quantitative PCR (RT-sqPCR or RT-qPCR)

RNA (for which quantity and purity were assessed from the A₂₆₀ and the A₂₆₀/A₂₈₀ ratio (1.8–2.0 deemed acceptable)) was isolated from cardiac myocytes with RNA-Bee (GE Healthcare) according to the manufacturer's instructions; cDNA was synthesized with random hexamers, and RT-sqPCR (see Table 1 for details of primers) measurements were performed as described in detail previously (3, 10). To ensure that there was no amplification of genomic DNA, primers were designed to cross exon boundaries where possible (Table 1), and controls in which the RT step was omitted were carried out. The resulting RT-sqPCR products were analyzed by ethidium bromide-agarose gel electrophoresis and the bands were captured under UV illumination. Sizes of PCR products were estimated by comparison with a ϕX174 RF DNA HaeIII digest DNA ladder or a 100-bp DNA ladder (Invitrogen). Densitometric analysis was performed using ImageMaster 1D Prime, version 3.0 (GE Healthcare). Values were normalized to the housekeeping gene Gapd. PCR cycles used (Table 1) were within the exponential phase of amplification (confirmed by construction of amplification curves from PCR over a range of cycle numbers). Sequences of PCR amplicons were determined using an ABI 3100 genetic analyzer by the Advanced Biotechnology Centre (Imperial College London).

TABLE 1.

Details of RT-sqPCR/qPCR primers used

For the Rdm2 transcript, the numbering used refers to the deduced sequence shown in supplemental Fig. 2. For Rdm2, primers for Rdm2(A) were used for sqPCR analysis of general Rdm2 expression; primers for Rdm2(B) were used for qPCR analysis of general Rdm2 expression; primers for Rdm2(C) and Rdm2(D) were used to investigate P1/P2 promoter usage (sqPCR); primers for Rdm2(E) were used for 5′-RACE (sqPCR); and primers for Rdm2(F) were used for ChIP analysis (sqPCR). The sequence of Rdm2 intron 1, in which the primers used for chromatin immunoprecipitation lie, is shown in Fig. 1. The predicted rat FosB transcript sequence (NM_001013146.1) is truncated at its 3′-end. The rat FosB gene is located on chromosome 1 at 78668523..78673784 (RGSC genome assembly version 3.4). The RT-sqPCR product corresponds to 78667636..78667773, sequence: ACCCCTGTGCAGTATTATGCCATGTCCCTCTCACCCCCACGGGCAACCCAGGCGCCCTTGGCCGTCCTCGTTGGGCCTTTCTGGTTTTGGGCAGCAGGGGGCGCTGCGACGCCCGTCTTGCTGGAGCGCTTTATACTG (primers underlined and boldface). In contrast, the mouse FosB mRNA sequence (NM_008036.2) is more extended and contains the PCR sequence described. The PCR sequence identifies two rat ESTs (C06882.1 and CB747983.1) and these lie 3′ to the predicted rat FosB transcript. In addition, the two rat ESTs correspond to the 3′-untranslated region of the mouse FosB mRNA. NA indicates not applicable.

For RT-qPCR, cDNA was prepared from 1 μg of RNA using high capacity reverse transcription kits (Applied Biosystems). Primers for qPCR were designed with Primer Express version 3.0 using deduced Rdm2 sequences (supplemental Fig. 2; Table 1). qPCR was performed using a 7500 real time PCR system (Applied Biosystems). A master mix containing (per reaction) 12.5 μl of Sybr-Green Jump Start Taq Readymix (Sigma) and 5 μl of oligonucleotides (200 mm forward and reverse primers) was dispensed into Optical 96-well reaction plates (Applied Biosystems), and cDNA template was added (7.5 μl, 1:15 dilution in water). PCR conditions for all primer pairs were 50 °C for 2 min, 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s and 60 °C for 60 s. Following qPCR, dissociation curve analysis was performed to check for aberrant amplification products. An absolute quantification protocol was used. Values were normalized to Gapd expression.

5′-RACE

Poly(A)⁺ RNA was purified from total RNA (150 μg) using oligo(dT)-cellulose (Sigma). Briefly, total RNA (in 600 μl of deionized water) was incubated for 5 min at 65 °C followed by addition of 1 volume of 2× binding buffer (20 mm Tris-HCl, pH 7.5, 1 m NaCl, 2 mm EDTA, 0.1% (w/v) SDS)). Oligo(dT)-cellulose (5 mg in 100 μl of binding buffer) was added followed by rotation at room temperature for 15 min. The oligo(dT)-cellulose was washed twice with binding buffer (1 ml) and twice with washing buffer (10 mm Tris-HCl, pH 7.5, 0.2 m NaCl, 1 mm EDTA, 0.05% (w/v) SDS). Poly(A)⁺ RNA was recovered in 250 μl of elution buffer (10 mm Tris-HCl, pH 7.5, 1 mm EDTA, 0.05% (w/v) SDS; two incubations of 5 min at 37 °C). The purification procedure was repeated, and poly(A)⁺ RNA was precipitated with 50 μl of 4 m NaCl and 2 volumes of ethanol followed by centrifugation (12,000 × g, 15 min, 4 °C). After washing with 70% (v/v) ethanol and centrifugation, poly(A)⁺ RNA was dissolved in deionized water (8 μl). Poly(A)⁺ RNA was reverse-transcribed using a SMART RACE cDNA amplification kit (Clontech). Poly(A)⁺ RNA (0.9 μg) was mixed with oligo(dT) primer (10 mm 5′-CDS primer A, 1 μl), 10 mm SMART II A oligonucleotide (10 mm, 1 μl) in a total volume of 5 μl. This was heated at 72 °C for 2 min and cooled on ice. Subsequently, 5× reaction buffer (2 μl; 250 mm Tris-HCl, pH 8.3, 30 mm MgCl₂, 375 mm KCl), 10 mm dNTP (1 μl), 20 mm DTT (1 μl), and PowerScript^TM reverse transcriptase (1 μl) were added, and the mixture was incubated at 42 °C for 90 min. Reverse transcriptase was inactivated at 72 °C for 7 min. The 5′-RACE cDNA reaction mixture (10 μl) was diluted to 250 μl with 10 mm Tricine, pH 8.5, 1 mm EDTA and stored at −80 °C. Sequence corresponding to 5′-end of Rdm2 mRNA was amplified using the SMART RACE cDNA amplification kit Universal Primer A Mix (complementary for the extended 5′-end of the 5′-RACE cDNA) and a reverse primer to mRNA in exon 3 of Rdm2 exon 3 (bases 269–302, see Table 1). For each PCR, 5′-RACE cDNA (2.5 μl) was combined with 10דAdvantage” polymerase buffer (10 μl; 40 mm Tricine/KOH, pH 9.2, 15 mm KOAc, 3.5 mm Mg(OAc)₂), 5′-RACE 1 primer (1 μl, 10 μm), 10× Universal Primer A Mix (5 μl), dNTP (10 mm), and Advantage 2 polymerase buffer (1 μl). The PCR program was as follows: 94 °C for 5 min; 5 cycles of 94 °C for 30 s and 72 °C for 2 min; 5 cycles of 94 °C for 30 s, 70 °C for 30 s, and 72 °C for 2 min; then 25 cycles of 94 °C for 30 s, 68 °C for 30 s, and 72 °C for 2 min), and finally 72 °C for 4 min. PCR products were analyzed on ethidium bromide-stained 2% agarose gels and the bands were captured under UV illumination. Products were excised from the gel and then purified using QIAquick PCR purification columns for automated sequencing (Advanced Biotechnology Centre, Imperial College London).

