Antioxidant function of the mitochondrial protein SP-22 in the cardiovascular system.

The mitochondrial protein SP-22 has recently been reported to be a member of the thioredoxin-dependent peroxide reductase family, suggesting that it may be one of the antioxidant systems in mitochondria, which are the major site of reactive oxygen intermediate generation. The aim of this study was to examine whether SP-22 is involved in mitochondrial antioxidant mechanisms and whether its expression is induced by oxidative stresses, particularly those in mitochondria. The expression of SP-22 protein was enhanced by about 1.5-4.6-fold when bovine aortic endothelial cells (BAEC) were exposed to various oxidative stresses, including mitochondrial respiratory inhibitors which increased the superoxide generation in BAEC mitochondria. The expression of SP-22 mRNA increased 2.0-3.5-fold with a peak at 3-6 h after exposure to Fe2+/dithiothreitol or a respiratory inhibitor, antimycin A. BAEC with an increased level of SP-22 protein caused by pretreatment with mild oxidative stress became tolerant to subsequent intense oxidative stress. On the other hand, BAEC that had been depleted of SP-22 with an antisense oligodeoxynucleotide against SP-22 mRNA became more labile to oxidative stress than control BAEC. The induction of SP-22 protein by oxidative stress in vivo was demonstrated in an experimental model of myocardial infarction in rat heart. These findings indicate that SP-22 functions as an antioxidant in mitochondria of the cardiovascular system.

Mitochondria play an important role in aerobic energy metabolism of living cells. The mitochondrial electron transport system consumes approximately 85% of the oxygen utilized by the cell, and about 5% of the oxygen is converted to reactive oxygen intermediates (ROIs) 1 (1,2). The generation of ROIs in mitochondria has been reported to impair various cellular func-tions by attacking reactive moieties of macromolecules such as protein, DNA, and lipid (3)(4)(5)(6). Mitochondrial permeability transition, which may be an initial event in the process of cell death induced by Ca 2ϩ and inorganic phosphate, has recently been reported to be mediated by ROIs and prevented by antioxidants in vitro (7).
The mitochondrial protein SP-22 was originally isolated from bovine adrenal cortex as a substrate protein for mitochondrial ATP-dependent protease (8,9). An analysis of its amino acid sequence revealed that SP-22 is a member of the thioredoxindependent peroxide reductase family like the C22 component of alkyl hydroperoxidase in Salmonella typhimurium (10), thiolspecific antioxidant enzyme (11), 23-kDa macrophage stress protein (MSP23) (12), natural killer cell enhancing factor (NKEF) (13), and MER5 (14) in mammalian cells. MER5 is 92% similar to SP-22 protein and considered to be a mouse homolog of SP-22 (15). Members of this family have a highly conserved active site sequence among a wide range of species and are believed to act as antioxidant systems together with the NADPH-thioredoxin-thioredoxin reductase system (11,16). Among the members of the thioredoxin-dependent peroxide reductase family, SP-22 is the only protein located in mitochondria. To test the hypothesis that SP-22 functions as an antioxidant system in mitochondria, which are the major site of cellular ROI generation, we investigated the oxidant-induced expression of SP-22 using cultured bovine aortic endothelial cells (BAEC) and an in vivo model of experimental myocardial infarction. Furthermore, we also examined the antioxidant function of SP-22 using BAEC with decreased or increased levels of SP-22 protein produced by treatment with antisense oligodeoxynucleotide or mild oxidant preconditioning. The present results indicate that SP-22 plays a crucial role in the antioxidant defense mechanism of mitochondria in the cardiovascular system.

