Introduction

Top Abstract Introduction Experimental Procedures Results Discussion References

The Na⁺-Ca²⁺ exchanger (NCX) is involved in regulation of intracellular Ca²⁺ concentration ([Ca²⁺]_i) via the forward mode (Ca²⁺ extrusion) or the reverse mode (Ca²⁺ influx) (Hryshko and Philipson, 1997; Matsuda et al., 1997). Protocols to inhibit selectively NCX activity are useful for investigating the roles of the exchanger. The antisense strategy has been successfully used in vitro (Lipp et al., 1995; Matsuda et al., 1996; Takuma et al., 1996a; Slodzinski and Blaustein, 1998; Van Eylen et al., 1998; White et al., 1998; Takahashi et al., 1999), but not in vivo. A variety of compounds are reported to inhibit NCX activity, but their use is limited because of lack of selectivity for NCX (Kaczorowski et al., 1989). Although XIP, a synthetic peptide, is the most selective inhibitor of NCX (Li et al., 1991), it does not appear to permeate through the cell membranes. In the circumstances, 2-[2-[4-(4-nitrobenzyloxy)phenyl]ethyl]isothiourea (KB-R7943) was synthesized as a potent inhibitor of NCX (IC₅₀ for Na⁺-dependent Ca²⁺ uptake, about 2 µM) (Iwamoto et al., 1996; Watano et al., 1996). Since KB-R7943 at 10 µM did not affect Na⁺/H⁺ exchange, dihydropyridine (DHP)-sensitive Ca²⁺ uptake, passive Na⁺ uptake, sarcolemmal Ca²⁺-ATPase, sarcoplasmic reticulum Ca²⁺-ATPase, and Na⁺,K⁺-ATPase (Iwamoto et al., 1996), it has been used as a selective inhibitor of NCX. However, it remains to be determined whether KB-R7943 is indeed a selective inhibitor of NCX. Sobolevsky and Khodorov (1999) reported that KB-R7943 blocked N-methyl-D-aspartate channels in acutely isolated hippocampal neurons. We have recently found KB-R7943 at 10 µM significantly inhibited the store-operated Ca²⁺ entry (SOCE; formerly referred to as capacitative Ca²⁺ entry) (Arakawa et al., 2000). Furthermore, Watano et al. (1999) could not exclude the possibility that the effect of KB-R7943 on ouabain-induced arrhythmias might be mediated by inhibition of voltage-gated Na⁺ channels. A more selective inhibitor of NCX is required for studies on physiological and pathological roles of NCX.

We found that Ca²⁺ paradox-like phenomenon occurred in cultured rat astrocytes: a persistent increase in [Ca²⁺]_i followed by delayed cell death was observed when the cells were incubated with Ca²⁺-containing medium after exposure to Ca²⁺-free medium (Matsuda et al., 1996). Furthermore, we have recently shown that the Ca²⁺ reperfusion induced apoptosis and reactive oxygen species (ROS) production might be involved in the cell death (Takuma et al., 1999). The injury is reduced by heat shock protein (Takuma et al., 1996b), calcineurin inhibitors (Matsuda et al., 1998), and anti-ischemic drugs (Takuma et al., 2000a,b). These findings, together with the previous report that a similar paradoxical change in extracellular Ca²⁺ concentration occurs in ischemic brain tissue (Siemkowicz and Hansen, 1981; Silver and Erecinska, 1992; Kristian et al., 1994), suggest that the Ca²⁺ paradox-like injury is an in vitro model of cerebral ischemia/reperfusion injury.

We report here that 2-[4-[(2,5-difluorophenyl)methoxy]phenoxy]-5-ethoxyaniline (SEA0400), a newly synthesized compound, is the most potent and selective inhibitor of NCX so far reported. This compound (Fig. 1) has been identified by screening a compound library for inhibition of Na⁺-dependent Ca²⁺ uptake into isolated cardiac sarcolemmal vesicles and cultured astrocytes. In addition, we demonstrate that SEA0400 protects astrocytes against Ca²⁺ paradox-like injury and reduces cerebral ischemic damage in rats with a transient middle cerebral artery occlusion.