Soluble and Nuclear Protein Extraction

For immunoblotting, myocytes were washed in ice-cold PBS (three times) and scraped into 100 μl (35-mm dishes) or (in the case of 60-mm dishes) into 80 μl (antisense experiments) or 150 μl (remaining experiments) of ice-cold buffer A (20 mm β-glycerophosphate pH 7.5, 20 mm NaF, 2 mm EDTA, 0.2 mm Na₃VO₄, 10 mm benzamidine, 5 mm DTT, 1 mm PMSF, 0.2 mm leupeptin, 2 μm microcystin LR, 10 μm E64, 1% (v/v) Triton X-100). After centrifugation (10,000 × g, 5 min, 4 °C), supernatant protein concentrations were determined using the Bradford method (26), boiled with 0.33 volume of SDS electrophoresis buffer (300 mm Tris-HCl, pH 6.8, 10% (w/v) SDS, 13% (w/v) glycerol, 130 mm DTT, 0.2% (w/v) bromphenol blue), and stored at −20 °C.

For crude nuclear extracts, PBS-washed myocytes were scraped into 150 μl of ice-cold buffer B (10 mm HEPES, pH 7.9, 10 mm KCl, 1.5 mm MgCl₂, 0.3 mm Na₃VO₄, 0.2 mm leupeptin, 10 μm E64, 5 mm DTT, 0.3 mm PMSF, 2 μm microcystin LR). Extracts were centrifuged (10,000 × g, 5 min, 4 °C), and the pellets resuspended in 100 μl of Buffer B containing 0.1% (v/v) Nonidet P-40. Samples were centrifuged (10,000 × g, 5 min, 4 °C), and the pellets were resuspended in 50 μl of ice-cold buffer C (20 mm HEPES, pH 7.9, 420 mm NaCl, 1.5 mm MgCl₂, 0.2 mm EDTA, 25% (v/v) glycerol, 0.3 mm Na₃VO₄, 0.2 mm leupeptin, 10 μm E64, 5 mm DTT, 0.3 mm PMSF, 2 μm microcystin LR). Suspensions were incubated on ice (1 h) with mixing every 15 min and then centrifuged (10,000 × g, 5 min, 4 °C). Supernatants were retained and stored at −80 °C or were boiled with SDS electrophoresis buffer.

Immunoblotting

Proteins were separated by SDS-PAGE using between 8 and 12% (w/v) resolving gels as appropriate with 4–6% (w/v) stacking gels and transferred to nitrocellulose membranes as described previously (27). Gels were standardized using “rainbow” markers (GE Healthcare). Nonspecific binding sites were blocked (60 min, room temperature) with 5% (w/v) nonfat milk powder in TBST (20 mm Tris-HCl, pH 7.5, 137 mm NaCl, 0.1% (v/v) Tween 20). Blots were incubated overnight with primary antibodies diluted in TBST containing 5% (w/v) bovine serum albumin. Membranes were washed in TBST (three times for 5 min, room temperature) and incubated (1 h, room temperature) with horseradish peroxidase-conjugated secondary antibodies in TBST containing 1% (w/v) nonfat milk powder. After washing in TBST (three times for 5 min, room temperature), bands were detected by the enhanced chemiluminescence method (reagents from Santa Cruz Biotechnology, Inc.). In some cases the sensitivity of detection was increased by using unconjugated secondary antibodies in TBST, 1% (w/v) nonfat milk powder, and after washing with TSBT, blots were incubated with horseradish peroxidase-conjugated tertiary antibodies before detection by enhanced chemiluminescence.

Antisense ODN Transfection

The 20-mer antisense phosphothioate-protected ODN, 5′-TGACATCTGCTCTCACTCGG-3′, was directed against Rdm2 mRNA exon 7, bases 602–621 (supplemental Fig. 2). A mixed backbone HDM2 antisense ODN (5′-UGACACCTGTTCTCACUCAC-3′) against the corresponding region (HDM2 mRNA bases 660–679, accession number NM_002392.2) has been used previously to inhibit HDM2 expression in colon cancer cells (28). A random sequence (5′-TGGATCCGACATGTCAGA-3′) was used as a control. For transfection experiments, cells in 60-mm dishes were washed three times with antibiotic-free Dulbecco's modified Eagle's medium before exposure to Dulbecco's modified Eagle's medium containing 200 nm ODN and 20 μg/ml Lipofectin (Invitrogen). Myocytes were incubated for 24 h at 37 °C and then the medium was then replaced with Lipofectin-free maintenance medium containing 200 nm ODN. Cells were preincubated for 1 h before exposure to 0.2 mm H₂O₂ for 6 h and extracted into 80 μl of buffer A (see above).

EMSAs

Double-stranded ODN EMSA probes (see Table 2) were prepared essentially as described previously (29). Briefly, ODNs were 5′-labeled with 15 μCi of [γ-³²P]ATP (GE Healthcare), 10 IU of T4 polynucleotide kinase (Promega) (12 μl final volume), and 1.2 μl of Promega 10× polynucleotide kinase buffer. The reaction mixture was incubated (37 °C, 30 min), and the reaction was then terminated by addition of 1 μl of EDTA (0.25 m). Excess [γ-³²P]ATP was removed by centrifugation (700 × g, 3 min, room temperature) through Chroma Spin columns (pore size 10) (Clontech) in 1× TE buffer (10 mm Tris-HCl, pH 8.0, 1 mm EDTA), and the EMSA probes were diluted to 0.4 ng/μl. Crude nuclear extracts (10 μg) were incubated (10 min, room temperature) with 4 μl of binding buffer (50 mm Tris-HCl, pH 7.5, 250 mm NaCl, 5 mm EDTA, 1 mm DTT, 25% (v/v) glycerol) and 1 μg of poly(dI-dC) (GE Healthcare) in a total volume of 20 μl. Radiolabeled EMSA probe (1 μl, 0.4 ng) was added, and the mixture was incubated (20 min, room temperature). Samples were subjected to electrophoresis (120 V, 3 h) on 5% (w/v) nondenaturing polyacrylamide gels formed in 1× TBE buffer (89 mm Tris, 89 mm boric acid, 2 mm EDTA, pH 8.3) with 0.5× TBE as running buffer. Gels were dried onto Whatman 3MM paper and exposed to x-ray film for 24–48 h at −80 °C. Images were analyzed semi-quantitatively by densitometry (ImageMaster 1D Prime, version 3.0, GE Healthcare). For competition experiments, crude nuclear extracts were incubated (10 min) with the required amount of unlabeled competing probe in binding buffer with poly(dI-dC) as in the standard assay, followed by addition of the radiolabeled wild type probe and continuation of the assay in the standard manner. For supershift assays, crude nuclear extracts (3 μg protein) were preincubated (1 h, 4 °C) with 1–2 μg of antibody.

TABLE 2.

Sequences of EMSA probes

See Fig. 1 for details of position in the Rdm2 gene and its intron 1. For p53 EMSAs, a consensus p53 probe was used. NA indicates not applicable.

