- Email: [email protected]
Ca2+-dependent cytotoxicity of H2O2 in L929 cells: the role of H2O2-induced Na+-influx
Free Radical Biology & Medicine, Vol. 25, No. 6, pp. 712–719, 1998 Copyright © 1998 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/98 $19.00 1 .00
PII S0891-5849(98)00159-2
Original Contribution Ca21-DEPENDENT CYTOTOXICITY OF H2O2 IN L929 CELLS: THE ROLE OF H2O2-INDUCED Na1-INFLUX ASTRID JUSSOFIE, MICHAEL KIRSCH,
and
HERBERT
DE
GROOT
Institut fu¨r Physiologische Chemie, Universita¨tsklinikum, Essen, Germany (Received 20 January 1998; Revised 5 June 1998; Accepted 8 June 1998)
Abstract—We investigated the mechanism by which H2O2 mediates an increase in [Na1]i in L929 cells and the relevance of this Na1 load for H2O2-induced cell injury. [Na1]i increased early after exposure to H2O2 as monitored by fluorescence spectrophotometry of cells loaded with SBFI. The omission of Na1 from the incubation buffer significantly reduced H2O2-cytotoxicity. This protection could not be mimicked by inhibition of either the Na1/H1antiporter, the Na1/HCO32-cotransporter, or the Na1/K1/2Cl2-cotransporter by using Hoechst 694 (0.02 mM) or 4-acetamido-49-isothio-cyanatostilbene-2,29-disulfonic acid (SITS) (0.02 mM) or furosemide (1 mM) and bumetanide (0.5 mM). Only the blocker of the Na1/Ca21-exchanger bepridil (0.2 mM) significantly reduced H2O2-cytotoxicity but without interfering with the increase in [Na1]i. H2O2 caused a rapid and sustained increase in [Ca21]i, which was significantly reduced in bepridil pretreated cells and after replacing extracellular Na1 by choline. H2O2 was found to initiate a cellular uptake of unphysiological Ni21 by using Newport Green diacetate as fluorescent dye. Our data suggest that H2O2 mediates Na1-influx across the plasma membrane rather unspecifically than through specific transporters. The protective effect of bepridil against H2O2-cytotoxicity occurs as a consequence of a reduced cellular Ca21-uptake. We conclude that H2O2-mediated unspecific accumulation of Na1 seems to favor a Ca21-influx into the cells, which takes place on the Na1/Ca21-exchanger operating in reverse mode in exchange for Na1-efflux. Therefore, H2O2-induced cellular Na1 accumulation appears to play a permissive rather than a triggering role in H2O2-mediated cell injury. © 1998 Elsevier Science Inc. Keywords—Hydrogen peroxide, Cytotoxicity, Intracellular Na1, Ca21
be partly due to the activation of the Na1/H1-antiporter and the Na1/HCO32-cotransporter associated with an intracellular Na1 load [2]. Because menadione generates H2O2 in cells [3], we focus on the question whether H2O2 may also induce disturbances in cellular Na1 homeostasis necessary for H2O2-mediated cell injury. Results obtained in our lab [4], indicate that the mechanism of H2O2 injury in L929 cells is highly dependent on the metabolic state of these cells given by the presence or absence of glucose. In the presence of glucose, cell death was mainly mediated by iron consistent with the ironcatalyzed production of hydroxyl radicals (Fenton reaction) as evidenced by Farber and coworkers [5]. These extremely reactive species are assumed to produce most of the covalent modification and damage to macromolecules, including DNA, proteins, and lipid membranes [5,6]. However, in the absence of glucose, disruption of ion homeostasis could be demonstrated as the decisive event responsible for H2O2 cytotoxicity as proposed by
INTRODUCTION
The importance of Na1 in the control of cytosolic pH and1 cell volume is reflected in the dynamics of cellular Na regulation. Extracellular Na1 can enter cells through different transporters such as the Na1/H1-antiporter, the Na1/HCO32-cotransporter or the Na1/K1/2Cl2-cotransporter. The activity of these transporters depends on the maintenance of the sodium electrochemical gradient across the membrane. Under physiological conditions, the excess Na1 will be pumped out by the Na1/K1ATPase, therefore, maintaining Na1 homeostasis (Fig. 1) [1]. However, an uncontrolled and persistent elevation of cellular Na1-levels promotes cell injury. The loss of viability with menadione in hepatocytes is proposed to Address correspondence to: PD Dr. A. Jussofie, Institut fu¨r Physiologische Chemie, Universita¨tsklinikum Essen, Hufelandstr. 55, D-45122 Essen, Germany; Tel: 149-201-723-4105; Fax: 149-201723-5943. 712
Role of Na1 in H2O2-cytotoxicity
713
nel blockers used as furosemide, SITS, bumetanide, bepridil were from Sigma (Deisenhofen, Germany) except Hoe 694 which was a gift from Hoechst (Frankfurt/Main, Germany). Cell culture The murine fibroblast-derived cells L929 (American Type Culture collection, NCTC clone 929 of strain L) were grown in minimum essential medium (MEM) Eagle (Sigma) supplemented with 25 mM bicarbonate, 10% fetal bovine serum (Sigma), 2 mM L-glutamine (Sigma), and 50 mg/ml gentamycin (Gibco BRL) in plastic flasks (Falcon, Heidelberg, Germany) at 37°C in a humidified atmosphere of 5% CO2 in air. Cells were trypsinized (0.25% v/v) twice weekly. Those from passage 5 to 18 were used for experiments. Fig.1. Cellular sodium regulatory mechanisms. Na1-influx through the plasma membrane can be attained by the Na1/H1-antiporter, the Na1/ Ca21-antiporter, the Na1/K1/2Cl2-cotransporter and the Na1/HCO32cotransporter. Na1 efflux involves the Na1/K1-ATPase.
Orrenius and coworkers [7]. Although free Ca21 seems to be required for H2O2-mediated cell injury, these data do not rule out the possibility that cellular accumulation of Ca21 after H2O2 exposure is not the initiating event in the development of membrane damage, but rather a consequence of a preceding event such as a Na1 challenge outside the physiological range. In the present study, we have examined whether and to what an extent loss of Na1 homeostasis is a mechanism responsible for H2O2-mediated cell injury by using a fluorescence spectrophotometric assay with the fluorescence indicators SBFI/AM and Fura-2/AM to monitor intracellular levels of Na1 and Ca21 time dependently. These experiments were performed with L929 cells incubated without glucose to assure a maximal disturbance of cellular ion homeostasis under the experimental conditions used sufficient to underlie H2O2-mediated cell injury. MATERIALS AND METHODS
Materials The acetoxymethyl esters of SBFI, Fura-2 and Newport green were obtained from Molecular Probes (Leiden, The Netherlands). Hydrogen peroxide was purchased from Aldrich (Steinheim, Germany). Pyruvate, NADH, and ethylenediaminetetraacetic acid, disodium salt (EDTA) were obtained from Boehringer Mannheim (Mannheim, Germany). All other chemicals were obtained from Merck (Darmstadt, Germany). All the chan-
Procedures L929 cells were seeded at a density of 9 3 104 cells/0.9 ml of MEM in 12-well tissue culture plates (Falcon) or, for fluorescent measurements, on glass coverslips (28 mm diameter, Fa. Assistent, Germany) in petri dishes and allowed to grow for 2 days. For each experiment the cells were washed once either with Krebs-Henseleit buffer, pH 7.4, supplemented with 20 mM Hepes or with the sodium-free variant containing choline in exchange for Na1. The Krebs-Henseleit buffer consisted of (in mM): NaCl 115.0, NaHCO3 25.0, KCl 5.9, MgCl2 1.2, NaH2PO4 1.2, Na2SO4 1.2, CaCl2 2.5. For the experiments performed in the absence of Na1 the sodium-free variant had been used containing (in mM): KCl 5.9, MgSO4 1.2, H3PO4 1.2, choline-HCO3 25, CaCl 2.5, choline-Cl 117.0. The two buffers were saturated with 5% CO2 and 95% air at 37°C. The experiment was started by the addition of H2O2 to a final concentration of 0.75 mM. The Na1-channel-blockers furosemide (1 mM), bumetanide (0.5 mM), SITS (0.2 mM) , Hoe 694 (0.02 mM), and bepridil (0.2 mM) were added directly prior to H2O2-exposure. Bepridil and SITS were dissolved in Krebs-Henseleit buffer, pH 7.4; furosemide was dissolved in the same buffer but supplemented with 1 mM NaOH. Bumetanide was added as an ethanolic solution (final concentration 1%), Hoe 694 as a DMSOsolution (final concentration 1%). Cytotoxicity was determined by measuring intra- and extracellular lactate dehydrogenase (LDH) activity as described previously [8]. Results obtained using the LDH-leakage viability test were compatible with our data using viability staining such as trypan blue exclusion [9]; therefore, cell damage induced by H2O2 will be expressed in the Results section as loss of viability.