View larger version (7K): [in this window] [in a new window]   Fig. 1.   Chemical structure of SEA0400.

	Experimental Procedures

Top Abstract Introduction Experimental Procedures Results Discussion References

Materials. Drugs were obtained from the following sources: fetal calf serum (FCS), mouse anti-human glial fibrillary acidic protein monoclonal antibody, isolectin B₄ (biotin labeled), and 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrazolium bromide (MTT), Sigma (St. Louis, MO); mouse anti-microtubule-associated protein-2 antibody, Chemicon International, Inc. (Temecula, CA); fluorescein-conjugated goat anti-mouse IgG antibody, Organon Teknika N.V.-Cappel Product (West Chester, PA); 2',7'-dichlorofluorescein diacetate (H₂DCF-DA), and Hoechst 33342, Molecular Probes, Inc. (Eugene, OR); and Eagle's minimum essential medium (MEM), Nissui Pharmaceutical Co., Ltd. (Tokyo, Japan). ⁴⁵Ca (5-50 mCi/mg) was purchased from Amersham (Tokyo, Japan). SEA0400 and KB-R7943 were synthesized in Taisho Pharmaceutical Co., Ltd. (Saitama, Japan). All other chemicals used were of the highest purity commercially available. In the in vitro studies, SEA0400 and KB-R7943 were dissolved in dimethyl sulfoxide (final concentration 0.1%). In an in vivo experiment, SEA0400 was administrated as a lipid emulsion containing 20% soybean oil, and MK-801 was dissolved in saline.

Cell Culture. Astrocytes were isolated from cerebral cortices of 1-day-old Wistar rats as previously reported (Takuma et al., 1994; Matsuda et al., 1996). Briefly, the tissue was dissociated with dispase and cultured in MEM containing 10% FCS and 2 mM glutamine. Cells were placed in 75 cm² tissue culture flasks and split once upon confluency (14-21 days). The cells (5 × 10⁴ cells/well) were plated on glass coverslips attached to silicon walls and grown for 1 to 2 days in experiments measuring fura-2 fluorescence. In experiments measuring NCX activity, the cells were plated in 24-well plastic tissue culture plates and grown for 14 to 20 days. The astrocytes in the plastic plates consisted of >90% flat polygonal astrocytes (type 1 astrocytes), as confirmed by phase-contrast microscopy and positive immunostaining with anti-glial fibrillary acidic protein antibody, and did not contain neuronal cells, as determined by negative immunostaining with anti-microtubule-associated protein-2 antibody (Sakaue et al., 2000). The dissociated cortical neurons were prepared from 18-day rat fetuses (Sakaue et al., 2000) and seeded onto culture plates. After 48 h, the cells were treated with 10 µM Ara-C for 48 h and cultured in Dulbecco's modified MEM containing 10% FCS for 6 to 10 days. The cell cultures consisted of more than 90% neurons as determined by positive immunostaining with anti-microtubule-associated protein-2 antibody. Microglia were obtained from cerebral cortices of 1-day-old Wistar rats (Kitanaka et al., 1996; Koyama et al., 2000). Briefly, mixed primary cultures were grown on 75 cm² tissue culture flasks in MEM containing 10% FCS and 2 mM glutamine. After 10 days in primary culture, microglial cells were harvested by mild shaking (140 rpm, 2 h) and plated on 48-well plastic tissue culture plates (1 × 10⁵ cell/well). The cells were incubated for 30 min, washed twice, and further incubated in MEM containing 10% FCS for more than 8 h. The cell cultures consisted of more than 95% microglia as determined by positive isolectin B₄ staining (Takuma et al., 2000b).