ChIP Assays

Myocytes were exposed to H₂O₂ (0.2 mm, 2 h), and formaldehyde (final concentration of 1% (w/v)) was added to medium (10 min, room temperature). Cross-linking was terminated with 0.125 m glycine (5 min). Myocytes were rinsed twice in ice-cold PBS and scraped into 200 μl of PBS containing 0.2 mm leupeptin, 10 μm E64, 5 mm DTT, 1 mm PMSF, and 2 μm microcystin LR. Suspensions were centrifuged (4000 × g, 10 min, 4 °C) and pellets resuspended in 0.8 ml of ChIP lysis buffer (50 mm Tris-HCl, pH 8.0, 85 mm KCl, 0.2 mm leupeptin, 10 μm E64, 5 mm DTT, 0.3 mm PMSF, 2 μm microcystin LR, 0.5% (v/v) Nonidet P-40). Suspensions were incubated on ice (10 min) and centrifuged (4000 × g, 10 min, 4 °C), and the pellets were resuspended in 0.8 ml of sonication buffer (50 mm Tris-HCl, pH 8.0, 10 mm EDTA, 1% (v/v) SDS, 0.2 mm leupeptin, 10 μm E64, 5 mm DTT, 1 mm PMSF, and 2 μm microcystin LR). Samples were sonicated (five times for 30 s, 0 °C, amplitude 30% (Sonics Vibra-Cell^TM sonicator with a 2-mm probe) followed after each sonication by 2 min of recovery in ice-water) to shear DNA into fragments of 200–800 bp. Sonicated samples were centrifuged (4000 × g, 10 min, 4 °C), and the supernatants were retained. Samples (20 μl) for total DNA input were taken at this stage. Sonicated extracts (750 μl) were diluted with 3.65 ml of ChIP buffer (20 mm Tris-HCl, pH 8.0, 150 mm NaCl, 2 mm EDTA, 0.01% (w/v) SDS, 1% (v/v) Triton X-100) and were precleared with 30 μl of 50% protein A-Sepharose slurry containing 5 μg/ml sonicated salmon sperm DNA (rotation at 4 °C for 30 min). Following centrifugation (200 × g, 1 min, 4 °C), anti-transcription factor antibodies (5 μg) were added followed by incubation (4 °C, 16 h, with mixing by rotation). Antibody-protein-DNA complexes were recovered by incubation (1 h, 4 °C, with mixing by rotation) with 50% protein A-Sepharose slurry (80 μl containing 5 μg/ml sonicated salmon sperm DNA). Protein A-Sepharose-antibody-transcription factor complexes were recovered by centrifugation (200 × g, 1 min, 4 °C), and pellets were washed in low salt buffer (20 mm Tris-HCl, pH 8.0, 150 mm NaCl, 2 mm EDTA, 0.1% (w/v) SDS, 1% (v/v) Triton X-100, 1 ml), high salt buffer (20 mm Tris-HCl, pH 8.0, 500 mm NaCl, 2 mm EDTA, 0.1% (w/v) SDS, 1% (v/v) Triton X-100, 1 ml), and LiCl buffer (10 mm Tris-HCl, pH 8.0, 250 mm LiCl, 1 mm EDTA, 1% (w/v) deoxycholic acid, 1% (v/v) Nonidet P-40) and then twice in 10 mm Tris-HCl, pH 8.0, 1 mm EDTA. Protein-DNA complexes were recovered by incubation (15 min, 65 °C) with 0.1 m NaHCO₃, 1% (w/v) SDS (250 μl), and samples were centrifuged (200 × g, 1 min, 4 °C) and supernatants retained. This elution step was repeated, and supernatants were combined. Cross-linking was reversed by incubation (16 h, 65 °C) with 0.2 m NaCl. DNA was extracted by mixing equal volumes of samples with phenol/chloroform/isoamyl alcohol (25:24:1, pH 8.0) and with separation of the phases by centrifugation (15,000 × g, 10 min, 4 °C). The upper aqueous phases were retained, and DNA was precipitated by adding 1 volume of isopropyl alcohol for 1 h on ice. Following recovery of DNA by centrifugation (21,000 × g, 10 min, 4 °C), DNA was washed (70% (v/v) ethanol) and resuspended in 20 μl of Milli-Q water for subsequent PCR. PCRs were performed in 1× Buffer IV® (AB Gene, 25 μl) containing Taq polymerase (1 unit), dATP, dCTP, dGTP, and dTTP (0.2 mm each) and 50 pmol of primers. PCR settings were as follows: 94 °C for 4 min; 35 cycles of (94 °C for 30 s; 59 °C for 30 s; 72 °C for 30 s), followed by 72 °C for 4 min. The resulting RT-PCR products were analyzed by ethidium bromide-agarose gel electrophoresis, and the bands were captured under UV illumination.

Statistical Analysis

Results are expressed as means ± S.E. Statistical significance was determined by either Student's t test or one-way analysis of variance (ANOVA) followed by Tukey's multiple comparison test, as appropriate. p values of <0.05 were considered significant.

RESULTS

Structure of the Rdm2 Gene

Mdm2 (NP_034916.1, 489 residues) and HDM2 (NP_002383.2, 497 residues) are encoded by established genes and are highly conserved, sharing precise identity in about 80% of residues (supplemental Fig. 1). In contrast, Rdm2 is a predicted gene on chromosome 7(q22) in the NCBI data base (www.ncbi.nlm.nih.gov) whose provisional processed transcript NM_001108099.1 (December 21, 2008) encodes a protein that differs significantly from Mdm2 and HDM2. A BLAST search of the mouse Mdm2 transcript NM_010786.3 against the rat genome identified a 24,233-bp sequence on the minus strand of contig NW_047774 corresponding to a 2930-nucleotide transcript from 12 exons with high homology to the Mdm2 transcript (supplemental Fig. 2). The deduced 2930-nucleotide sequence of the full-length transcript is identical to the 2855-nucleotide NM_001108099.1 except that it contains an extra exon, the 75-nucleotide (protein-coding) exon 4. A BLAST search of this exon revealed 14 rat ESTs with a region of precise sequence identity within them, thus establishing that exon 4 is transcribed and is present in mature Rdm2 mRNA. The structural relationship of the first four exons in Rdm2 is shown in Fig. 1. The initiation codon for p90-Rdm2 lies in exon 3 and that for p76-Rdm2 lies in-frame in exon 4 (Fig. 1). Two short and overlapping upstream open reading frames are present in exon 1 of the deduced Rdm2 mRNA sequence (supplemental Fig. 2). The HDM2 and Mdm2 transcripts also each contain two (differing) upstream open reading frames (17, 18). The deduced amino acid sequence of Rdm2 and the Mdm2 sequence are ∼90% identical, although Mdm2 possesses a more extended Ser-rich region at residues 203–211 (supplemental Fig. 1).

FIGURE 1.

Deduced structure of the 5′ region of the Rdm2 gene. Further details are provided under “Results.” The complete deduced sequence of mature Rdm2 mRNA is shown in supplemental Fig. 2. The numbering refers to the positions of the nucleotides in the gene sequence. The positions of the primers used to determine P1/P2 promoter usage in exons 1, 2 (forward primers F1 and F2, respectively), and in exon 4 are indicated, as are the two alternative translational initiation codons (ATG(1) for p90 Rdm2 and ATG(2) for p76 Rdm2). The consensus sequences for the transcription factor response elements are as follows: MZF-1, AGTGGGGA or the more extended CGGGNGAGGGGGAA (36); Ets, C(A/C)GGA(A/T)G(C/T) (34); AP-1, TGA(C/G)TCA (33); NF-E2, (C/T)TGCTGA(C/G)TCA(C/T) (35); p53, two (A/G)(A/G)(A/G)C(A/T)(A/T)G(C/T)(C/T)(C/T) elements separated by 0–13 nucleotides (31). The positions of the response elements and the 5′ to 3′ directions applicable are marked by arrows. The positions of the primers used in the chromatin immunoprecipitation studies of the AP-1 response element are indicated by underlying asterisks (see also Table 1).