714
A. JUSSOFIE et al.
Fluorescence measurements The cover slips with the cells attached were washed twice with Krebs-Henseleit buffer or the sodium-free variant and transferred into a chamber (Pentz, chamber system, Heraeus, Hanau, Germany), which was mounted on the thermostatic table of an inverted microscope (Zeiss, Oberkochen, Germany). The cells were flushed with 5% CO2 and 95% air during loading with the fluorescent dye and the whole incubation period. To measure changes in the intracellular Ca21-concentration the Ca21 fluorescence indicator Fura-2/AM was added to the coverslips at a final concentration of 6 mM in the respective buffer and cells were returned to the 37°C/5% CO2 incubator for 60 min. Cells were then washed twice with the respective buffer. Fura-2 fluorescence was measured by using a videomicroscope system (Attofluor TM fluorescence analyser, Zeiss, Oberkochen). Cells were detected using 334 nm and 380 nm excitation and monitoring fluorescence by means of an emission barrier filter at 500 –530 nm. Calibration and calculation of free cytosolic Ca21 were performed according to the procedure described by Grynkiewicz and coworkers [10]. Rmax was obtained by adding 50 mM ionomycin, Rmin was evaluated by adding 10 mM EGTA. For [Na1]i-measurement cells were loaded with SBFI-AM for 75 min at 37°C in Krebs-Henseleit or sodium-free buffer both containing 20 mM SBFI and the nonionic detergent Pluoronic at a concentration of 0.06%. The excitation monochromators were set at 334 and 380 nm; emitted fluorescence was measured as described previously. The intracellular Na1 concentration was calibrated in situ by the method of Donoso and coworkers [11]. Two calibration solutions were used to contain: NaCl or KCl 140 mM, Hepes 10 mM, EGTA 1 mM, glucose 10 mM, gramicidin D 4 mM, monensin 10 mM, nigericin 10 mM; adjusted to pH 7.4 with either NaOH or KOH. The mixture of these two stocks solutions yielded calibration solutions with the following sodium concentrations: 17.5, 35, 52.5, 70, and 105 mM. To test the hypothesis concerning an H2O2 induced unspecific cation influx into cells Ni21 was added extracellularly at a final concentration of 10 mM. We used the cell permanent Newport Green diacetate as fluorescent indicator for nickel which exhibits an increase in fluorescence emission upon binding Ni21. Cells were loaded with Newport Green (final concentration 5 mM) for 15 min at 37°C in 5% CO2 and 95% air. The Ni21-influx was measured by ratio imaging of Newport Green fluorescence excited at 480 and 340 nm. Fluorescence was monitored as described.
Fig. 2. Toxicity of H2O2 to L929 cells incubated in the presence of Na1. Cells were exposed for 5 h to 0.75 mM H2O2 in Krebs-Henseleit buffer, pH 7.4, at 37°C in 5% CO2 and 95% air. For the experiments conducted without Na1, choline was used in exchange for Na1. At the times indicated cell injury was estimated by LDH leakage and trypan blue exclusion, as described in Materials and Methods. Data expressed as loss of viability are means 6 SEM of 6 independent experiments, each performed in triplicate.
RESULTS
Cytotoxicity of H2O2 L929 cells cultured for 2 days were exposed to 0.75 mM H2O2. Fig. 2 shows that this H2O2 treatment led to a large loss of viability when it was performed in KrebsHenseleit buffer, i.e., in the presence of Na1. The most pronounced reduction of surviving cells was observed during the second hour of exposure to H2O2 where cell death increased from 3 to 69%. After 3 h, almost all of the cells had already lost their viability. When the cells were exposed to H2O2 in Krebs-Henseleit buffer where Na1 was exchanged for choline the cytotoxicity of H2O2 was significantly reduced although cell death was still much higher than that observed in controls. Cell death remained low after 2 h, reaching only 11% but then increased steadily to 85% after 5 h.
Role of Na1 in H2O2-cytotoxicity
715
Fig. 3. Effect of H2O2 on the cytosolic Na1-concentration. Cytosolic Na1-levels were monitored in L929 cells loaded with the flurorescent dye SBFI/AM (20 mM for 75 min). After baseline fluorescence levels were monitored for 8 min 0.75 mM H2O2 and 1 mM ouabain were added to cells incubated at 37°C in 5% CO2 and 95% air with Krebs-Henseleit buffer, pH 7.4 at the time indicated by the arrow (A). The rise of [Na]i could not be blocked if the cells were preincubated with 0.2 mM bepridil for 8 min (B). Control cells were incubated either in Krebs-Henseleit buffer without H2O2 or with the Na1-free variant containing choline instead of Na1 in the presence of 0.75 mM H2O2. Data are means of 5 experiments 6 SEM each registered as the average fluorescence of 30 –50 cells.
Effect of H2O2 on the intracellular Na1 concentration
Effect of sodium channel blockers on H2O2-induced cell injury
Choline is a substitute for Na1. Therefore, we evaluated the effect of H2O2 on intracellular sodium concentration using the fluorescence ratio of SBFI as a measure of sodium levels. The addition of 0.75 mM H2O2 increased the intracellular Na1-concentration in a timedependent manner compared to both the untreated control in KHR-buffer or the H2O2-treated cells in choline buffer (Fig. 3). This rise of [Na1]i already began 4 min following the addition of H2O2 and continued for 60 min in a linear fashion reaching concentrations above 70 mM. This intracellular Na1-accumulation could not be mimicked when the Na1/K1-ATPase was inhibited by 1 mM ouabain, indicative of rather an increased Na1-influx than a decreased Na1-efflux elicited by H2O2. In the absence of H2O2 only a slight increase up to 32 mM was detectable. The H2O2-induced increase in Na1 concentration was dependent on extracellular Na1; when Na1 was replaced by choline the H2O2-treated cells maintained their basal Na1 level of 18 mM over the whole course of incubation.
Next we examined the effects of different sodium channel blockers such as furosemide [12] (1 mM), bumetanide [13] (0.5 mM), SITS [14] (0.2 mM), Hoe 694 (0.02 mM), and bepridil (0.2 mM) on H2O2-induced cell injury. None of the blockers tested was effective in blocking H2O2-induced cell damage except bepridil (Table 1). The co-administration of 200 mM bepridil produced only a 18% loss of viability after 2 h, compared to 61% obtained with cells solely treated with H2O2. Bepridil levels lower than 200 mM were effective in a concentration-dependent manner, whereas higher concentrations of the channel blocker did not additionally improve protection (data not shown). Bepridil is well known to block the Na1/Ca21-exchanger. In view of its protective effect on cells exposed to H2O2 we tested whether bepridil might reduce or delay the H2O2-induced increase in [Na1]i. However, bepridil failed to affect the H2O2-dependent Na1 uptake at all; the intracellular Na1 concentration determined in the presence of bepridil did not differ significantly from that
716
A. JUSSOFIE et al. Table 1. Effects of the Sodium Channel Blockers on the Cytotoxic Action of H2O2 in L929 Cells Loss of viability (% of total)
Conditions A. B. C. D. E. F. G. H. I. J.