EXPERIMENTAL PROCEDURES
Cell Culture and Materials-BAEC were harvested from bovine thoracic aorta obtained from a local slaughterhouse and cultured in RPMI 1640 medium supplemented with 10% fetal calf serum (FCS) as described by Kaku et al. (17). The identity of endothelial cells was verified by their characteristic morphology and the presence of factor VIIIassociated antigen. Endothelial cells at up to passage 10 were used for these experiments. Mouse anti-manganese superoxide dismutase (Mn-SOD) monoclonal antibody was purchased from Chemicon International (Temecula, CA), anti-4-hydroxy-2-nonenal (HNE) monoclonal antibody was from NOF Co. (Tokyo, Japan), 1-methyl-4-phenylpyridinium was from Aldrich, paraquat from Nacalai Tesque (Kyoto, Japan), antimycin A and superoxide dismutase were from Sigma, and glucose oxidase and xanthine oxidase were from Boehringer Mannheim (Mannheim, Germany). Other chemicals were standard commercial products * This work was presented in part at the 71st Scientific Sessions of the American Heart Association, November 11, 1998, Dallas, TX. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Polyclonal Rabbit Antibodies against SP-22-The peptide SPTAS-REYFEKVNQ, corresponding to residues 183-195 of the SP-22 protein (9), was synthesized and conjugated with hemocyanin. Female Japanese white rabbits were subcutaneously immunized with the conjugated peptide (500 g) emulsified with adjuvant (Titer Max, Sigma). The first booster injection (250 g) was given 4 weeks later, and this was followed by three booster injections (250 g each) at 2-week intervals. Sera were obtained 2 weeks after the last booster injection.
Immunocytochemistry-For immunofluorescence microscopy, KB cells grown on coverslips in minimal essential medium/10% FCS were fixed with 4% paraformaldehyde in phosphate-buffered saline and permeabilized with 0.1% saponine. The permeabilized cells were reacted with anti SP-22 antibody and then stained with rhodamine-conjugated second antibody. Electron immunocytochemistry of bovine mitochondria was carried out as described previously (18). The crude mitochondrial fraction of KB cells was pelleted, fixed with 8% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, and embedded in LR white resin at 50°C. Thin sections of the mitochondria were incubated with the antibody against the SP-22, and then incubated with anti-rabbit IgG-gold. Immunolabeled sections were then stained with uranyl acetate and lead citrate and examined under a Hitachi HU 12 electron microscope at 100 kV. Control experiments done with preimmune serum gave no immunoreactive signals.
Superoxide Production in BAEC Submitochondrial Particles-BAEC submitochondrial particles (SMP) were prepared essentially as described by Kang et al. (18). Briefly, BAEC were washed in an isotonic sucrose buffer composed of 10 mM Tris-HCl, pH 7.4, 1 mM EDTA, 0.25 M sucrose, 15 g/ml leupeptin, 5 g/ml (p-amidinophenyl)methanesulfonyl fluoride hydrochloride, and 50 ng/ml pepstatin, and suspended in the same buffer (1 ϫ 10 8 cells/ml). The cells were homogenized in a Potter-Elvehjem homogenizer and centrifuged twice at 600 ϫ g for 10 min to obtain post-nuclear supernatant. The post-nuclear supernatant was centrifuged at 7000 ϫ g for 10 min. The pellet (crude mitochondrial fraction) was sonicated and centrifuged at 320,000 ϫ g for 1 h. The resultant pellet was homogenized in an isotonic sucrose buffer and served as the SMP fraction. Superoxide production by SMP was determined by the oxidation of adrenaline to adrenochrome, and was corrected by subtracting the rate in the presence of 10 g/ml superoxide dismutase as described by Takeshige et al. (19).