Expression of Rdm2 mRNA and Protein in Response to H₂O₂

We examined the effects of H₂O₂ on Rdm2 expression by RT-sqPCR and RT-qPCR. Up-regulation of Rdm2 mRNA was rapid, being detectable at 1 h (Fig. 2, A–C), maximal at 2 h, and maintained for at least 6–24 h. The absolute extent of maximal up-regulation depended on the technique used with sqPCR giving 3–4-fold (Fig. 2A) but with qPCR giving 20–30-fold (Fig. 2, B and C). This probably reflects differences in primer design and the semi-quantitative nature of densitometric analysis. At 1 h, Rdm2 mRNA was detectably induced with 0.1 mm H₂O₂ and maximally induced with 0.2–0.3 mm H₂O₂ (Fig. 2, D and E). At higher concentrations of H₂O₂, Rdm2 mRNA abundances remained at control values (Fig. 2, D and E). At 2 h of exposure, the maximal response was greater and was shifted to higher H₂O₂ concentrations (Fig. 2E). These “bell-shaped” concentration dependences are typical of a number of responses of the cardiac myocyte to H₂O₂. The stimulation of Rdm2 expression by H₂O₂ at 1 h was prevented by cycloheximide (Fig. 3, A and B), indicating that Rdm2 is not an immediate early gene (i.e. the stimulation of its transcription by H₂O₂ is dependent on preceding de novo protein synthesis).

FIGURE 2.

Expression of Rdm2 transcripts and protein in cardiac myocytes exposed to H₂O₂. RNA or protein was extracted, and Rdm2 mRNA or protein abundances were determined by RT-sqPCR, RT-qPCR, or immunoblotting as described under “Experimental Procedures.” A–C, myocytes were exposed to 0.2 mm (A and B) or 1 mm H₂O₂ (C) for the times indicated and Rdm2 transcripts measured by RT-sqPCR (A) or RT-qPCR (B and C). D and E, dependence of Rdm2 mRNA expression at 2 h (D, RT-sqPCR; E, RT-qPCR) or 1 h (E, RT-qPCR) on H₂O₂ concentration. No RT, as a control for genomic contamination, the RT step was omitted. F and G, Rdm2 protein abundances were determined by immunoblotting of whole cells extracts at the times indicated. Primary antibody dilutions were both at 1:750. Representative primary data are shown, and the quantified results are expressed as means ± S.E. for at least three independent observations on separate preparations of myocytes.

FIGURE 3.

Rdm2 is not an immediate early gene nor is Rdm2 mRNA stabilized by H₂O₂. A and B, cardiac myocytes were pretreated with cycloheximide (CHX, 20 μm) for 30 min and then exposed additionally to 0.2 mm H₂O₂ for 2 h (A, RT-sqPCR) or 1 h (B, RT-qPCR). C, H₂O₂ (0.2 mm) exposure does not affect Rdm2 transcript stability. Myocytes were exposed (●) or not exposed (▴) to 0.2 mm H₂O₂ for 2 h and then to 4 μm actinomycin D for the times indicated, followed by RT-sqPCR. No RT, as a control for genomic contamination, the RT step was omitted. Results were fitted to monophasic exponential decay curves. Representative primary data are shown, and the quantified results are expressed as means ± S.E. for 3–4 independent observations on separate preparations of myocytes. Statistical significance was as follows: *, p < 0.001 by one-way ANOVA with Tukey's multiple comparison test.

Because H₂O₂ can inhibit overall protein synthesis in cardiac myocytes (3), we investigated whether the increased abundance of Rdm2 transcripts was reflected in their translation into protein (Fig. 2, F and G). Rdm2 migrated at about 90 kDa on SDS-PAGE, and protein abundance increased about 5-fold over 45 min to 1.5 h (Fig. 2F). In some blots, multiple bands were observed (Fig. 2, F and G). Although one of these may represent p76-Rdm2, they may be degradation products or cross-reacting species. Abundance of Rdm2 protein was maximal at 1–2 h and declined gradually thereafter (Fig. 2G) but remained elevated over control values (about 2–3-fold greater) for up to 8 h.

Abundance of any individual mRNA species is regulated by the rate of its transcription and the rate of its degradation. The 3′-untranslated region of Rdm2 mRNA contains an AU-rich instability element (probably a class I ARE consisting of overlapping or adjacent AUUUA pentamers (30), bases 2072–2085, see supplemental material Fig. 2), which is orthologous with a putative Mdm2 mRNA ARE (see the Organism ARED (23, 30), although the Mdm2 ARE may be nonfunctional (23). We therefore examined the half-life of Rdm2 transcripts when transcription was inhibited (by actinomycin D) in the absence or presence of 0.2 mm H₂O₂ (Fig. 3C). The data sets (n = 4 for each condition) were treated as monophasic exponential decays. Rdm2 mRNA was degraded with a half-life of 1.16 h (k = 0.596 ± 0.140 h⁻¹) in the absence of H₂O₂, whereas it was 1.82 h (k = 0.381 ± 0.141 h⁻¹) in the presence of H₂O₂. There was thus no significant difference in the stability of the transcripts under the two conditions, and thus up-regulation of Rdm2 expression results primarily from increased transcription.

Alternative Promoter Usage in Rdm2 Transcription

Transcription of Mdm2 and HDM2 is regulated by two alternative promoters, P1 or P2, lying 5′ to exon 1 or exon 2, respectively (16). To assess alternative promoter usage for Rdm2, we used forward primers in exon 1 (F1) or in exon 2 (F2) and a reverse primer in exon 4 (Fig. 1 and supplemental Fig. 2) to produce amplicons of 333 bp from P1 (long mRNA) or 220 bp from either P1 or P2 (long + short mRNAs). In unstimulated cardiac myocytes, mRNA was detectable at 333 bp (Fig. 4A) using primer F1, and this would contribute to the 220-bp band seen with primer F2 to an inaccessible extent. Following exposure to H₂O₂ (0.2 mm for 2 h), there were no changes in the amplicon abundances using primer F1, but primer F2 amplicon abundances increased about 4–5-fold. The sequences of the 220- and 333-bp PCR products were confirmed, indicating that H₂O₂-inducible transcription was regulated by the P2 promoter.

FIGURE 4.

Promoter usage for Rdm2 expression. A, cardiac myocytes were exposed to or not treated with 0.2 mm H₂O₂ for 2 h. RNA was extracted, and RT-sqPCR was used with a reverse primer in exon 4 and forward primers lying in exon 1 (primer F1) to assess P1 promoter usage (long product, 333 bp) or exon 2 (primer F2) to assess combined P1 and P2 promoter usage (short product, 220 bp) (Fig. 1 and Table 1). Representative primary data are shown, and the quantified results are expressed as means ± S.E. for six independent observations on separate preparations of myocytes. Statistical significance versus control: *, p < 0.01 by one-way ANOVA with Tukey's multiple comparison test. B, 5′-RACE for Rdm2 transcripts in myocytes exposed or not exposed to 0.2 mm H₂O₂ for 2 h. The theoretical lengths of amplicon A (from P1 promoter use) and amplicon B (from P2 promoter use) are 332 and 162 bp, respectively.