Control H2O2 H2O2 1 1% DMSO H2O2 1 1% ethanol H2O2 1 1 mM NaOH H2O2 1 1 mM furosemide H2O2 1 0.5 mM bumetanide H2O2 1 0.2 mM sits H2O2 1 0.02 mM Hoe H2O2 1 0.2 mM bepridil
1h
2h
3h
4h
5h
2.4 6 0.26 4.4 6 0.9 4.0 6 1.1 3.7 6 0.9 3.9 6 0.8 5.0 6 1.2 4.9 6 1.1 6.0 6 0.9 4.9 6 0.8 5.3 6 1.3
3.2 6 0.5 61 6 3.5 55 6 2.9 59 6 3.2 60 6 3.6 62 6 2.8 56 6 3.1 73 6 2.9 64 6 3.0 18 6 3.0
3.6 6 0.5 89 6 2.5 75 6 3.2 88 6 3.5 87 6 3.8 83 6 3.7 88 6 3.4 93 6 1.9 91 6 1.8 67 6 2.1
3.6 6 0.8 93 6 3.0 88 6 1.9 93 6 2.1 92 6 2.5 90 6 2.3 92 6 2.5 93 6 2.1 94 6 1.7 89 6 1.6
3.4 6 0.6 94 6 2.1 91 6 2.2 95 6 1.9 95 6 2.0 92 6 1.8 95 6 1.8 94 6 1.5 95 6 1.6 92 6 0.8
Cells were grown for 2 days and exposed to 0.75 mM H2O2 in Krebs-Henseleit buffer, pH 7.4, at 37°C. After the exposure period indicated the loss of viability was determined by estimating LDH leakage and trypan blue exclusion. The results are means 6 SEM of five experiments. Student’s t-test for all time periods after 1 h of exposure to H2O2 demonstrated a significantly larger cell injury than control cells incubated without H2O2. The loss of viability in J was significantly ( p , .001, two-way ANOVA) lower than that observed in B.
in cells incubated without bepridil over the whole course of 90 min (Fig. 3). Effect of H2O2 on membrane permeability for Ni21 Because neither the H2O2-mediated cellular Na1-uptake could be blocked by bepridil nor a reduction of the cytotoxic action of H2O2 could be achieved with the other commonly used blockers of Na1-dependent transporters, we determined whether H2O2 produces an unspecific influx of Na1 across the plasma membrane possibly due to an increased permeability of cations. Therefore, we used unphysiological Ni21, which cannot move, per se, through the cellular channels when added extracellularly. To monitor the effect of H2O2 on Ni21 uptake we used the fluorescent dye Newport Green, which increases its fluorescence upon binding Ni21. As shown in Fig. 4, by the increasing ratio of 480/340 nm values of Newport Green fluorescence Ni21 comes into the cells that were exposed to H2O2, whereas Ni21 permeability of the untreated control cells was negligable. After the addition of H2O2, cells displayed a rapid onset of Ni21 influx, which continued steadily for 50 min similarly to that of Na1. The H2O2-induced increase in fluorescence was dependent on extracellular Ni21; when extracellular Ni21 was removed by complexing with EDTA, H2O2 failed to increase the cellular fluorescence.
of bepridil, the basal Ca21- level of around 50 nM immediately increased to 75 nM after the addition of H2O2. The additional increase to 135 nM during the next 35 min of incubation was only small but then the cytosolic Ca21-level dramatically increased reaching a concentration of 452 nM after 88 min. The pretreatment with bepridil caused a significantly smaller increase in the cytosolic Ca21-level, because the maximal Ca21-concentration was only 246 nM (p , .05). A similar effect was observed when Na1 was replaced by choline. Although in this case the Ca21 concentration was enhanced
Effects of bepridil and choline on H2O2-dependent cellular Ca21concentration In view of the H2O2-mediated unspecific cellular Na1-uptake the protective role of bepridil in H2O2 cytotoxicity was unclear. Therefore, we evaluated the effect of bepridil (0.2 mM) on the Ca21 concentration in cells exposed to 0.75 mM H2O2 (Fig. 5). In the absence
Fig. 4. Cellular Ni21-uptake induced by H2O2. The influx of extracellularly added Ni21 (10 mM) was measured by ratio imaging of Newport Green fluorescence, as described under Materials and Methods. 0.75 mM H2O2 was added to the cells at the time indicated by the arrow. Control incubations were performed in Krebs-Henseleit buffer without H2O2 (control) or in the presence of 2 mM EDTA to chelate divalent cations prior to addition of 0.75 mM H2O2.
Role of Na1 in H2O2-cytotoxicity
Fig. 5 Effects of H2O2 on cytosolic Ca21. Ca21 measurements were made in Fura-2 loaded cells treated with 0.75 mM H2O2 at the time indicated. Cells were incubated with Krebs-Henseleit buffer, pH 7.4, (KH) or with the Na1-free variant containing choline instead of Na1 (choline). Bepridil (0.2 mM) was added to the Krebs-Henseleit buffer 8 min before starting the experiment with H2O2.
to 323 nM, this was a significant reduction as compared to H2O2-treated cells in Krebs-Henseleit buffer (p , .01). DISCUSSION
Influx of extracellular Na1 during H2O2 exposure Two lines of evidence indicate that Na1 is involved in H2O2-mediated injury of L929 cells in the absence of glucose. First, in cells incubated with Na1 exogenously added H2O2 caused a significant loss of viability (about 69%) after 2 h, whereas in the presence of choline as substitute for external Na1 we observed only an 11% cell death. Second, using SBFI-fluorescence as a measure for [Na1]i in single cells, H2O2-treatment resulted in an early increase in [Na1]i suggesting that this rise in Na1 rather triggers than mediates cell death directly. Principally, H2O2 may produce cellular Na1 accumulation by either increasing the influx of Na1 into cells or blocking their efflux associated with an impairment of the Na1/ K1-ATPase. The inhibition of Na1/K1-pump by ouabain was not sufficient to achieve a rise in [Na1]i comparable to that obtained after exposure of cells to H2O2. This result was taken as evidence that H2O2induced cellular Na1 accumulation was rather due to an increased Na1-uptake than a decreased Na1-efflux. However, a detrimental action of H2O2 on the Na1/K1pump cannot be excluded. The ability of H2O2 to inhibit
717
Na1/K1-ATPase activity has repeatedly been demonstrated [15,16]. The lack of detection of H2O2-induced impairment of the Na1/K1-pump in our present study may be related to a strong inhibition of the Na1/K1pump caused by our experimental conditions, which included cellular incubation without glucose. This is presumably responsible for a depletion of cellular ATP content associated with an inhibition of the Na1/K1ATPase. Therefore, the predominance of an H2O2 stimulated cellular Na1-uptake over a decreased Na1-efflux refers to L929 cells incubated without glucose only and may not be necessarily applied to experimental conditions with glucose. Experiments using several blockers of Na1-dependent transporters revealed that H2O2 elevated [Na1]i was hardly exerted through alterations in specific protein molecules, because none of the blockers tested except bepridil had any protective effect on cell survival during exposure to H2O2. That bepridil failed to prevent the H2O2 dependent Na1-uptake additionally underlines the notion of a lacking role for Na1-dependent transporters in H2O2-mediated Na1-accumulation. This led us to assume an unspecific Na1-influx into cells possibly due to an interaction of H2O2 with the lipid component of the cell membrane. In contrast, a previous study [2] using menadione as toxic agent has found a reduction of cellular Na1-influx after blockade of the Na1/H1-antiporter and of the Na1/HCO32-cotransporter by, respectively, amiloride and 4,49-di-isothiocyano-2,29-disulfonic acid stilbene; the Na1-influx after activation of these two Na1-transporters during menadione exposure occurred in response to the decrease of intracellular pH. The explanation may lie in the differences between hepatocytes and L929 cells as has been reported previously [17]. Instead of hepatocytes we used in our study tumor cells, which are usually faced with a higher acidic load than cells of normal tissues. We postulate an unspecific Na1-influx in L929 tumor cells exposed to H2O2 leading to an alkaline load. The activity of the Na1/H1-antiporter is not stimulated under these conditions [18]. Therefore, an inhibitory effect of Hoe 694 could not be found. Consistent with this result, SITS tended to be cytotoxic because SITS inhibits, in addition to the Na1dependent, the Na1-independent HCO32/Cl2-antiport acting as a cell acidifying mechanism following an alkaline load [19]. In contrast, DMSO seemed to produce a slight protection against the cytotoxic effects of H2O2, which may result from its hydroxyl radical scavenging activity. The notion that H2O2 mediated cellular Na1-influx proceeds unspecifically across the plasma membrane was substantiated by our finding revealing in response to H2O2-exposure an increase in membrane permeability to extracellular Ni21, which cannot penetrate, per se,