Na⁺-Ca²⁺ Exchange Activity. Na⁺-Ca²⁺ exchange activity was determined by assaying Na⁺-dependent ⁴⁵Ca²⁺ uptake as reported previously (Takuma et al., 1994). Briefly, the cells were preincubated in Hanks' balanced saline solution (HBSS) for 20 min, and the medium was switched to HBSS containing ⁴⁵Ca²⁺ and incubated for 5 min. To increase intracellular Na⁺ concentration, 1 mM ouabain plus 20 µM monensin (for astrocytes and microglia) and 10 µM monensin (for neurons) were used. Monensin was added simultaneously with the isotope. Ouabain was added 5 min before monensin in astrocytes and microglia. NCX inhibitors were added 5 min before monensin and present during ⁴⁵Ca²⁺ uptake reaction.

Binding Assays to Various Receptors, Ion Transporters, and Ion Channels. The binding assays were performed as a contract study by CEREP (Celle l'Evescault, France). The assays were performed under experimental conditions as shown in Tables 1 and 2. Na⁺/H⁺ exchange (Jean et al., 1986), Na⁺,K⁺-ATPase (Fiske and Subbarow, 1925), Ca²⁺-ATPase (Jean and Klee, 1986), phospholipase A₂ (Katsumata et al., 1986), phospholipase C (Nakanishi et al., 1985), 5-lipoxygenase (Coffey et al., 1992), inducible nitric-oxide synthetase (Estrada et al., 1992), and constitutive nitric-oxide synthetase (cNOS) (Bredt and Snyder, 1990) activities were determined as reported previously.

SOCE. SOCE was determined by measuring cyclopiazonic acid (CPA)-induced change in [Ca²⁺]_i in cultured astrocytes (Arakawa et al., 2000). [Ca²⁺]_i was determined as reported previously (Takuma et al., 1994; Sakaue et al., 2000). Briefly, astrocytes were perfused with HBSS consisting of (in mM): NaCl, 137; KCl, 5.4; CaCl₂, 1.3; MgCl₂, 0.49; MgSO₄, 0.406; Na₂HPO₄, 0.335; KH₂PO₄, 0.44; NaHCO₃, 4.17; and glucose, 5.55. Fura-2 fluorescence (510 nm emission excited by 340 and 380 nm illumination) from cells, as well as background fluorescence, was imaged using an ARGUS-HiSCA image processor (Hamamatsu, Japan). Since CPA-induced Ca²⁺ signal varied from cell to cell, the effect of drug on SOCE was assessed in each cell; the change in [Ca²⁺]_i after drug application was measured as previously reported (Arakawa et al., 2000).

Ca²⁺ Paradox-Like Injury. Reperfusion experiments were carried out using confluent astrocytes in fetal calf serum-free medium. The cells were washed and exposed to Earle's solution (control) and Ca²⁺-free Earle's solution (Ca²⁺ paradox) for 30 min, and then incubated with normal Earle's solution for 3 days (for MTT assay) and 5 days (for DNA ladder and Hoechst 33342 staining) (Matsuda et al., 1996; Takuma et al., 1999). In an experiment, the cells were treated with thapsigargin at 100 nM for 3 days. Cell viability was measured by a colorimetric assay using MTT (Matsuda et al., 1996). MTT reduction activity was found to be proportional to the number of cells (Matsuda et al., 1993).

Measurement of ROS Production. Cells, plated in 96-well plastic tissue culture plates, were incubated at 37°C for 30 min in normal or Ca²⁺-free HBSS containing 10 µM H₂DCF-DA and 0.25 µg/ml Cremophor EL, and then rinsed twice with normal HBSS to remove excess dye. The cells were reperfused in normal HBSS for 1 h, and the conversion of H₂DCF-DA to its fluorescent product dichlorofluorescein by ROS, presumably H₂O₂ and hydroxyl radical, was determined with excitation at 485 nm and emission at 535 nm using a Wallac Multilabel counter (Behl et al., 1994; Takuma et al., 1999). ROS production is expressed as a percentage of control cells. The linearity and sensitivity of ROS assay were confirmed using H₂O₂ prior to the experiment.