To confirm increased P2 promoter usage, we conducted 5′-RACE. The theoretical lengths of the amplicons were 332 bp for a transcript expressed from the P1 promoter and 162 bp for a transcript expressed from the P2 promoter (Fig. 1 and supplemental Fig. 2). Under control conditions, the major product (amplicon A) was 300–400 bp in length with a weaker band (amplicon B) at 100–200 bp (Fig. 4B). Following exposure to H₂O₂ (0.2 mm, 2 h), the product corresponded essentially completely to the shorter amplicon B. DNA sequencing confirmed that amplicon A corresponded to exons 1–3 (up to the reverse primer site), whereas amplicon B corresponded to exons 2 and 3. These results confirm that transcripts originate from P1 with some expression from P2 in the control, whereas promoter usage shifts primarily to P2 following exposure to H₂O₂.

Transcription Factor Response Elements in Intron 1 of the Rdm2 Gene

We identified potential regulatory elements in Rdm2 intron 1 by analogy with published data for intron 1 of the Mdm2 and HDM2 genes and by using TFSEARCH. Starting the number of intron 1 from its 5′-end, a region in intron 1 of the Mdm2 (residues 402–523), Rdm2 (residues 374–494), and HDM2 genes (residues 406–524) is highly conserved with ∼75% identity (supplemental Fig. 3). This contains the two p53-response elements (31, 32) and a composite AP-1-Ets element (Ets^(AP-1)) (33, 34). An upstream Ets (Ets^(upstream)) element in Rdm2 and Mdm2 lies outside the conserved region (supplemental Fig. 3 and Fig. 1). High stringency searching also reveals an NF-E2-like site (35) in Rdm2 that overlaps with the AP-1 site and is conserved between Mdm2 and HDM2. Rdm2 and Mdm2 also contain a sequence reminiscent of an MZF-1 element (36) that is not conserved in HDM2, although HDM2 contains an alternative MZF-1 site (supplemental Fig. 3). However, neither NF-E2 nor MZF-1 are thought to be significantly expressed outside hematopoietic cells (35, 37), and they were not studied further.

p53

Induction of p53 is frequently a response to mutational or cytotoxic interventions and is of established importance in the regulation of inducible expression of Mdm2 (16). We examined whether p53 mRNA expression increased in response to H₂O₂. Although an amplicon of the appropriate size for p53 was detectable by RT-sqPCR, exposure of myocytes to 0.2 mm H₂O₂ for up to 24 h did not alter its expression (results not shown).

The matter of whether p53 protein in induced by ROS is contentious, and doubts have been raised about the reliability of antibodies obtained from Santa Cruz Biotechnology, Inc. (see under “Discussion”). We characterized six different antibodies using lysates of 293T cells with (positive control) or without (negative control) transfection with mouse p53 (Fig. 5). Three of these reliably detected p53 by immunoblotting in our hands (FL-393 (Fig. 5A), Pab 240 (Fig. 5B) and 2B2.71 (Fig. 5C)). In all cases, p53 migrated slightly more rapidly than a 52-kDa marker (RPN800E, GE Healthcare). With FL-393 (Fig. 5A) and 2B2.71 (Fig. 5C), cross-reacting proteins were detectable in whole cell extracts of 293T cells, one of which migrated slightly more slowly than p53 at ∼52 kDa. We confirmed that p53 in crude nuclear extracts of rat cardiac myocytes, which had been exposed to the anthracycline daunorubicin (anthracyclines induce p53 in these cells (38)), co-migrated with murine p53 in the positive-control 293T cells (Fig. 5D).

FIGURE 5.

Characterization of p53 antibodies. Extracts of 293T cells that had either been transfected with mouse p53 (positive control) or not transfected (negative control) were examined by SDS-PAGE and immunoblotting with FL-393 antibody (A), Pab 240 antibody (B), or 2B2.71 antibody (C and D), all at 1:1000 dilution. D, a crude nuclear extract from cardiac myocytes exposed to 0.5 μm daunorubicin for 6.5 h was included. The position of the 52-kDa molecular mass marker is indicated.

Following exposure of cardiac myocytes to 0.2 mm H₂O₂, Rdm2 transcript abundances reach their maximum levels in crude nuclear extracts after about 2 h (Fig. 2, A–C). If de novo induction of transcription factor proteins is a prerequisite for this event, it should occur within this period. We assessed whether p53 protein was induced over a period of up to 8 h in crude nuclear extracts of cardiac myocytes exposed to 0.2 mm H₂O₂ using antibodies FL-393 (Fig. 6A), 2B2.71 (Fig. 6, B and C), and Pab 240 (Fig. 6D). We also assessed whether increasing the H₂O₂ concentration to 0.5 mm had any additional effect (Fig. 6C). In all cases, a positive control for p53 protein induction was included using crude nuclear extracts of cardiac myocytes exposed to 0.5 μm daunorubicin for 6.5–7.5 h. In no instance could we detect induction of p53 by H₂O₂ (Fig. 6, A–D), although p53 protein was reliably detected following treatment with daunorubicin (0.5 μm for 6.5–7.5 h).

FIGURE 6.

H₂O₂ does not induce expression of p53 in cardiac myocytes. Myocytes were exposed to 0.2 mm (A–F) or 0.5 mm (C) H₂O₂ for the times indicated or to daunorubicin. A–D, immunoblots of crude nuclear extracts from myocytes. Blots were probed with the p53(FL-393) antibody (A) at a dilution of 1:1200, with the 2B2.71 antibody (B and C) at a dilution of 1:1000, or with the Pab 240 antibody (D) at a dilution of 1:1000. For 2B2.71 and Pab 240, an amplification step with an rabbit anti-mouse antibody and a goat anti-rabbit horseradish peroxidase-conjugated tertiary antibody were used. E, EMSAs of crude nuclear extracts were carried out with a p53 consensus binding sequence probe. F, myocytes were exposed to 0.5 μm daunorubicin for 7.5 h, and supershift EMSAs were carried out with the antibodies indicated. Representative primary data are shown for three independent observations on separate preparations of myocytes.

To confirm the immunoblotting results (Fig. 6, A–D), we assessed whether any p53 DNA binding activity could be detected in crude nuclear extracts of myocytes exposed to H₂O₂ using EMSAs. Although daunorubicin increased protein binding to a p53-response element consensus sequence, 0.2 mm H₂O₂ did not induce any detectable binding (Fig. 6E). DO-1 antibodies to p53 produced a supershifted band on EMSAs with daunorubicin, whereas FL-393 antibodies prevented binding to the probe (Fig. 6F), thus confirming that the band detected on EMSAs with daunorubicin was indeed p53. An irrelevant antibody (to JunD) did not supershift the complex (Fig. 6F). Thus, by using two independent methodologies and a variety of antibodies, we were unable to find evidence for p53 induction by H₂O₂. We conclude that it is unlikely that p53 is responsible for the increased Rdm2 P2 promoter activity induced by H₂O₂.

Ets

We examined whether Ets proteins may regulate Rdm2 expression. An EMSA probe corresponding exactly to the Ets^(upstream) region (Table 2 and Fig. 1) detected three complexes in crude nuclear extracts of myocytes exposed to 0.2 mm H₂O₂ for 0 or 30 min, with complex 2 showing evidence of inducibility (Fig. 7A). Unlabeled wild type probe competed effectively for binding to complexes 1 and 2 (but not to complex 3) even at an equimolar concentration, whereas an unlabeled probe with two mutated bases did not completely compete for binding even when in a 100-fold excess (Fig. 7A). Thus complex 3 represents nonspecific binding. Binding of complex 1 was “constitutive” and remained constant over 2 h (Fig. 7B). Nevertheless, complex 1 may still be important in regulation of the P2 promoter as its transactivating activity (as opposed to DNA binding activity) could be subject to regulation (e.g. phosphorylation (39)). Complex 2 was transiently induced to a maximum of about 8-fold after 30 min (Fig. 7B). By 1 h, ∼60% of the induction had been lost, and abundances had returned to base line by 2 h (Fig. 7B), i.e. its peak induction significantly preceded Rdm2 induction. We also examined the Ets^(AP-1) element using EMSAs with a probe corresponding exactly to the Ets^(AP-1) region (Table 2 and Fig. 1). Although a complex was detectable, its abundance was not altered by exposure of myocytes to 0.2 mm H₂O₂ for up to 8 h (results not shown). Furthermore, unlabeled probe for the Ets^(AP-1) region did not effectively compete until present at a 50–100-fold excess (results not shown). Overall, any binding of proteins to the Ets sequence EMSA probes is either not inducible, not specific, or does not follow a time course appropriate for induction of Rdm2 expression.