718
A. JUSSOFIE et al.
through the cellular transporters. In particular, the similar time course of cellular ion influx between Ni21 and Na1 supports the view that H2O2 mediates the Na1-influx just as unspecifically across the plasma membrane as the Ni21-influx. We are not aware of other studies suggesting such an unspecific route of Na1-entry into cells as an event relevant for the cytotoxicity of H2O2. However, the ability of H2O2 to increase the Na1-uptake was also reported previously in cultured bovine pulmonary arterial endothelial cells showing an increasing influx of 22 Na in response to a nonlytic concentration of H2O2 [20]. Role of cellular Na1-accumulation in H2O2-mediated cell injury We continued the study analyzing the protective role of bepridil in the cytotoxic action of H2O2 determining the effect of bepridil on the intracellular Ca21 concentration in H2O2-treated cells, because bepridil inhibits the Ca21/Na1-antiporter. From the fluorescence ratio of Fura-2 it was obvious that bepridil was able to reduce H2O2-induced Ca21-influx significantly. Therefore, the Ca21-influx through the Na1/Ca21-antiporter may be a compensatory response to the increased Na1-influx induced by H2O2 in an attempt to raise Na1 back to a normal range. Several lines of evidence support the hypothesis that the Na1/Ca21-antiporter operating in reversed mode is an important mechanism for Na1 extrusion in L929 cells exposed to H2O2: The presence of the Na1/Ca21-exchanger in different tissues and cell types [21–23] such as brain, heart muscle, as well as endocrine, bone, and epithelial cells led us assume this exchanger to occur also in L929 cells. Because the Na1/Ca21-exchanger is electrogenic, the directions in which the ions move depend on the membrane potential. Therefore, this exchanger promotes Ca21-influx in muscle upon depolarization. According to the calculations of Haigney and coworkers [24], the exchanger favors Ca21-entry rather than extrusion, when cytosolic Na1 would exceed 17 mM at a plasma membrane potential of -80 mV. Apparently, H2O2 produced Na1 concentrations of above 70 mM are far higher than the minimum concentration required for the reverse mode of Ca21/Na1-exchange. This conclusion was confirmed by the partial antagonism of the H2O2-dependent Ca21-rises by the replacement of external Na1 by choline. Our data are in line with the previous demonstration that endogenous Na1/Ca21-exchanger in the plasma membrane of Xenopus oocytes run in reverse mode in the absence of high external Na1 and the presence of external Ca21 [25]. We did not obtain complete abolition of H2O2-dependent cellular Ca21accumulation by either bepridil or choline indicating the presence of other [Ca21]i increasing mechanisms in re-
sponse to H2O2-exposure [26]. The reverse mode of Ca21/Na1-exchange following exposure of cells to H2O2 agrees with the results of Lidofsky and coworkers [27], showing that this exchanger is not involved in the physiological regulation of cytosolic Ca21-levels in hepatocytes. In conclusion, the H2O2-dependent Ca21-entry is augmented by the initial Na1-influx that helps the intracellular Ca21 to reach cytotoxic levels. Accordingly, suppression of the reverse mode of Ca21/Na1-exchange by bepridil or omission of Na1 was effective against this additional source of Ca21-elevation and, consequently, ameliorated the cytotoxic action of H2O2. Therfore, the effect of Na1 on cytotoxicity is mediated via excessive Ca21, known as an ultimate forerunner of cell death [28,29]. Taken together, the results of the present investigation confirmed and extended the view that Na1 play a critical role in the cytotoxic action of H2O2. The exposure of L929 cells to H2O2 in the absence of glucose resulted in an Na1-influx, which was rather due to an increased membrane permeablitiy for Na1, per se, than through specific Na1-dependent transporters. From our finding that the rise in [Na]i led in turn to the activation of the Na1/Ca21-antiporter in the reversed mode in an attempt to counteract this Na1-load it was concluded that Na1accumulation serves to amplify cytosolic Ca21 rises thereby triggering H2O2-mediated cell death. REFERENCES [1] Lambotte, L. Effect of anoxia and ATP depletion on the membrane potential and permeability of dog liver. J. Physiol. 269:53– 76; 1977. [2] Carini, R.; Bellomo, G.; Benedetti, A.; Fulceri, R.; Gamerucci, A.; Rarola M.; Dianzani M. U.; Albano, E. Alteration of Na1 homeostasis as a critical step in the development of irreversible hepatocyte injury after adenosine triphosphate depletion. Hepatology 21:1089 –1098; 1995. [3] Nath, K. A.; Ngo, E. O.; Hebbel, R.P.; Croatt, A. J.; Zhou B.; Nutter, L. M. a-Ketoacids scavenge H2O2 in vitro and in vivo and reduce menadione-induced DNA injury and cytotoxicity. Am. J. Physiol. 268:C227–C236; 1995. [4] Lomonosowa, E.; Kirsch M.; de Groot, H. Calcium versus ironmediated processes in hydrogen peroxide toxicity to L929 cells. Effects of glucose. Free Radic. Biol. Med. (in press). [5] Farber, J. L.; Kyle, M. E.; Coleman, J. B. Biology of diseasemechanisms of cell injury by activated oxygen species. Laboratory investigation 62:670 – 679; 1990. [6] Coyle, J. T.; Puttfarcken, P. Oxidative stress, glutamate, and neurodegenerative disorders. Science 262:689 – 695; 1993. [7] Orrenius, S; McConkey, D. J.; Bellomo, G; Nicotera, P. Role of Ca21 in toxic cell killing. Trends Pharmacol. Sci. 10:281–285; 1989. [8] Murphy, M. E.; Piper, H. M.; Watanabe, H.; Sies, H. Nitric oxide production by cultured aortic endothelial cells in response to thiol depletion and replenishment. J. Biol. Chem. 266:19378 –19383; 1991. [9] Hugo-Wissemann, D.; Anundi, I.; Lauchart, W.; Viebahn, R.; de Groot, H. Differences in glycolytic capacity and hypoxia tolerance between hepatoma cells and hepatocytes. Hepatology 13: 297–303; 1991. [10] Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. A new generation of
Role of Na1 in H2O2-cytotoxicity
[11]
[12] [13]
[14] [15] [16] [17]
[18] [19] [20]
[21] [22]
Ca21 indicators with greatly improved fluorescence properties. J. Biol. Chem. 260:3440 –3450; 1985. Donoso, P.; Mill, J. G.; O’Neill, S. C.; Eisner, D. A. Fluorescence measurements of cytoplasmic and mitochondrial sodium concentration in rat ventricular myocytes. J. Physiol. 448:493–509; 1992. Grewal, K. K.; Rafeiro, E.; Racz, W. J. Bromobenzene and furosemide hepatotoxicity: alterations in glutathione, protein thiols, and calcium. Can. J. Physiol. Pharmacol. 74:257–264; 1996. Boyer, J. L.; Ananthanarayanan, M.; Hofmann, A. F.; Scheingart, C. D.; Hagenbuch, B; Stieger, B.; Meier, P. J. Expression and characterization of a functional rat liver Na1 bile acid cotransport system in COS-7 cells. Am. J. Physiol. 266:G382–G387; 1994. Sakai, H.; Kakinoki, B.; Diener, M.; Takeguchi, N. Endogenous arachidonic acid inhibits hypotonically-activated Cl2 channels in isolated rat hepatocytes. Jpn. J. Physiol. 46:311–318; 1996. Garner, M. H.; Garner, W. H.; Spector, A. Kinetic cooperativity change after H2O2 modification of (Na,K)-ATPase. J. Biol. Chem. 259:7712–7718; 1984. Matsuoka, T.; Kato, M.; Kako, K. J. Effect of oxidants on Na, K,ATPase and its reversal. Basic Res. Cardiol. 85:330 –341; 1990. Yamamoto, K.; Tsukidate, K.; Farber, J. Differing effects of the inhibition of poly(ADP-ribose)polymerase on the course of oxidative cell injury in hepatocytes and fibroblasts. Biochem. Pharmacol. 46:483– 491; 1993. Horvat, B.; Shahram, T.; Salihagic, A. Tumour cell proliferation is abolished by inhibitors of Na1/H1 and HCO32 exchange. Eur. J. Cancer 29A:132–137; 1993. Frelin, C.; Vigne, P.; Ladoux, A.; Lazdunski, M. The regulation of the intracellular pH in cells from vertebrates. Eur. J. Biochem. 174:3–14; 1988. Meharg, J. V; McGowan, J. J.; Charles, A.; Parmelee, J. T; Cutaia, M. V.; Rounds, S. Hydrogen peroxide stimulates sodiumpotassium pump activity in cultured pulmonary arterial endothelial cells. Am. J. Physiol. 265:L613–L621; 1993. Carafoli, E. Intracellular calcium homeostasis. Annu. Rev. Biochem. 56:395– 433; 1987. Blaustein, M. P. Goldman, W. F., Fontana, G., Kreuger, B. K., Santiago, E. M., Steele, T. D., Weiss, D. N., Yarowsky, P. J. Physiological roles of sodium-calcium exchanger in nerve and muscle. Ann. N. Y. Acad. Sci. 639:253–273; 1991.