Analysis of the DNA Ladder. The cells collected by centrifugation were suspended in 10 mM Tris-HCl buffer (pH 8.0) containing 1 mM EDTA, 0.5% N-lauroylsarcosine, and 0.2 mg/ml proteinase K, and incubated at 37°C overnight. DNA was extracted by phenol/chloroform (1:1; v/v) and precipitated with ethanol. The pellet was dissolved in the Tris-EDTA buffer containing 0.5 mg/ml RNase A and incubated at 37°C for 30 min to digest RNA. Equal amounts of DNA samples were subjected to 1.8% agarose gel electrophoresis. DNA in the gel was stained with ethidium bromide and photographed (Takuma et al., 1999).

Hoechst 33342 Staining. The cells, plated on a chamber slide, were fixed with 10% formaldehyde and stained with Hoechst 33342 as previously reported (Takuma et al., 1999). An inverted microscope (Olympus, IX70) equipped with a reflected fluorescence illuminator (Olympus, IX-FLA) and a 40× objective lens was used to visualize individual nuclei.

Focal Cerebral Ischemia. The effect of SEA0400 on ischemia/reperfusion injury was performed as a contract study by Panapharm Laboratories Co., Ltd. (Uto, Kumamoto, Japan). Male Sprague-Dawley rats, weighing 290 to 320 g, were used for the study. Temporary focal ischemia was induced by intraluminal vascular occlusion with a suture as described previously (Longa et al., 1989; Kuge et al., 1995). Rats were anesthetized with isoflurane in a mixture of 70% N₂O and 30% O₂, and mechanically ventilated during surgery. Rectal temperature was controlled at 37°C throughout the experiment with a feedback-regulated heating lamp. The femoral vein was cannulated to infuse drugs. A 19-mm length of 4-0 surgical monofilament nylon suture with its tip rounded by heating and coated with silicone was inserted into the external carotid artery stump and carefully advanced from the carotid bifurcation to occlude the origin of the right middle cerebral artery (MCA). During MCA occlusion, the rats were without anesthesia. After 2 h of MCA occlusion, the suture was withdrawn to allow reperfusion. Thirty minutes after MCA occlusion, the neurological status (failure to extend forepaw) of the rats was assessed. Rats were sacrificed at 22 h after reperfusion. The brains were quickly removed and coronally sectioned in 2-mm-thick slices. Slices were stained with 2% 2,3,5-triphenyltetrazolium chloride and then photographed. The infarct volumes were expressed in absolute terms (mm³). SEA0400 was i.v. injected at 1 and 3 mg/kg immediately after MCA occlusion, and then infused at 1 and 3 mg/kg/h for 2 h. MK-801 was i.p. injected at 3 mg/kg immediately after MCA occlusion.

Blood Pressure and Regional Cortical Blood Flow. Male Sprague-Dawley rats, weighing 289 to 325 g, were anesthetized with 1 to 2% halothane. A catheter was inserted into the femoral artery and connected to a pressure transducer to record blood pressure. Regional cortical blood flow was measured by a laser Doppler flowmeter (Neuroscience, Tokyo, Japan; FLO-N1), with probe placement at 2 mm posterior and 6 mm lateral to the bregma, as previously reported (Toung et al., 1999). SEA0400 or its vehicle with an equivalent volume was i.v. injected at 3 mg/kg and then infused at 3 mg/kg/h for 2 h under normal conditions without MCA occlusion.

Statistical Analysis. Statistical analysis of the experimental data was carried out by Student's t test or Dunnett's test, using a software package (StatView) for Apple Macintosh. Values of P < 0.05 were considered to be significant, and results were expressed as means ± S.E.

	Results

Top Abstract Introduction Experimental Procedures Results Discussion References

Effects on NCX Activity. Figure 2 shows the effects of SEA0400 and KB-R7943 on Na⁺-dependent ⁴⁵Ca²⁺ uptake in cultured neurons, astrocytes, and microglia. IC₅₀ values of SEA0400 were 33 nM (neurons), 5.0 nM (astrocytes), and 8.3 nM (microglia), and those of KB-R7943 were 3.8 µM (neurons), 2.0 µM (astrocytes), and 3.1 µM (microglia). In these cells, SEA0400 was >100 times more potent than KB-R7943 in inhibiting the NCX activity.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Effects of SEA0400 and KB-R7943 on Na+-dependent 45Ca2+ uptake in cultured neurons, astrocytes, and microglia. A, neurons; B, astrocytes; C, microglia. SEA0400 (open symbols) and KB-R7943 (closed symbols) at the indicated concentrations were added 5 min before 45Ca2+. Results are means ± S.E. of 4 wells (A), 4 to 12 wells (B), and 14 to 15 wells (C).