Reverse Transcription-Polymerase Chain Reaction (RT-PCR)-Semiquantitative RT-RCR was performed to quantitate SP-22 mRNAs in oxidant-treated BAEC, as described previously (17). Briefly, total RNA was isolated from confluent BAEC using TRIzol reagent (Life Technologies, Inc.) based on the method reported by Chomczynski and Sacchi (20). RT reactions were carried out with Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.). The reverse-transcribed cDNA products were amplified with Taq DNA polymerase in a PJ2000 DNA thermal cycler (Perkin-Elmer). The conditions for each cycle were 94°C for 45 s, 55°C for 1 min, and 72°C for 2 min (31 cycles). The following gene-specific primers for SP-22 were used: 5Ј-tcacgatgtgaactgcgaagttg-3Ј(bases 366 -388, sense) and 5Ј-ggatggggcttgattgtaggag-3Ј(bases 710 -731, antisense), which amplified a 366-base pair product. Each primer set yielded a single PCR product of the predicted size. The identity of the PCR products was confirmed by direct cycle sequencing. RT-PCR was also performed for the housekeeping gene ␤-actin as a control for the amount of RNA used in the RT reaction. A negative control, in which reverse transcriptase was omitted, was also performed to exclude the possibility of the amplification of contaminating genomic DNA. Linear relationships were observed between the quantity of RNA subjected to the RT reaction and the amount of amplified PCR product under the PCR conditions used for SP-22 (31 cycles) and ␤-actin (24 cycles) ( Fig. 1).
Northern Blotting Analysis-SP-22 PCR products (366 base pairs) were subcloned into the pGEM-T Easy Vector (Promega) by means of the TA cloning technique. After confirming the direction of the PCR insert by sequencing, the plasmids with inserts were linearized with BSP 120I (Fermentas Ltd., Vilnius, Lithuania) and transcribed in vitro into digoxigenin (DIG)-labeled cRNA with SP6 RNA polymerase using a DIG RNA labeling kit (Boehringer Mannheim).
Total RNA (5 g) was dissolved in 12 l of sample buffer containing 10 mM sodium phosphate, pH 7.0, 50% (w/v) dimethyl sulfoxide, and 1 M glyoxal. After denaturation (1 h at 50°C), the samples were electrophoresed in 1% agarose gel with a 10 mM sodium phosphate buffer, and then transferred to a nylon membrane (Boehringer Mannheim) and immobilized by incubation for 30 min at 121°C. Hybridization with the DIG-labeled cRNA probe was carried out overnight at 68°C in 500 mM sodium phosphate, 7% SDS, 1 mM EDTA, and 1 mg/ml yeast tRNA (21). Blots were washed twice at room temperature with 200 mM sodium phosphate, 5% SDS, and 1 mM EDTA, and for 15 min at 65°C with 0.2ϫ SSC plus 0.2% SDS before color reaction using a DIG nucleic acid detection kit (Boehringer Mannheim) according to the instructions provided by the supplier. The amount of SP-22 mRNA was normalized by 18 S ribosomal RNA.
Preconditioning of BAEC with a Mild Oxidant-BAEC were preincubated with 3.3 M Fe 2ϩ , 330 M dithiothreitol (DTT) for 24 h, which maximally increased the expression of SP-22 protein. These treated BAEC were then exposed to more intense oxidative stress consisting of 3.3 M Fe 2ϩ , 1 mM DTT for 24 h, and cell viabilities were evaluated by measuring the lactate dehydrogenase activities of the cell lysate.
Antisense Oligodeoxynucleotide-A 20-mer antisense phosphorothioate oligodeoxynucleotide (ODN) was synthesized by Toagosei Inc. Japan. The first antisense ODN sequence (antisense 1; 5Ј-catcttcgttatgcagggct-3Ј) is directed against the translation initiation region of the SP-22 mRNA. The second antisense ODN (antisense 2; 5Ј-ccttcaccaagcggagggtc-3Ј) is directed against the internal region of SP-22 mRNA. We also prepared a sense ODN of antisense 1 and 2, and several random ODNs for a control experiment. All of these ODNs were designed to not possess sequences homologous to other genomic sequences or strong secondary structures (22). For efficient transfection of BAEC, we used the cationic lipid Tfx-50 reagent (Promega). ODNs that had been diluted in 30 mM HEPES buffer, pH 7.4 (final concentration of 3.3 M), and 5 l of Tfx-50 reagent (final concentration of 5 M), which was prepared according to the instructions provided by the supplier, were mixed and incubated for 15 min at room temperature. The ODN/Tfx-50 reagent mixture was added to each plate with 1 ml of RPMI 1640 medium without FCS, and the cells were incubated for 1 h at 37°C in a 5% CO 2 atmosphere. At the end of the incubation period, the cells were gently overlaid with 2 ml of RPMI 1640 medium containing 10% FCS. This process was performed twice more at 24-h intervals.