719
[23] Herman, B.; Gores, G. J.; Nieminen, A.-L; Kawanishi, T.; Harman, A:; Lemasters, J. J. Calcium and pH in anoxic and toxic injury. Crit. Rev. Toxicol. 21:127–148; 1990. [24] Haigney, M.C.; Miyata, H.; Lakatta, E.G.; Stern, M.D.; Silverman, H. S. Dependence of hypoxic cellular calcium loading on Na1-Ca21 exchange. Circ. Res. 71:547–557; 1992. [25] Schlief, T; Heinemann, S. H. H2O2-induced chloride currents are indicative of an endogenous Na1-Ca21 exchange mechanism in Xenopus oocytes. J. Physiol. 486:123–130; 1995. [26] Richter C. Pro-oxidants and mitochondrial Ca21: their relationship to apoptosis and oncogenesis. FEBS-Lett. 325:104 –107; 1993. [27] Lidofsky, S. D.; Xie, M. H.; Scharmschmidt, B. F. Na1-Ca21 exchange in cultured rat hepatocytes: evidence against a role in cytosolic Ca21 regulation or signaling. Am. J. Physiol. 259:G56 – G61; 1990. [28] Nicotera, P.; Bellomo, G.; Orrenius, G. The role of Ca21 in cell killing. Chem. Res. Toxicol. 3:484 – 494; 1990. [29] Beal, M. F. Mechanisms of excitotoxicity in neurologic diseases. FASEB J. 6:3338 –3344; 1992.
ABBREVIATIONS
H2O2— hydrogen peroxide SBFI—sodium-binding benzofuran isophthalate EDTA— ethylenediaminetetraacetic acid SITS— 4-acetamido-49-isothio-cyanatostilbene-2,29disulfonic acid Hoe 694 —Hoechst 694 MEM—minimal essential medium Hepes— 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid DMSO— dimethylsulphoxide LDH—lactate dehydrogenase EGTA—(2-aminoethoxyethane)-N,N,N9,N9-tetraacetic acid
PII S0891-5849(98)00159-2
Original Contribution Ca21-DEPENDENT CYTOTOXICITY OF H2O2 IN L929 CELLS: THE ROLE OF H2O2-INDUCED Na1-INFLUX ASTRID JUSSOFIE, MICHAEL KIRSCH,
and
HERBERT
DE
GROOT
Institut fu¨r Physiologische Chemie, Universita¨tsklinikum, Essen, Germany (Received 20 January 1998; Revised 5 June 1998; Accepted 8 June 1998)
Abstract—We investigated the mechanism by which H2O2 mediates an increase in [Na1]i in L929 cells and the relevance of this Na1 load for H2O2-induced cell injury. [Na1]i increased early after exposure to H2O2 as monitored by fluorescence spectrophotometry of cells loaded with SBFI. The omission of Na1 from the incubation buffer significantly reduced H2O2-cytotoxicity. This protection could not be mimicked by inhibition of either the Na1/H1antiporter, the Na1/HCO32-cotransporter, or the Na1/K1/2Cl2-cotransporter by using Hoechst 694 (0.02 mM) or 4-acetamido-49-isothio-cyanatostilbene-2,29-disulfonic acid (SITS) (0.02 mM) or furosemide (1 mM) and bumetanide (0.5 mM). Only the blocker of the Na1/Ca21-exchanger bepridil (0.2 mM) significantly reduced H2O2-cytotoxicity but without interfering with the increase in [Na1]i. H2O2 caused a rapid and sustained increase in [Ca21]i, which was significantly reduced in bepridil pretreated cells and after replacing extracellular Na1 by choline. H2O2 was found to initiate a cellular uptake of unphysiological Ni21 by using Newport Green diacetate as fluorescent dye. Our data suggest that H2O2 mediates Na1-influx across the plasma membrane rather unspecifically than through specific transporters. The protective effect of bepridil against H2O2-cytotoxicity occurs as a consequence of a reduced cellular Ca21-uptake. We conclude that H2O2-mediated unspecific accumulation of Na1 seems to favor a Ca21-influx into the cells, which takes place on the Na1/Ca21-exchanger operating in reverse mode in exchange for Na1-efflux. Therefore, H2O2-induced cellular Na1 accumulation appears to play a permissive rather than a triggering role in H2O2-mediated cell injury. © 1998 Elsevier Science Inc. Keywords—Hydrogen peroxide, Cytotoxicity, Intracellular Na1, Ca21
be partly due to the activation of the Na1/H1-antiporter and the Na1/HCO32-cotransporter associated with an intracellular Na1 load [2]. Because menadione generates H2O2 in cells [3], we focus on the question whether H2O2 may also induce disturbances in cellular Na1 homeostasis necessary for H2O2-mediated cell injury. Results obtained in our lab [4], indicate that the mechanism of H2O2 injury in L929 cells is highly dependent on the metabolic state of these cells given by the presence or absence of glucose. In the presence of glucose, cell death was mainly mediated by iron consistent with the ironcatalyzed production of hydroxyl radicals (Fenton reaction) as evidenced by Farber and coworkers [5]. These extremely reactive species are assumed to produce most of the covalent modification and damage to macromolecules, including DNA, proteins, and lipid membranes [5,6]. However, in the absence of glucose, disruption of ion homeostasis could be demonstrated as the decisive event responsible for H2O2 cytotoxicity as proposed by
INTRODUCTION
The importance of Na1 in the control of cytosolic pH and1 cell volume is reflected in the dynamics of cellular Na regulation. Extracellular Na1 can enter cells through different transporters such as the Na1/H1-antiporter, the Na1/HCO32-cotransporter or the Na1/K1/2Cl2-cotransporter. The activity of these transporters depends on the maintenance of the sodium electrochemical gradient across the membrane. Under physiological conditions, the excess Na1 will be pumped out by the Na1/K1ATPase, therefore, maintaining Na1 homeostasis (Fig. 1) [1]. However, an uncontrolled and persistent elevation of cellular Na1-levels promotes cell injury. The loss of viability with menadione in hepatocytes is proposed to Address correspondence to: PD Dr. A. Jussofie, Institut fu¨r Physiologische Chemie, Universita¨tsklinikum Essen, Hufelandstr. 55, D-45122 Essen, Germany; Tel: 149-201-723-4105; Fax: 149-201723-5943. 712
Role of Na1 in H2O2-cytotoxicity
713
nel blockers used as furosemide, SITS, bumetanide, bepridil were from Sigma (Deisenhofen, Germany) except Hoe 694 which was a gift from Hoechst (Frankfurt/Main, Germany). Cell culture The murine fibroblast-derived cells L929 (American Type Culture collection, NCTC clone 929 of strain L) were grown in minimum essential medium (MEM) Eagle (Sigma) supplemented with 25 mM bicarbonate, 10% fetal bovine serum (Sigma), 2 mM L-glutamine (Sigma), and 50 mg/ml gentamycin (Gibco BRL) in plastic flasks (Falcon, Heidelberg, Germany) at 37°C in a humidified atmosphere of 5% CO2 in air. Cells were trypsinized (0.25% v/v) twice weekly. Those from passage 5 to 18 were used for experiments. Fig.1. Cellular sodium regulatory mechanisms. Na1-influx through the plasma membrane can be attained by the Na1/H1-antiporter, the Na1/ Ca21-antiporter, the Na1/K1/2Cl2-cotransporter and the Na1/HCO32cotransporter. Na1 efflux involves the Na1/K1-ATPase.