FIGURE 7.

Binding of proteins to the Ets^(upstream) and AP-1-response elements in the Rdm2 P2 promoter as assessed by EMSAs. Cardiac myocytes were exposed to 0.2 mm H₂O₂ as indicated and crude nuclear extracts prepared. Sequences of probes are given in Table 2. A and C, EMSAs using the Ets^(upstream) probe (A) or the AP-1 probe (C). Binding specificities were assessed using unlabeled probe. Two specific complexes (1 and 2) were detected with the Ets^(upstream) probe (A), and one was detected for the AP-1 probe (C). B and D, time courses of binding to the Ets^(upstream) probe (B) or the AP-1 probe (D). Representative primary data are shown, and the quantified results are expressed as means ± S.E. for 3–5 independent observations on separate preparations of myocytes.

AP-1

The AP-1 element is situated immediately 5′ to the Ets^(AP-1) element in the Rdm2 P2 promoter (Fig. 1). An EMSA probe corresponding exactly to this AP-1 sequence (Table 2 and Fig. 1) detected a complex in crude nuclear extracts of myocytes that was competed by unlabeled wild type but not mutant probe (Fig. 7C). Levels of this complex were increased by H₂O₂ to a maximum at 1–2 h (Fig. 7D). Supershift EMSAs were used to identify the AP-1 proteins (the Jun and Fos family members) binding to the AP-1 probe. In the absence of stimulation, the complex was partially supershifted by antibodies to JunB or JunD but not by antibodies to c-Jun (Fig. 8A). Following exposure to H₂O₂ (0.2 mm 1.5 h), antibodies to JunB, JunD, and, additionally, c-Jun partially supershifted the complex (Fig. 8B). Antibodies to c-Fos, FosB, and Fra-1 also supershifted this complex, but antibodies to Fra-2 or ATF2 were ineffective (Fig. 8C). When a mixture of all supershifting antibodies (c-Jun, JunB, JunD, c-Fos, FosB, and Fra-1) was used, supershifting of the complex was essentially complete (Fig. 8D). For c-Jun and JunB, ChIP assays confirmed that binding to an Rdm2 P2 sequence containing the AP-1 site (Fig. 1 and Table 1) increased following exposure of cardiac myocytes to H₂O₂ (0.2 mm, 1.5 h) (Fig. 8E). These results indicate that Fos and/or Jun family transcription factors bind to the Rdm2 P2 AP-1 sequence, and this is increased by H₂O₂.

FIGURE 8.

Binding of AP-1 proteins to the AP-1-response element in the Rdm2 P2 promoter as assessed by supershift EMSAs or ChIP assays. Cardiac myocytes were exposed to 0.2 mm H₂O₂ as indicated and crude nuclear extracts prepared. A–D, supershift EMSAs were performed using the antibodies indicated. SS, supershifted complex. E, ChIP assays using c-Jun or JunB antibodies, as indicated. Representative primary data are shown, and the quantified results are expressed as means ± S.E. for three independent observations on separate preparations of myocytes. Statistical significance versus control: *, p < 0.05 by an unpaired two-way Student's t test.

Expression of Jun and Fos Family Members

We examined the time courses of expression of c-Jun, JunB, and JunD transcripts and proteins in response to 0.2 mm H₂O₂. Transcripts for both c-Jun (Fig. 9A) and JunB (Fig. 9B) were increased maximally (about 8- and 4-fold for c-Jun and JunB, respectively) by H₂O₂ after about 1 h but had declined to control values by 4 h. Protein abundances followed time courses essentially similar to those for transcripts with c-Jun showing a 7–8-fold increase (Fig. 9, C and D), and JunB showing a 2.0–2.5-fold increase (Fig. 9, E and F). There were no changes detectable in JunD transcript or protein abundances (results not shown).

FIGURE 9.

Expression of c-Jun and JunB transcripts and proteins. Cardiac myocytes were exposed to 0.2 mm H₂O₂ as indicated. mRNA abundance was determined by RT-sqPCR, and protein abundance in crude nuclear extracts was assessed by immunoblotting. A and B, expression of c-Jun (A) and JunB (B) transcripts. No RT, no reverse transcriptase step. C–F, expression of c-Jun (C and D) and JunB (E and F) proteins. Antibody dilutions are as follows: c-Jun, 1:750; JunB, 1:5000. Representative primary data are shown, and the quantified results are expressed as means ± S.E. for 3–4 independent observations on separate preparations of myocytes.

A similar pattern of mRNA expression was observed for the Fos family members implicated by EMSAs in the binding to the Rdm2 P2 AP-1 sequence (Fig. 10, A–D). Transcript abundances increased by ∼20-fold for c-Fos (Fig. 10A), 15-fold for FosB (Fig. 10B), and 10-fold for Fra-1 (Fig. 10C) over 1–2 h before subsequently declining. Abundances of Fra-2 did not change significantly (results not shown) in accord with the results of supershift EMSAs (Fig. 8C). For c-Fos protein, abundance increased by ∼15-fold over the same time period before declining (Fig. 10D). Antibodies against FosB and Fra-1 were not satisfactory for unequivocal use in immunoblotting in our hands. These results suggest that one factor in the increase in binding of AP-1 proteins to the Rdm2 P2 sequence is an increase in their abundances.

FIGURE 10.

Expression of c-fos, Fos-B, and Fra-1 transcripts and c-Fos protein. Cardiac myocytes were exposed to 0.2 mm H₂O₂ as indicated. mRNA abundance was determined by RT-sqPCR, and protein abundance in crude nuclear extracts was assessed by immunoblotting. A–C, expression of c-Fos (A), FosB (B), and (C) Fra-1 transcripts assessed by RT-sqPCR. No RT, no reverse transcriptase step. D, expression of c-Fos protein. Antibody dilution for c-Fos: 1:20,000. Representative primary data are shown and the quantified results are expressed as means ± S.E. for 3–4 independent observations on separate preparations of myocytes.

Effects of Rdm2 on Caspase 3 Cleavage in H₂O₂-treated Cardiac Myocytes

Mdm2 and its orthologues are generally considered to be antiapoptotic. We examined whether Rdm2 expression delayed the pro-apoptotic effects of oxidative stress in cardiac myocytes using an antisense ODN to a region of Rdm2 exon 7 (supplemental Fig. 2). This reduced H₂O₂-stimulated expression of Rdm2 protein by ∼75% (Fig. 11, A, top panel, and B). Apoptosis was assessed by formation of the 17-kDa active (cleaved) form of caspase 3. H₂O₂ increased the abundance of 17-kDa caspase 3 in the absence or presence of lipofectin by about 2.5–4-fold (Fig. 11, A, middle panel, and C), thus demonstrating that, under the conditions used, H₂O₂ was pro-apoptotic. The Rdm2 antisense ODN (but not a scrambled ODN) further increased the H₂O₂-mediated stimulation of 17-kDa caspase 3 formation by some 2–3.5-fold (depending on the value of 17-kDa caspase 3 formation (lipofectin, no H₂O₂, or nonsense oligodeoxynucleotide + H₂O₂) used as denominator) (Fig. 11, A, middle panel, and C). As judged by immunoblotting for α-actin, the adherence of the myocytes to the culture dishes was unaffected by any treatment (Fig. 11A, bottom panel). These results suggest that one role of increased Rdm2 expression in myocytes exposed to oxidative stress is to reduce or delay apoptosis.