Experimental Myocardial Infarction and Immunocytochemistry-Adult male Wistar rats weighing about 250 g were used for these experiments. The experiments were performed in accordance with the guidelines specified for institutional animal care and use of the University of Occupational and Environmental Health, Japan. The animals were anesthetized and ventilated by a Harvard small animal respirator. The left descending coronary artery was ligated via a left-sided thoracotomy, which results in infarction of the free left ventricular wall (23). In a sham procedure, a superficial suture was placed in the epicardium of the left ventricle, near the left descending coronary artery. A total of 9 infarcted rats and 9 sham-operated rats (n ϭ 3 per time point) were killed either 1 or 2 days after surgery. The heart was frozen in liquid nitrogen and stored at Ϫ80°C. The hearts from the sham group and infarct group, at 1 or 2 days after the operation (n ϭ 3), were cut and the sections were fixed with acetone for 5 min. Endogenous peroxidase was then blocked with 0.3% H 2 O 2 in methanol. The thin sections were reacted with the anti-SP-22 polyclonal antibody and anti-HNE monoclonal antibody, and stained by a Vectastain ® Elite ABC kit (Vector Laboratories, Burlingame, CA). The sections were counterstained with hematoxylin.
Other Procedures-Protein concentrations were determined by the Bio-Rad protein assay kit using bovine serum albumin as a standard. The intensity of the stained bands in immunoblotting, Northern blotting, and RT-PCR was quantitated by densitometric analysis using the public domain computer program NIH Image (Wayne Rasband, NIH, Research Service Branch, National Institute of Mental Health, Bethesda, MD), and the results are expressed as means Ϯ S.E. A statistical analysis was performed using Student's t-test. Differences were considered statistically significant at p Ͻ 0.05.

RESULTS
Immunological Detection of SP-22-We prepared an antibody against SP-22 protein and used it for immunological analysis. In immunoblotting, this antibody detected a single band with a molecular mass corresponding to SP-22 (22 kDa) in total homogenate and the mitochondrial fraction of BAEC, but not in the cytosol fraction ( Fig. 2A). Intracellular distribution of SP-22 protein was further examined by immunofluorescence microscopy. When cultured KB cells were stained with the anti-SP-22 antibody, the signals of immunoreactive SP-22 exhibited the mitochondrial staining pattern characterized by a reticular staining appearance (Fig. 2B). Immunostaining signals were hardly visible when preimmune sera for SP-22 were used. The localization of the SP-22 protein in the mitochondria was further confirmed by electron microscopic immunocytochemistry. The mitochondrial fraction isolated from KB cells was stained with the anti-SP-22 antibody, followed by goldlabeled second antibody. Proteins reactive to the antibody were located in the mitochondria (Fig. 2C). These results indicate that SP-22 protein is located in mitochondria of culture cells, which is consistent with the results of the biochemical analysis by Watabe et al. (8).