Orrenius and coworkers [7]. Although free Ca21 seems to be required for H2O2-mediated cell injury, these data do not rule out the possibility that cellular accumulation of Ca21 after H2O2 exposure is not the initiating event in the development of membrane damage, but rather a consequence of a preceding event such as a Na1 challenge outside the physiological range. In the present study, we have examined whether and to what an extent loss of Na1 homeostasis is a mechanism responsible for H2O2-mediated cell injury by using a fluorescence spectrophotometric assay with the fluorescence indicators SBFI/AM and Fura-2/AM to monitor intracellular levels of Na1 and Ca21 time dependently. These experiments were performed with L929 cells incubated without glucose to assure a maximal disturbance of cellular ion homeostasis under the experimental conditions used sufficient to underlie H2O2-mediated cell injury. MATERIALS AND METHODS
Materials The acetoxymethyl esters of SBFI, Fura-2 and Newport green were obtained from Molecular Probes (Leiden, The Netherlands). Hydrogen peroxide was purchased from Aldrich (Steinheim, Germany). Pyruvate, NADH, and ethylenediaminetetraacetic acid, disodium salt (EDTA) were obtained from Boehringer Mannheim (Mannheim, Germany). All other chemicals were obtained from Merck (Darmstadt, Germany). All the chan-
Procedures L929 cells were seeded at a density of 9 3 104 cells/0.9 ml of MEM in 12-well tissue culture plates (Falcon) or, for fluorescent measurements, on glass coverslips (28 mm diameter, Fa. Assistent, Germany) in petri dishes and allowed to grow for 2 days. For each experiment the cells were washed once either with Krebs-Henseleit buffer, pH 7.4, supplemented with 20 mM Hepes or with the sodium-free variant containing choline in exchange for Na1. The Krebs-Henseleit buffer consisted of (in mM): NaCl 115.0, NaHCO3 25.0, KCl 5.9, MgCl2 1.2, NaH2PO4 1.2, Na2SO4 1.2, CaCl2 2.5. For the experiments performed in the absence of Na1 the sodium-free variant had been used containing (in mM): KCl 5.9, MgSO4 1.2, H3PO4 1.2, choline-HCO3 25, CaCl 2.5, choline-Cl 117.0. The two buffers were saturated with 5% CO2 and 95% air at 37°C. The experiment was started by the addition of H2O2 to a final concentration of 0.75 mM. The Na1-channel-blockers furosemide (1 mM), bumetanide (0.5 mM), SITS (0.2 mM) , Hoe 694 (0.02 mM), and bepridil (0.2 mM) were added directly prior to H2O2-exposure. Bepridil and SITS were dissolved in Krebs-Henseleit buffer, pH 7.4; furosemide was dissolved in the same buffer but supplemented with 1 mM NaOH. Bumetanide was added as an ethanolic solution (final concentration 1%), Hoe 694 as a DMSOsolution (final concentration 1%). Cytotoxicity was determined by measuring intra- and extracellular lactate dehydrogenase (LDH) activity as described previously [8]. Results obtained using the LDH-leakage viability test were compatible with our data using viability staining such as trypan blue exclusion [9]; therefore, cell damage induced by H2O2 will be expressed in the Results section as loss of viability.
714
A. JUSSOFIE et al.
Fluorescence measurements The cover slips with the cells attached were washed twice with Krebs-Henseleit buffer or the sodium-free variant and transferred into a chamber (Pentz, chamber system, Heraeus, Hanau, Germany), which was mounted on the thermostatic table of an inverted microscope (Zeiss, Oberkochen, Germany). The cells were flushed with 5% CO2 and 95% air during loading with the fluorescent dye and the whole incubation period. To measure changes in the intracellular Ca21-concentration the Ca21 fluorescence indicator Fura-2/AM was added to the coverslips at a final concentration of 6 mM in the respective buffer and cells were returned to the 37°C/5% CO2 incubator for 60 min. Cells were then washed twice with the respective buffer. Fura-2 fluorescence was measured by using a videomicroscope system (Attofluor TM fluorescence analyser, Zeiss, Oberkochen). Cells were detected using 334 nm and 380 nm excitation and monitoring fluorescence by means of an emission barrier filter at 500 –530 nm. Calibration and calculation of free cytosolic Ca21 were performed according to the procedure described by Grynkiewicz and coworkers [10]. Rmax was obtained by adding 50 mM ionomycin, Rmin was evaluated by adding 10 mM EGTA. For [Na1]i-measurement cells were loaded with SBFI-AM for 75 min at 37°C in Krebs-Henseleit or sodium-free buffer both containing 20 mM SBFI and the nonionic detergent Pluoronic at a concentration of 0.06%. The excitation monochromators were set at 334 and 380 nm; emitted fluorescence was measured as described previously. The intracellular Na1 concentration was calibrated in situ by the method of Donoso and coworkers [11]. Two calibration solutions were used to contain: NaCl or KCl 140 mM, Hepes 10 mM, EGTA 1 mM, glucose 10 mM, gramicidin D 4 mM, monensin 10 mM, nigericin 10 mM; adjusted to pH 7.4 with either NaOH or KOH. The mixture of these two stocks solutions yielded calibration solutions with the following sodium concentrations: 17.5, 35, 52.5, 70, and 105 mM. To test the hypothesis concerning an H2O2 induced unspecific cation influx into cells Ni21 was added extracellularly at a final concentration of 10 mM. We used the cell permanent Newport Green diacetate as fluorescent indicator for nickel which exhibits an increase in fluorescence emission upon binding Ni21. Cells were loaded with Newport Green (final concentration 5 mM) for 15 min at 37°C in 5% CO2 and 95% air. The Ni21-influx was measured by ratio imaging of Newport Green fluorescence excited at 480 and 340 nm. Fluorescence was monitored as described.
Fig. 2. Toxicity of H2O2 to L929 cells incubated in the presence of Na1. Cells were exposed for 5 h to 0.75 mM H2O2 in Krebs-Henseleit buffer, pH 7.4, at 37°C in 5% CO2 and 95% air. For the experiments conducted without Na1, choline was used in exchange for Na1. At the times indicated cell injury was estimated by LDH leakage and trypan blue exclusion, as described in Materials and Methods. Data expressed as loss of viability are means 6 SEM of 6 independent experiments, each performed in triplicate.
RESULTS
Cytotoxicity of H2O2 L929 cells cultured for 2 days were exposed to 0.75 mM H2O2. Fig. 2 shows that this H2O2 treatment led to a large loss of viability when it was performed in KrebsHenseleit buffer, i.e., in the presence of Na1. The most pronounced reduction of surviving cells was observed during the second hour of exposure to H2O2 where cell death increased from 3 to 69%. After 3 h, almost all of the cells had already lost their viability. When the cells were exposed to H2O2 in Krebs-Henseleit buffer where Na1 was exchanged for choline the cytotoxicity of H2O2 was significantly reduced although cell death was still much higher than that observed in controls. Cell death remained low after 2 h, reaching only 11% but then increased steadily to 85% after 5 h.
Role of Na1 in H2O2-cytotoxicity
715
Fig. 3. Effect of H2O2 on the cytosolic Na1-concentration. Cytosolic Na1-levels were monitored in L929 cells loaded with the flurorescent dye SBFI/AM (20 mM for 75 min). After baseline fluorescence levels were monitored for 8 min 0.75 mM H2O2 and 1 mM ouabain were added to cells incubated at 37°C in 5% CO2 and 95% air with Krebs-Henseleit buffer, pH 7.4 at the time indicated by the arrow (A). The rise of [Na]i could not be blocked if the cells were preincubated with 0.2 mM bepridil for 8 min (B). Control cells were incubated either in Krebs-Henseleit buffer without H2O2 or with the Na1-free variant containing choline instead of Na1 in the presence of 0.75 mM H2O2. Data are means of 5 experiments 6 SEM each registered as the average fluorescence of 30 –50 cells.
Effect of H2O2 on the intracellular Na1 concentration
Effect of sodium channel blockers on H2O2-induced cell injury
Choline is a substitute for Na1. Therefore, we evaluated the effect of H2O2 on intracellular sodium concentration using the fluorescence ratio of SBFI as a measure of sodium levels. The addition of 0.75 mM H2O2 increased the intracellular Na1-concentration in a timedependent manner compared to both the untreated control in KHR-buffer or the H2O2-treated cells in choline buffer (Fig. 3). This rise of [Na1]i already began 4 min following the addition of H2O2 and continued for 60 min in a linear fashion reaching concentrations above 70 mM. This intracellular Na1-accumulation could not be mimicked when the Na1/K1-ATPase was inhibited by 1 mM ouabain, indicative of rather an increased Na1-influx than a decreased Na1-efflux elicited by H2O2. In the absence of H2O2 only a slight increase up to 32 mM was detectable. The H2O2-induced increase in Na1 concentration was dependent on extracellular Na1; when Na1 was replaced by choline the H2O2-treated cells maintained their basal Na1 level of 18 mM over the whole course of incubation.