Selectivity. We have recently reported that KB-R7943 inhibited not only NCX, but also SOCE in cultured astrocytes (Arakawa et al., 2000). Figure 3 shows the effect of SEA0400 on SOCE in cultured astrocytes. SEA0400 at the concentrations up to 3 µM did not affect SOCE, although it was inhibited at 10 µM.

View larger version (26K): [in this window] [in a new window]   Fig. 3.   Effect of SEA0400 on SOCE in cultured astrocytes. Left, a typical CPA-induced Ca2+ signal in astrocytes, and the horizontal bar indicates the presence of 10 µM CPA. Right, dose response of the effect of SEA0400 on the Ca2+ signal. SEA0400 at the indicated concentrations was added 5 min after addition of CPA. Results are means ± S.E. of 25 to 38 cells. **P < 0.01, compared with the control (Dunnett's test).

Ca²⁺ Paradox-Like Injury. In cultured astrocytes, SEA0400 attenuated Ca²⁺ paradox-like injury at concentrations more than 0.1 µM, although it did not affect thapsigargin-induced cell injury (Fig. 4). Similar protection was observed by KB-R7943 at the concentrations more than 1 µM (data not shown). SEA0400 reduced paradoxical Ca²⁺ challenge-induced ROS production in a dose-dependent manner (Fig. 5). Figures 6 and 7 show the effect of SEA0400 on paradoxical Ca²⁺ challenge-induced apoptosis in cultured astrocytes. SEA0400 decreased the DNA ladder formation in a dose-dependent manner (Fig. 6). Hoechst 33342 staining showed that SEA0400 at 1 µM blocked almost completely nuclear condensation (Fig. 7).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   Effect of SEA0400 on Ca2+ paradox-like injury (A) and thapsigargin-induced injury (B) in cultured rat astrocytes. Open and closed circles indicate control and Ca2+ paradox- or thapsigargin-treated cells, respectively. The indicated concentrations of drugs were added at 10 min before Ca2+ reperfusion or thapsigargin treatment and present until assay. Results are means ± S.E. of eight wells obtained from four separate experiments. **P < 0.01, significant from control; P < 0.01, significant from Ca2+ paradox without SEA0400 (Dunnett's test).

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Effect of SEA0400 on Ca2+ reperfusion-induced ROS production in astrocytes. ROS was measured in the cells reperfused for 1 h after exposure to normal (control, open circles) or Ca2+-free Earle's solution (Ca2+ paradox, closed circles). SEA0400 at the indicated concentrations was added 10 min before Ca2+ reperfusion and present until assay. Results are means ± S.E. of 10 wells obtained from two separate experiments. **P < 0.01, significant from control; P < 0.01, significant from Ca2+ paradox without SEA0400 (Dunnett's test).

View larger version (127K): [in this window] [in a new window]   Fig. 6.   Effect of SEA0400 on Ca2+ reperfusion-induced DNA fragmentation in astrocytes. DNA was extracted from the cells reperfused for 5 days after exposure to Ca2+-free medium. SEA0400 at the indicated concentrations was added 10 min before Ca2+ reperfusion and present until assay. A typical experiment is shown. M, 100-base pair marker.

View larger version (199K): [in this window] [in a new window]   Fig. 7.   Effect of SEA0400 on Ca2+ reperfusion-induced nuclear condensation in astrocytes. Nuclear condensation was examined by Hoechst 33342 staining. Cells were exposed to normal (A, C) or Ca2+-free Earle's solution (B, D) for 30 min, and incubated with Earle's solution for 5 days. SEA0400 at 1 µM was added 10 min before Ca2+ reperfusion and present for 5 days (C, D). A typical experiment is shown.