Induction of the SP-22 Protein by Oxidative Stress-Exposing BAEC to Fe 2ϩ /DTT, which produces the hydroxyl radical through the Fenton reaction (24,25), increased the expression of SP-22 protein in a time-dependent manner, with the maximal elevation occurring after 24 h of exposure. On the other hand, the expression of MnSOD was induced only slightly in this condition (Fig. 3). SP-22 expression was completely suppressed with Fe chelators, such as deferoxamine and diethylene triamineacetic acid (Table I). Table II summarizes  hanced SP-22 protein by about 50 -100% increase. However, diethylmaleate and buthionine sulfoximine, which are glutathione depletors that have been shown to induce other thioredoxindependent peroxide reductase proteins (12), exhibited a rather poor induction of SP-22 protein. Rotenone, paraquat, 1-methyl-4-phenylpyridinium, antimycin A, and KCN, which are respiratory chain inhibitors that increase ROI generation in mitochondria (29), also enhanced SP-22 protein by 1.6 -3.2fold. ROI generation in SMP treated with antimycin A and rotenone showed 40 -100% increase compared with untreated SMP (Table III). Based on the rate of superoxide generation by antimycin A-treated SMP (4.1 nmol/min/mg protein), the antimycin A-treated confluent BAEC (10 7 cells) produced superoxide roughly at a rate of 0.13 nmol/min/mg protein (assuming that the protein content of SMP in 10 7 cells is about 0.065 mg), which is about 800-fold lower than the rate of superoxide gen-eration produced by exogenous xanthine/xanthine oxidase (50 milliunits/ml), whereas the SP-22 protein induced by antimycin A treatment was about 3-fold higher than that induced by the xanthine/xanthine oxidase treatment, suggesting that superoxide production in mitochondria induced SP-22 expression more effectively.
Respiratory chain inhibitors also deplete ATP by blocking oxidative energy metabolism in mitochondria (30,31). To examine the possibility that the stress of ATP depletion may enhance SP-22 expression, we compared the protein levels of SP-22 between BAECs grown in medium that contained either 5 mM glucose or 20 mM 2-deoxyglucose (a glycolytic inhibitor) instead of glucose. Six hours after the addition of glucose or 2-deoxyglucose, the ATP level of BAEC incubated with 20 mM 2-deoxyglucose was reduced to 48.1 Ϯ 7.8% of that in the control (with 5 mM glucose). The expression of SP-22 protein did not differ between two different media (103 Ϯ 4%, in 20 mM 2-deoxyglucose medium versus 5 mM glucose). This result suggests that ATP depletion per se does not influence the induction of SP-22 expression in BAEC.
Induction of SP-22 mRNA-The induction of SP-22 protein in BAEC exposed to oxidative stress was also examined in terms of the mRNA level. Total RNA was obtained from BAEC exposed to oxidants for 0, 0.5, 1, 3, 6, and 24 h, and the amount of SP-22 mRNA was analyzed by Northern blotting and semiquantitative RT-PCR. In Northern blotting, SP-22 mRNA began to increase after 0.5 h of exposure to Fe 2ϩ /DTT or antimycin A, reached a maximal level at 3 h, and decreased to the initial level after 24 h (Fig. 4A). In semiquantitative RT-PCR, almost the same results were obtained (Fig. 4B). To investigate the mechanism of the apparent induction of SP-22 in antimycin A-treated BAEC, we examined the effect of actinomycin D on the increase in SP-22 mRNA expression after antimycin A treatment. As shown in Fig. 4C, actinomycin D suppressed the antimycin A-mediated increase in SP-22 mRNA expression to the control level. This result indicates that the observed increase in SP-22 mRNA was due at least in part to the transcriptional activation of SP-22 mRNA.
Effect of Preconditioning with a Low Dose of Oxidant-The pretreatment of BAEC with a low dose of oxidant (3.3 M Fe 2ϩ , 330 M DTT), which did not affect cell viability, increased the level of SP-22 protein about 3.3-fold compared with that in control cells (Fig. 5A). In contrast to the untreated control, the cell viability of pretreated BAEC did not markedly decrease with subsequent treatment with a high dose of oxidant (Fig.  5B). These results indicate that BAEC in which SP-22 protein had been induced by preconditioning with a mild oxidant became tolerant to subsequent intense oxidative attack.    inhibitors SMP were prepared from BAEC as described under "Experimental Procedures." Superoxide production by SMP was determined by the oxidation of adrenaline to adrenochrome and corrected by subtracting the rate in the presence of 10 g/ml superoxide dismutase, as described by Takeshige et al. (19). The control system contained 0.5 mg/ml SMP and 1 mM adrenaline, 0.25 M sucrose, and 50 mM Hepes/NaOH, pH 7.5, in 1 ml. After preincubation for 5 min, the reaction was started under the conditions indicated adding of 0.2 mM NADPH, and adrenochrome formation was determined at 485-575 nm by a dual-wavelength spectrophotometer. The results represent the means Ϯ S.E. of three separate experiments.