Next we examined the effects of different sodium channel blockers such as furosemide [12] (1 mM), bumetanide [13] (0.5 mM), SITS [14] (0.2 mM), Hoe 694 (0.02 mM), and bepridil (0.2 mM) on H2O2-induced cell injury. None of the blockers tested was effective in blocking H2O2-induced cell damage except bepridil (Table 1). The co-administration of 200 mM bepridil produced only a 18% loss of viability after 2 h, compared to 61% obtained with cells solely treated with H2O2. Bepridil levels lower than 200 mM were effective in a concentration-dependent manner, whereas higher concentrations of the channel blocker did not additionally improve protection (data not shown). Bepridil is well known to block the Na1/Ca21-exchanger. In view of its protective effect on cells exposed to H2O2 we tested whether bepridil might reduce or delay the H2O2-induced increase in [Na1]i. However, bepridil failed to affect the H2O2-dependent Na1 uptake at all; the intracellular Na1 concentration determined in the presence of bepridil did not differ significantly from that
716
A. JUSSOFIE et al. Table 1. Effects of the Sodium Channel Blockers on the Cytotoxic Action of H2O2 in L929 Cells Loss of viability (% of total)
Conditions A. B. C. D. E. F. G. H. I. J.
Control H2O2 H2O2 1 1% DMSO H2O2 1 1% ethanol H2O2 1 1 mM NaOH H2O2 1 1 mM furosemide H2O2 1 0.5 mM bumetanide H2O2 1 0.2 mM sits H2O2 1 0.02 mM Hoe H2O2 1 0.2 mM bepridil
1h
2h
3h
4h
5h
2.4 6 0.26 4.4 6 0.9 4.0 6 1.1 3.7 6 0.9 3.9 6 0.8 5.0 6 1.2 4.9 6 1.1 6.0 6 0.9 4.9 6 0.8 5.3 6 1.3
3.2 6 0.5 61 6 3.5 55 6 2.9 59 6 3.2 60 6 3.6 62 6 2.8 56 6 3.1 73 6 2.9 64 6 3.0 18 6 3.0
3.6 6 0.5 89 6 2.5 75 6 3.2 88 6 3.5 87 6 3.8 83 6 3.7 88 6 3.4 93 6 1.9 91 6 1.8 67 6 2.1
3.6 6 0.8 93 6 3.0 88 6 1.9 93 6 2.1 92 6 2.5 90 6 2.3 92 6 2.5 93 6 2.1 94 6 1.7 89 6 1.6
3.4 6 0.6 94 6 2.1 91 6 2.2 95 6 1.9 95 6 2.0 92 6 1.8 95 6 1.8 94 6 1.5 95 6 1.6 92 6 0.8
Cells were grown for 2 days and exposed to 0.75 mM H2O2 in Krebs-Henseleit buffer, pH 7.4, at 37°C. After the exposure period indicated the loss of viability was determined by estimating LDH leakage and trypan blue exclusion. The results are means 6 SEM of five experiments. Student’s t-test for all time periods after 1 h of exposure to H2O2 demonstrated a significantly larger cell injury than control cells incubated without H2O2. The loss of viability in J was significantly ( p , .001, two-way ANOVA) lower than that observed in B.
in cells incubated without bepridil over the whole course of 90 min (Fig. 3). Effect of H2O2 on membrane permeability for Ni21 Because neither the H2O2-mediated cellular Na1-uptake could be blocked by bepridil nor a reduction of the cytotoxic action of H2O2 could be achieved with the other commonly used blockers of Na1-dependent transporters, we determined whether H2O2 produces an unspecific influx of Na1 across the plasma membrane possibly due to an increased permeability of cations. Therefore, we used unphysiological Ni21, which cannot move, per se, through the cellular channels when added extracellularly. To monitor the effect of H2O2 on Ni21 uptake we used the fluorescent dye Newport Green, which increases its fluorescence upon binding Ni21. As shown in Fig. 4, by the increasing ratio of 480/340 nm values of Newport Green fluorescence Ni21 comes into the cells that were exposed to H2O2, whereas Ni21 permeability of the untreated control cells was negligable. After the addition of H2O2, cells displayed a rapid onset of Ni21 influx, which continued steadily for 50 min similarly to that of Na1. The H2O2-induced increase in fluorescence was dependent on extracellular Ni21; when extracellular Ni21 was removed by complexing with EDTA, H2O2 failed to increase the cellular fluorescence.
of bepridil, the basal Ca21- level of around 50 nM immediately increased to 75 nM after the addition of H2O2. The additional increase to 135 nM during the next 35 min of incubation was only small but then the cytosolic Ca21-level dramatically increased reaching a concentration of 452 nM after 88 min. The pretreatment with bepridil caused a significantly smaller increase in the cytosolic Ca21-level, because the maximal Ca21-concentration was only 246 nM (p , .05). A similar effect was observed when Na1 was replaced by choline. Although in this case the Ca21 concentration was enhanced
Effects of bepridil and choline on H2O2-dependent cellular Ca21concentration In view of the H2O2-mediated unspecific cellular Na1-uptake the protective role of bepridil in H2O2 cytotoxicity was unclear. Therefore, we evaluated the effect of bepridil (0.2 mM) on the Ca21 concentration in cells exposed to 0.75 mM H2O2 (Fig. 5). In the absence
Fig. 4. Cellular Ni21-uptake induced by H2O2. The influx of extracellularly added Ni21 (10 mM) was measured by ratio imaging of Newport Green fluorescence, as described under Materials and Methods. 0.75 mM H2O2 was added to the cells at the time indicated by the arrow. Control incubations were performed in Krebs-Henseleit buffer without H2O2 (control) or in the presence of 2 mM EDTA to chelate divalent cations prior to addition of 0.75 mM H2O2.
Role of Na1 in H2O2-cytotoxicity
Fig. 5 Effects of H2O2 on cytosolic Ca21. Ca21 measurements were made in Fura-2 loaded cells treated with 0.75 mM H2O2 at the time indicated. Cells were incubated with Krebs-Henseleit buffer, pH 7.4, (KH) or with the Na1-free variant containing choline instead of Na1 (choline). Bepridil (0.2 mM) was added to the Krebs-Henseleit buffer 8 min before starting the experiment with H2O2.
to 323 nM, this was a significant reduction as compared to H2O2-treated cells in Krebs-Henseleit buffer (p , .01). DISCUSSION
Influx of extracellular Na1 during H2O2 exposure Two lines of evidence indicate that Na1 is involved in H2O2-mediated injury of L929 cells in the absence of glucose. First, in cells incubated with Na1 exogenously added H2O2 caused a significant loss of viability (about 69%) after 2 h, whereas in the presence of choline as substitute for external Na1 we observed only an 11% cell death. Second, using SBFI-fluorescence as a measure for [Na1]i in single cells, H2O2-treatment resulted in an early increase in [Na1]i suggesting that this rise in Na1 rather triggers than mediates cell death directly. Principally, H2O2 may produce cellular Na1 accumulation by either increasing the influx of Na1 into cells or blocking their efflux associated with an impairment of the Na1/ K1-ATPase. The inhibition of Na1/K1-pump by ouabain was not sufficient to achieve a rise in [Na1]i comparable to that obtained after exposure of cells to H2O2. This result was taken as evidence that H2O2induced cellular Na1 accumulation was rather due to an increased Na1-uptake than a decreased Na1-efflux. However, a detrimental action of H2O2 on the Na1/K1pump cannot be excluded. The ability of H2O2 to inhibit
717
Na1/K1-ATPase activity has repeatedly been demonstrated [15,16]. The lack of detection of H2O2-induced impairment of the Na1/K1-pump in our present study may be related to a strong inhibition of the Na1/K1pump caused by our experimental conditions, which included cellular incubation without glucose. This is presumably responsible for a depletion of cellular ATP content associated with an inhibition of the Na1/K1ATPase. Therefore, the predominance of an H2O2 stimulated cellular Na1-uptake over a decreased Na1-efflux refers to L929 cells incubated without glucose only and may not be necessarily applied to experimental conditions with glucose. Experiments using several blockers of Na1-dependent transporters revealed that H2O2 elevated [Na1]i was hardly exerted through alterations in specific protein molecules, because none of the blockers tested except bepridil had any protective effect on cell survival during exposure to H2O2. That bepridil failed to prevent the H2O2 dependent Na1-uptake additionally underlines the notion of a lacking role for Na1-dependent transporters in H2O2-mediated Na1-accumulation. This led us to assume an unspecific Na1-influx into cells possibly due to an interaction of H2O2 with the lipid component of the cell membrane. In contrast, a previous study [2] using menadione as toxic agent has found a reduction of cellular Na1-influx after blockade of the Na1/H1-antiporter and of the Na1/HCO32-cotransporter by, respectively, amiloride and 4,49-di-isothiocyano-2,29-disulfonic acid stilbene; the Na1-influx after activation of these two Na1-transporters during menadione exposure occurred in response to the decrease of intracellular pH. The explanation may lie in the differences between hepatocytes and L929 cells as has been reported previously [17]. Instead of hepatocytes we used in our study tumor cells, which are usually faced with a higher acidic load than cells of normal tissues. We postulate an unspecific Na1-influx in L929 tumor cells exposed to H2O2 leading to an alkaline load. The activity of the Na1/H1-antiporter is not stimulated under these conditions [18]. Therefore, an inhibitory effect of Hoe 694 could not be found. Consistent with this result, SITS tended to be cytotoxic because SITS inhibits, in addition to the Na1dependent, the Na1-independent HCO32/Cl2-antiport acting as a cell acidifying mechanism following an alkaline load [19]. In contrast, DMSO seemed to produce a slight protection against the cytotoxic effects of H2O2, which may result from its hydroxyl radical scavenging activity. The notion that H2O2 mediated cellular Na1-influx proceeds unspecifically across the plasma membrane was substantiated by our finding revealing in response to H2O2-exposure an increase in membrane permeability to extracellular Ni21, which cannot penetrate, per se,