FIGURE 11.

“Knockdown” of Rdm2 protein in cardiac myocytes with an antisense ODN to the Rdm2 transcript increases H₂O₂-induced activation of caspase 3. AS, antisense ODN; NS, scrambled ODN. A, abundances of Rdm2 (top panel), active caspase 3 (middle panel), and sarcomeric α-actin (bottom panel) in whole cell extracts were assessed by immunoblotting. B and C, quantification of Rdm2 (B) or active caspase 3 abundances (C). Antibody dilutions are as follows: Rdm2, 1:500; caspase 3, 1:1000; α-actin, 1:2000. Representative primary data are shown, and the quantified results are expressed as means ± S.E. for four independent observations on separate preparations of myocytes. Statistical significance versus control: *, p < 0.05; **, p < 0.001 by one-way ANOVA with Tukey's multiple comparison test.

DISCUSSION

Effects of H₂O₂ on Mdm2 Expression in the Heart

The effects of ROS generally and H₂O₂ in particular on Mdm2 expression in cardiac and noncardiac cells have not been extensively examined previously. In noncardiac cells, H₂O₂ up-regulates Mdm2 mRNA and protein expression in concert with up-regulation of p53, and the assumption is that induction of p53 protein is responsible for up-regulation of Mdm2 (40–42). However, this may not be universal (43). Of relevance to cardiac myocytes, in the H9c2 rat ventricular myoblast line, H₂O₂ (0.4 mm, 24 h) increases Rdm2 protein abundance (44), whereas in contrast to our microarray data (3, 10), Rdm2 transcript abundance was reported to decrease in cardiac myocytes exposed to H₂O₂ (0.1 mm, 1 h) (45). To study Rdm2 in more detail, we first needed to characterize the Rdm2 gene and its transcript and Rdm2 protein using bioinformatics approaches (Fig. 1 and supplemental Figs. 1–3). The current details for Rdm2 mRNA in the NCBI data base (www.ncbi.nlm.nih.gov) are misleading as the 75-bp exon 4 has been omitted. It could be argued that this shorter species lacking “exon 4” is specific to the rat and could arise, for example, by alternative splicing/exon skipping. However, in addition to identifying exon 4 in rat ESTs, the reverse primer (Table 1) used in the experiments described in Fig. 3A lies within exon 4, and products of the predicted sizes were generated, thus establishing that exon 4 is expressed in mature Rdm2 mRNA.

We showed that Rdm2 transcripts and Rdm2 protein were rapidly induced (within 1–2 h) by H₂O₂ and that this occurs over the 0.1–1 mm range (Fig. 2, A–E). At higher concentrations, expression declines (Fig. 2, D and E). Rdm2 was not regulated as an immediate early gene (Fig. 3, A and B), and the up-regulation of the transcript did not involve transcript stabilization (Fig. 3C). As in other cells (16), constitutive Rdm2 expression involved the 5′-P1 promoter, whereas H₂O₂-inducible expression was largely regulated through the P2 promoter within intron 1 (Fig. 4, A and B). Intron 1 contains several potential transcription factor binding sites, including sites for p53, AP-1 and Ets (Fig. 1), any or all of which may be important in the inducible expression of Rdm2.

Induction of p53 by H₂O₂ in Cardiac Myocytes

At least two studies report that H₂O₂ induced expression of p53 transcripts and protein in neonatal rat cardiac myocytes in culture (2, 45). Because our original data contradicted these studies, we have been encouraged to go to great lengths to prove that our data are correct. In addition to the antibody (FL-393) we originally used, we used two further antibodies (Pab 240 and 2B2.71) for immunoblotting. We first confirmed the selectivity and specificity of these antibodies for p53 by immunoblotting commercially available whole cell extracts of 293T cells transfected with mouse p53 (Fig. 5, A–C). All three antibodies detected p53 as a band that migrated slightly more rapidly that the 52-kDa marker (Fig. 5, A–C). This band was absent from the negative control cell line, although it should be noted that cross-reacting proteins of different apparent molecular masses could be detected with two of the antibodies, including one that migrated with the 52-kDa marker (Fig. 5, A, C, and D). Furthermore, rat p53 (induced by treatment of cardiac myocytes with daunorubicin) co-migrated with murine p53 in the positive control cell line (Fig. 5D). None of the three antibodies detected any increase in p53 expression in cardiac myocytes following exposure to H₂O₂, despite a clear induction with daunorubicin (Fig. 6, A–D). Furthermore, we could not detect any induction of p53 by H₂O₂ using EMSAs despite the clear induction by daunorubicin of a band that could be supershifted with p53 antibodies (Fig. 6, E and F). Thus, we are confident that 0.2 mm H₂O₂ does not induce expression of p53 in cardiac myocytes, and this probably excludes any role for p53 in H₂O₂-induced Rdm2 expression.

It is not entirely clear why others have detected p53 induction by H₂O₂ (2, 45), but there are at least four potential areas of concern as follows: (i) the antibody used; (ii) the cell extracts used; (iii) the conditions used; and (iv) the absence of positive controls for p53. First, both von Harsdorf et al. (2) and Long et al. (45) used a monoclonal antibody from Calbiochem of which there are six currently listed. Not surprisingly, given the time elapsed, none of the authors of Ref. 2 could recall which antibody was used, and indeed the choice may have been limited or nomenclature not specified by the supplier at this stage. In the report by Long et al. (45), Dr. José Marín-García informed us that they had used the Calbiochem Ab-4 antibody (catalogue number OP32) from the Pab 246 clone, but he now purchases this from Santa Cruz Biotechnology, Inc. (catalogue number sc-100). We found that Pab 246 cross-reacted with several proteins in the 52-kDa region in 293T whole cell extracts (results not shown), and we did not use this antibody. Second, in both reports, whole cell extracts of cardiac myocytes were used rather than crude nuclear extracts. Generally, we find that whole cell extracts are unsatisfactory for detection of nuclear transcription factors by immunoblotting, mainly because of cross-reaction with other proteins. Indeed, two of the three antibodies showed cross-reactivity with a band at 52 kDa in 293T whole cell extracts (Fig. 5, A, C and D), which was not detectable in nuclear extracts from rat cardiac myocytes (Fig. 5D and Fig. 6, A–D) but which, in the absence of a positive control, could be mistaken for p53. Third, although the conditions used by Long et al. (45) (0.1 mm H₂O₂ for 30 min, 1 h, or 24 h) resembled our own, von Harsdorf et al. (2) used a potentially more severe oxidative stress (0.1 mm H₂O₂, 0.1 mm Fe²⁺ for 1 h, followed by further culture in H₂O₂/Fe²⁺-free medium for up to 23 h). The combination of H₂O₂ and Fe²⁺ leads to a catalytic disproportionation of H₂O₂ into the highly reactive hydroxyl radical (which is toxic and induces a severe oxidative stress), and a peroxide radical, and is commonly known as the Fenton reaction (Fe²⁺ + H₂O₂ → Fe³⁺ + OH^• + OH⁻; Fe³⁺ + H₂O₂ → Fe²⁺ + OOH^• + H⁺). Fourth, for both of the published studies (2, 45), there were no positive controls for p53, nor were the positions of molecular mass markers included.