Cerebral Ischemia. SEA0400 did not affect the mean blood pressure, the regional cortical blood flow during the 2-h observation period (Table 3), and body temperature (data not shown). Infarcts were assessed by 2,3,5-triphenyltetrazolium chloride staining 24 h after 2 h of MCA occlusion in the cerebral cortex and striatum. The infarct areas were more predominant in the cerebral cortex than in the striatum. Figure 8 shows the effects of SEA0400 and MK-801 on the infarct volume in the cerebral cortex and striatum. SEA0400 (3 mg/kg bolus i.v. + 3 mg/kg/h for 2-h continuous infusion) attenuated the infarct volume in the cerebral cortex and striatum. In contrast, MK-801 (3 mg/kg i.p.) showed a protective effect only in the cerebral cortex.

View larger version (35K): [in this window] [in a new window]   Fig. 8.   Effects of SEA0400 and MK-801 on infarct volume in the cerebral cortex (A) and striatum (B) after MCA occlusion in rats. Results are means ± S.E. of 14-17 animals. *P < 0.05, compared with vehicle (Dunnett's test); **P < 0.01, compared with saline (Student's t test).

	Discussion

Top Abstract Introduction Experimental Procedures Results Discussion References

The present study shows that SEA0400 is an extremely potent and selective inhibitor of NCX. SEA0400 inhibited NCX activity in neurons, astrocytes, and microglia, with IC₅₀ values of 5 to 33 nM. We have found in the separate experiment that SEA0400 also inhibited NCX activity in dog sarcolemmal vesicles and cultured rat myocytes: the IC₅₀ values were 90 and 92 nM, respectively. These results indicate that SEA0400 inhibits markedly NCX activity in cells and cell-free systems. In view of the previous reports (Iwamoto and Shigekawa, 1998; Sakaue et al., 2000), a slight difference in the potency of SEA0400 among preparations may be due to that in the sensitivity of NCX isoforms to the inhibitor. With respect to selectivity, SEA0400 showed a slight affinity for the Na⁺ channel site 2, the Ca²⁺ channel DHP site, and the Ca²⁺ channel diltiazem site, but their IC₅₀ values were >360 times greater than that of NCX. This selectivity seems to be important for studies on the effect on ischemic injury, since Ca²⁺ and Na⁺ channel inhibitors attenuate cerebral ischemic injury (Gemba et al., 1993; Taylor and Meldrum, 1995). Furthermore, SEA0400 at the concentrations up to 3 µM did not affect other ion channels, receptors, and enzymes that were considered to be involved in key processes in ischemic diseases, including Na⁺/H⁺ exchange (Phillis et al., 1998), K⁺ channels (Bari et al., 1996; Fink et al., 1998), adrenoceptors (Zhu et al., 1998; Semkova and Krieglstein, 1999), adenosine receptors (Von Lubitz, 1999), glutamate receptors (Martin et al., 1998; Bordi and Ugolini, 1999), mACh receptors (Jiang et al., 2000), bradykinin receptors (Walker et al., 1995; Relton et al., 1997), LTB₄ receptors (Barone et al., 1992; Bonventre et al., 1997), PAF receptors (Bonventre et al., 1997), phospholipases (Bonventre et al., 1997; Farooqui et al., 1997), lipoxygenase (Rao et al., 1999), and inducible nitric-oxide synthetase (Del Zoppo et al., 2000). In contrast, KB-R7943 affected mACh, LTB₄ and PAF receptors, and NE transporter at 3 µM, corresponding to the IC₅₀ value for NCX inhibition. In addition, the previous study showed that KB-R7943 inhibited SOCE (Arakawa et al., 2000), whereas SEA0400 did not affect SOCE. Taken together, the present study suggests that SEA0400 is a valuable new tool for elucidating the pathophysiological roles of NCX.