Conditions
Adrenochrome formation nmol/min/mg Antisense ODN for SP-22-We examined the cell viability of BAEC with a decreased level of SP-22 due to the treatment with antisense ODNs against SP-22 mRNA. After BAEC had been exposed to SP-22 antisense ODNs three times, the protein levels of SP-22 in BAEC were examined by immunoblotting. The amount of SP-22 in BAEC treated with antisense 1 and 2 decreased to 43.9% and 72.1%, respectively, of that in control cells, whereas that of the sense ODN-treated BAEC remained close to the level in the control (Fig. 6A). As shown in Fig. 6B, oxidative stress (3.3 M Fe 2ϩ , 750 M DTT) markedly decreased the cell viability of the antisense 1 ODN-treated BAEC (about 77% decrease), whereas the same oxidative stress produced only a slight decrease (ϳ20%) in cell viability for the control and sense ODN-treated cells. Treatment of BAEC with antisense or sense ODNs by itself did not appreciably affect cell viability (Fig. 6B). These results indicate that BAEC with a lower level of SP-22 protein were more susceptible to oxidative stress.
Induction of SP-22 in Infarcted Rat Heart-To examine the in vivo induction of SP-22 protein, an immunohistochemical study was carried out using an experimental model of myocardial infarction in rat heart. Immunohistochemistry with anti-SP-22 and anti-HNE-protein adducts (32) showed that immunoreactive signals of both antibodies were markedly increased in the infarcted zone of the rat heart (Fig. 7, B and C), indicating that lipid peroxidation occurred in the infarct zone and SP-22 in rat heart was specifically induced corresponding to the area exposed to oxidative stress. In sham-operated animals, immunopositive signals against both antibodies were less evident, and there were no differences between the infarct and non-infarct zones (data not shown).

DISCUSSION
In cardiovascular systems, including the heart and vascular endothelial cells, mitochondria are particularly important for generating ATP. At the same time, the mitochondrial electron transport system consumes a great deal of oxygen and produces ROIs (1, 2). The generation of ROIs in cardiovascular systems has harmful effects on tissues and leads to atherosclerosis (33), ischemia reperfusion injury (34), and hypertension (35). ROIs are removed by various scavenging systems, such as CuZnSOD or MnSOD (36), glutathione peroxidase (37), and catalase (38). SOD dismutates superoxide radical and produces hydrogen peroxide. Although catalase removes hydrogen peroxide in the cytosolic compartment, there has been no report on mitochondrial catalase, except for that in rat heart mitochondria (39). SP-22 is a member of the thioredoxin-dependent peroxide reductase family (11), and by accepting electrons from the NADPH-thioredoxin-thioredoxin reductase system, it removes peroxides in vitro (15). In this study, the induction of SP-22 mRNA and protein were enhanced by various oxidative stresses, including mitochondrial respiratory chain inhibitors, in vascular endothelial cells. It is well known that respiratory chain inhibitors increase ROIs in mitochondria by blocking electron transport (29,30), and we confirmed that antimycin A and rotenone actually increased superoxide production in submitochondrial particles of BAEC. The result that intramitochondrial ROI generation may be more effective for enhancing the expression of SP-22 protein than the extracellular admin-istration of an oxidative stressor can be explained by supposing that a redox sensor, which exists in or near mitochondria, specifically responds to intramitochondrial ROIs. Aconitase, an iron-sulfur protein located in both cytosol and mitochondria, has been reported to serves as the redox sensor and play a regulatory role in gene expression (40).