Regulation of Inducible Mdm2 by p53-independent Mechanisms

Expression of the Mdm2 gene can also be regulated by p53-independent mechanisms (19–21). Thus, point mutations in any one of the AP-1, Ets^(AP-1), or Ets^(upstream) elements reduce induction of Mdm2 expression by oncogenic Ras or Raf in NIH-3T3 cells (19). Furthermore, elevated HDM2 expression in a breast cancer cell line is partly dependent on the composite AP-1-Ets^(AP-1) element (20). In contrast, up-regulation of Mdm2 in certain mouse erythroleukemia lines is partly attributable to increased expression of Fli-1 (an Ets family transcription factor) and may involve the Ets^(upstream) element (21).

Ets family transcription factors number >20 members (39) that require a core GGA motif for response element recognition with slightly varying preferences in the flanking sequences (34). In cardiac myocytes exposed to H₂O₂, constitutive (complex 1) and transiently inducible (complex 2) protein binding to the Ets^(upstream) element was detected by EMSAs (Fig. 7, A and B). We attempted to supershift the Ets^(upstream) complexes with antibodies to the biologically activated form of Elk-1 (phospho-Elk-1(Ser-383)) (the orthologous residue in rat Elk-1 is in fact Ser-382) and to Fli-1 but were unsuccessful (results not shown). We consider it unlikely that the Ets^(upstream) element is significant in Rdm2 up-regulation because of the following: (i) the level of complex 1 does not change and (ii) levels of complex 2 were returning to zero time values before any significant induction of Rdm2 transcripts or Rdm2 protein was detectable (Fig. 2, A–C and F, and Fig. 7B). Equally, we could not obtain any convincing evidence that the Ets^(AP-1) element might be involved. The EMSA band with the Ets^(AP-1) probe was not inducible and was not competed by unlabeled probe.

We turned our attention to the AP-1 site. AP-1 is a heterodimer of Fos and Jun family members or possibly a homodimer of Jun family members, and it controls the transcription of numerous genes (46). The AP-1 complex can either activate or inhibit gene expression, and its two subunits do not necessarily act in concert. Thus, whereas c-Fos and FosB act in concert with c-Jun (the most powerful transcriptional regulator), c-Jun may be antagonized by JunB and by Fra-1 and Fra-2, which do not contain transactivation domains. Furthermore, the transactivating activities of AP-1 proteins can be influenced by post-translational modification (e.g. phosphorylation). As with Ets transcription factors, there is evidence that AP-1 may participate in regulation of Mdm2 expression (19). In our study, EMSAs showed specific binding of proteins to the Rdm2 AP-1 element (Fig. 7C) with a time course consistent with induction of Rdm2 expression (Fig. 2, A–C, and Fig. 6D). c-Jun and JunD bound to the probe in unstimulated cells (Fig. 8A). In cells exposed to H₂O₂, JunB (Fig. 8B) and Fos family members (c-Fos, FosB, Fra-1, but not Fra-2) were also bound (Fig. 8C). The binding of Fra-1 is a little surprising given that expression of Rdm2 is increased, but Fra-1 is supposed to reduce AP-1-dependent stimulation of gene expression (46), and we do not have any explanation for this at present. A mixture of antibodies to c-Jun, JunB, JunD, c-Fos, FosB, and Fra-1 resulted in essentially a complete supershift of the complex (Fig. 8D). Of the Jun and Fos members mentioned but with the exceptions of JunD and Fra-2, transcripts (and usually proteins) were induced by H₂O₂ (Figs. 9 and 10) within the time period chosen for supershifting (1.5 h). Using ChIP assays, we demonstrated increased binding of c-Jun and JunB to an Rdm2 P2 promoter region encompassing the AP-1 site (Fig. 8E). ChIP assays using antibodies to the remaining Fos/Jun proteins were less successful, presumably because of their inadequacy in immunoprecipitation protocols. These results strongly suggest that the Rdm2 AP-1 element is of significance in the up-regulation of Rdm2 expression by H₂O₂, and that increased abundances of AP-1 transcription factors (several of which have been confirmed as immediate early genes in cardiac myocytes in the context of endothelin stimulation (47)) participate in Rdm2 mRNA up-regulation.

H₂O₂ and ROS are established regulators of AP-1 (48). All of the three major MAPK cascades (JNK, ERK1/2, and p38 MAPK) are activated by H₂O₂ in cardiac myocytes (49) and may influence AP-1 expression and activity. JNKs phosphorylate c-Jun in its transactivation domain and increase its transactivating activity (46). In addition, c-Jun mRNA and protein are up-regulated by a combination of activation of ERK1/2 (to increase transcription of c-Jun) and JNKs (to increase protein stability) (50). Thus, H₂O₂-mediated stimulation of MAPK cascades might be a factor in the up-regulation of AP-1 transcription factor levels and activities in our study.

Cardioprotective Effects of Mdm2

p90 Mdm2 is present in whole murine heart in greater abundance than p76 Mdm2 (although both at much lower abundances than in other tissues such as testis) (23). Cardioselective deletion of Mdm2 is embryonically lethal, and this lethality can be rescued by deletion of p53 (51), thus emphasizing the importance of Mdm2 in limiting p53 expression during cardiac development. In the developed heart, the role of Mdm2 in cardioprotection is implicit from a number of studies. Increasing Mdm2 abundance by overexpression of the transcriptional coactivator p300 protects against apoptosis induced by the cardiotoxic anthracycline doxorubicin in mice in vivo (52). Cardiospecific overexpression of insulin-like growth factor 1 increases the association between p53 and Mdm2 thus reducing the binding of p53 to the promoter of the proapoptotic bax gene (53). The growth factor also reduces mechanical strain-induced apoptosis in isolated cardiac myocytes, one possible mechanism involving an Rdm2-mediated reduction in p53 abundance (54). Overexpression of Mdm2 also protects against the pro-apoptotic effects of hypoxia, and inhibition of the p53-Rdm2 interaction increases the sensitivity to hypoxia (55). Furthermore, reduced expression of Mdm2 increases the susceptibility of the isolated heart to injury induced by ischemia and reperfusion (55).

In general agreement with these earlier studies, our studies here suggest that the induction of Rdm2 expression by H₂O₂ represents a cytoprotective defense against oxidative stress in the cardiac myocyte (Fig. 11). Thus, diminished H₂O₂-mediated induction of Rdm2 using an antisense ODN to Rdm2 promoted caspase 3 activation (Fig. 11, A and C). It is pertinent to ask how Rdm2 can decrease apoptosis, given that H₂O₂ did not induce p53 expression (Fig. 6). One explanation may be that other pro-apoptotic proteins expressed in myocytes are inhibited or down-regulated by Rdm2. Precedents for this exist. For example, treatment of p53-null neuroblastoma cells with the HDM2-protein interaction inhibitor, Nutlin-3, stimulates apoptosis by inhibiting the interaction with the p53-related protein p73 (56). It should perhaps also be noted that inhibition of HDM2 is currently under intensive investigation as a chemotherapeutic intervention (57). From the results of our own and other studies, such therapies may decrease cardiac myocyte survival, particularly under conditions of oxidative stress, which probably occur during myocardial ischemia and heart failure.