Previous studies showed that the NCX inhibitor KB-R7943 attenuated ischemic injury in cardiac (Nakamura et al., 1998; Ladilov et al., 1999) and renal (Kuro et al., 1999) preparations. However, it is obscure whether NCX is involved in the injury, since the specificity of KB-R7943 for NCX is questionable as described above. We previously used an antisense strategy to study the involvement of NCX in Ca²⁺ paradox-like injury in cultured astrocytes (Matsuda et al., 1996). In the present study, we further examined whether the selective NCX inhibitor SEA0400 attenuates Ca²⁺ paradox-like injury in astrocytes. In astrocytes, SEA0400 attenuated Ca²⁺ reperfusion injury in a dose-dependent manner, although the concentration required for the protective effect was 10 times higher than that required for NCX inhibition. We have recently shown that Ca²⁺ paradox-like injury is mediated by an increase in [Ca²⁺]_i resulting in ROS production (Takuma et al., 1999). The present study showed that SEA0400 blocked ROS production in a dose-dependent manner. The significant effect of SEA0400 was also observed in DNA ladder formation and Hoechst 33342 staining, suggesting an antiapoptotic effect of the compound. It is likely that the compound inhibits the NCX-mediated Ca²⁺ influx, but not Ca²⁺-mediated processes, resulting in cell toxicity since SEA0400 does not affect thapsigargin-induced cell toxicity. It is considered that NCX inhibition further increases [Ca²⁺]_i in thapsigargin-treated cells, since the forward mode of NCX contributes to a recovery of the increased level of [Ca²⁺]_i to the resting level. However, SEA0400 did not aggravate thapsigargin-induced cell toxicity. The exact reason for the apparent discrepancy is not known. Iwamoto et al. (1996) reported that KB-R7943 selectively inhibits the reverse mode of NCX, but we found in a preliminary experiment that SEA0400 inhibited Na⁺-dependent Ca²⁺ influx and Na⁺-dependent Ca²⁺ efflux with a similar potency in dog cardiac sarcolemmal membrane vesicles. Alternatively, thapsigargin-induced increase in [Ca²⁺]_i may reach the level that caused maximal cell toxicity, and then the increase in [Ca²⁺]_i by inhibition of the forward mode of NCX may not aggravate cell toxicity. It should be noted that SEA0400 did not show any cell toxicity even when the cells were exposed to SEA0400 until the assay. This contrasted with 3,4-dichlorobenzamil, a NCX inhibitor: treatment with 3,4-dichlorobenzamil was limited for the first 1 h of Ca²⁺ repletion because of cell toxicity (Matsuda et al., 1996).

The present study provides the first evidence that a NCX inhibitor is effective in reducing focal cerebral ischemic lesions in rats. SEA0400, like MK-801, was effective in reducing infarct volume in the cerebral cortex after MCA occlusion. In this study, we used SEA0400 as a lipid emulsion, because of its low solubility in water. We found in separate experiments that SEA0400 was rapidly cleared from the plasma, and brain penetration of this compound was excellent. Maximal concentrations of SEA0400 in brain were about 2.7 and 7.3 µM when it was i.v. injected at 1 and 3 mg/kg, respectively. In view of the selectivity of SEA0400 as described above, it is likely that the compound inhibits selectively brain NCX in vivo and that NCX is involved in reperfusion injury in a MCA occlusion model. We found that the administration of SEA0400 did not alter the blood pressure and regional cortical blood flow in normal rats. This suggests that SEA0400 is not altering the ischemic insult by vascular mechanisms. The present in vitro study on Ca²⁺ paradox injury implies that glial mechanism may be involved at least partly in the in vivo effect of SEA0400. It is also considered that SEA0400 may affect indirectly the cerebral circulation after MCA occlusion. The precise mechanism underlying the protective effect of SEA0400 on the cerebral ischemic injury is not known. Studies on the therapeutic time window of SEA0400 in this model and the effects in permanent MCA occlusion are in progress.

In conclusion, the results of the present study demonstrate that SEA0400 is the most potent and selective inhibitor of NCX, and it has a beneficial effect in in vitro and in vivo models of cerebral ischemic injury.