The cytosolic members of the thioredoxin-dependent peroxide reductase family have been reported to be induced by oxidative stresses. MSP23 is inducibly expressed when mouse peritoneal macrophages are exposed to diethylmaleate (12). This expression has also been reported to be enhanced by exposure to oxidized low density lipoprotein, hydrogen peroxide, and heavy metals (12,41,42). Thiol-specific antioxidant from yeast is also induced by high concentrations of mercaptoethanol, iron and oxygen (43). Yeast thiol-specific antioxidant exerts a protective effect in vitro on mitochondrial oxidative damage (7). Regarding protein induction by oxidative stresses, SP-22 seems to be induced in a manner similar to other cytosolic members of the thioredoxin-dependent peroxide reductase family, except for its response to sulfhydryl-reactive agents. Diethylmaleate and buthionine sulfoximine induce the expres- sion of MSP23 in the cytosol (12), but not of SP-22 protein in the mitochondria. This discrepancy could be explained by the difference of the subcellular localization of the two proteins. Extracellularly added sulfhydryl agents has been reported to exert a different effect on the redox states of glutathione in the cytosol and mitochondria (44).
It is well known that MnSOD (45) and glutathione peroxidase (37) function as defensive enzymes against ROIs in mitochondria. Therefore, we were interested in the relative contribution of SP-22 to the antioxidant function in mitochondria. BAEC in which the expression of the SP-22 protein was enhanced by exposure to a radical-generating system, Fe 2ϩ /DTT, which only slightly induced MnSOD (Fig. 3), acquired tolerance to a higher concentration of Fe 2ϩ /DTT (Fig. 5). Furthermore, BAEC that had been depleted of SP-22 by antisense phosphorothioate ODN, were labile to the Fe 2ϩ /DTT system (Fig. 7).
These results suggest that SP-22 plays an integral role in antioxidant systems of BAEC. We found that thioredoxin and thioredoxin reductase, which participate in a cellular antioxidant defense (46) and are essential for SP-22 activity, are also located in mitochondria of BAEC. 2 Previous reports have mentioned that some cytokines such as tumor necrosis factor-␣ and ␤ (TNF-␣ and ␤) and interleukin-1␣ and ␤ regulated the expression of MnSOD (47). Shull et al. reported xanthine/xanthine oxidase increased MnSOD mRNA in epithelial cells without induction of catalase, CuZn-SOD, and glutathione peroxidase mRNAs (48). Yamamoto et al. reported that the promotor region of murine MER5 protein (homolog of bovine SP-22) had sites for several DNA-binding proteins (AP1, AP2, SP1) and MER5 protein was enhanced after induction of differentiation by dimethyl sulfoxide (14). Kang et al. recently reported overexpression of peroxiredoxin II (thioredoxin-dependent peroxide reductase) prevented the TNF-␣-induced NFB activation by removing intracellular hydrogen peroxide (49). These results suggested that the expression of cellular antioxidant enzymes, such as the thioredoxindependent peroxide reductase and MnSOD, can be regulated by a cytokine/ROI-mediated signal transduction. We also observed SP-22 mRNA was induced by TNF-␣. 3 In vivo, the expression of SP-22 was enhanced in an experimental model of rat myocardial infarction. The induction of SP-22 expression peaked at 24 h after ligation, and rapidly decreased at 48 h when necrosis reached a peak. At 24 h after coronary occlusion, the process of infarction is still incomplete in the rat heart (23). With the progression of ischemia, ROIs such as superoxide, hydrogen peroxide, and hydroxyl radical are generated through several pathways, including the leakage of electrons from mitochondria, which are especially abundant in the heart (29,50), the metabolism of arachidonic acid released by activated phospholipases (51), and the increase in xanthine oxidase activity, which produces superoxide (52). It has been reported that exogenous SOD and/or catalase reduce the extent of myocardial infarction (53,54), suggesting that they may protect against the progression of myocardial infarction. The present results obtained in experimental myocardial infarction also suggest that SP-22 may participate in protecting against myocardial infarction.