718
A. JUSSOFIE et al.
through the cellular transporters. In particular, the similar time course of cellular ion influx between Ni21 and Na1 supports the view that H2O2 mediates the Na1-influx just as unspecifically across the plasma membrane as the Ni21-influx. We are not aware of other studies suggesting such an unspecific route of Na1-entry into cells as an event relevant for the cytotoxicity of H2O2. However, the ability of H2O2 to increase the Na1-uptake was also reported previously in cultured bovine pulmonary arterial endothelial cells showing an increasing influx of 22 Na in response to a nonlytic concentration of H2O2 [20]. Role of cellular Na1-accumulation in H2O2-mediated cell injury We continued the study analyzing the protective role of bepridil in the cytotoxic action of H2O2 determining the effect of bepridil on the intracellular Ca21 concentration in H2O2-treated cells, because bepridil inhibits the Ca21/Na1-antiporter. From the fluorescence ratio of Fura-2 it was obvious that bepridil was able to reduce H2O2-induced Ca21-influx significantly. Therefore, the Ca21-influx through the Na1/Ca21-antiporter may be a compensatory response to the increased Na1-influx induced by H2O2 in an attempt to raise Na1 back to a normal range. Several lines of evidence support the hypothesis that the Na1/Ca21-antiporter operating in reversed mode is an important mechanism for Na1 extrusion in L929 cells exposed to H2O2: The presence of the Na1/Ca21-exchanger in different tissues and cell types [21–23] such as brain, heart muscle, as well as endocrine, bone, and epithelial cells led us assume this exchanger to occur also in L929 cells. Because the Na1/Ca21-exchanger is electrogenic, the directions in which the ions move depend on the membrane potential. Therefore, this exchanger promotes Ca21-influx in muscle upon depolarization. According to the calculations of Haigney and coworkers [24], the exchanger favors Ca21-entry rather than extrusion, when cytosolic Na1 would exceed 17 mM at a plasma membrane potential of -80 mV. Apparently, H2O2 produced Na1 concentrations of above 70 mM are far higher than the minimum concentration required for the reverse mode of Ca21/Na1-exchange. This conclusion was confirmed by the partial antagonism of the H2O2-dependent Ca21-rises by the replacement of external Na1 by choline. Our data are in line with the previous demonstration that endogenous Na1/Ca21-exchanger in the plasma membrane of Xenopus oocytes run in reverse mode in the absence of high external Na1 and the presence of external Ca21 [25]. We did not obtain complete abolition of H2O2-dependent cellular Ca21accumulation by either bepridil or choline indicating the presence of other [Ca21]i increasing mechanisms in re-
sponse to H2O2-exposure [26]. The reverse mode of Ca21/Na1-exchange following exposure of cells to H2O2 agrees with the results of Lidofsky and coworkers [27], showing that this exchanger is not involved in the physiological regulation of cytosolic Ca21-levels in hepatocytes. In conclusion, the H2O2-dependent Ca21-entry is augmented by the initial Na1-influx that helps the intracellular Ca21 to reach cytotoxic levels. Accordingly, suppression of the reverse mode of Ca21/Na1-exchange by bepridil or omission of Na1 was effective against this additional source of Ca21-elevation and, consequently, ameliorated the cytotoxic action of H2O2. Therfore, the effect of Na1 on cytotoxicity is mediated via excessive Ca21, known as an ultimate forerunner of cell death [28,29]. Taken together, the results of the present investigation confirmed and extended the view that Na1 play a critical role in the cytotoxic action of H2O2. The exposure of L929 cells to H2O2 in the absence of glucose resulted in an Na1-influx, which was rather due to an increased membrane permeablitiy for Na1, per se, than through specific Na1-dependent transporters. From our finding that the rise in [Na]i led in turn to the activation of the Na1/Ca21-antiporter in the reversed mode in an attempt to counteract this Na1-load it was concluded that Na1accumulation serves to amplify cytosolic Ca21 rises thereby triggering H2O2-mediated cell death. REFERENCES [1] Lambotte, L. Effect of anoxia and ATP depletion on the membrane potential and permeability of dog liver. J. Physiol. 269:53– 76; 1977. [2] Carini, R.; Bellomo, G.; Benedetti, A.; Fulceri, R.; Gamerucci, A.; Rarola M.; Dianzani M. U.; Albano, E. Alteration of Na1 homeostasis as a critical step in the development of irreversible hepatocyte injury after adenosine triphosphate depletion. Hepatology 21:1089 –1098; 1995. [3] Nath, K. A.; Ngo, E. O.; Hebbel, R.P.; Croatt, A. J.; Zhou B.; Nutter, L. M. a-Ketoacids scavenge H2O2 in vitro and in vivo and reduce menadione-induced DNA injury and cytotoxicity. Am. J. Physiol. 268:C227–C236; 1995. [4] Lomonosowa, E.; Kirsch M.; de Groot, H. Calcium versus ironmediated processes in hydrogen peroxide toxicity to L929 cells. Effects of glucose. Free Radic. Biol. Med. (in press). [5] Farber, J. L.; Kyle, M. E.; Coleman, J. B. Biology of diseasemechanisms of cell injury by activated oxygen species. Laboratory investigation 62:670 – 679; 1990. [6] Coyle, J. T.; Puttfarcken, P. Oxidative stress, glutamate, and neurodegenerative disorders. Science 262:689 – 695; 1993. [7] Orrenius, S; McConkey, D. J.; Bellomo, G; Nicotera, P. Role of Ca21 in toxic cell killing. Trends Pharmacol. Sci. 10:281–285; 1989. [8] Murphy, M. E.; Piper, H. M.; Watanabe, H.; Sies, H. Nitric oxide production by cultured aortic endothelial cells in response to thiol depletion and replenishment. J. Biol. Chem. 266:19378 –19383; 1991. [9] Hugo-Wissemann, D.; Anundi, I.; Lauchart, W.; Viebahn, R.; de Groot, H. Differences in glycolytic capacity and hypoxia tolerance between hepatoma cells and hepatocytes. Hepatology 13: 297–303; 1991. [10] Grynkiewicz, G.; Poenie, M.; Tsien, R. Y. A new generation of
Role of Na1 in H2O2-cytotoxicity
[11]
[12] [13]
[14] [15] [16] [17]
[18] [19] [20]
[21] [22]
Ca21 indicators with greatly improved fluorescence properties. J. Biol. Chem. 260:3440 –3450; 1985. Donoso, P.; Mill, J. G.; O’Neill, S. C.; Eisner, D. A. Fluorescence measurements of cytoplasmic and mitochondrial sodium concentration in rat ventricular myocytes. J. Physiol. 448:493–509; 1992. Grewal, K. K.; Rafeiro, E.; Racz, W. J. Bromobenzene and furosemide hepatotoxicity: alterations in glutathione, protein thiols, and calcium. Can. J. Physiol. Pharmacol. 74:257–264; 1996. Boyer, J. L.; Ananthanarayanan, M.; Hofmann, A. F.; Scheingart, C. D.; Hagenbuch, B; Stieger, B.; Meier, P. J. Expression and characterization of a functional rat liver Na1 bile acid cotransport system in COS-7 cells. Am. J. Physiol. 266:G382–G387; 1994. Sakai, H.; Kakinoki, B.; Diener, M.; Takeguchi, N. Endogenous arachidonic acid inhibits hypotonically-activated Cl2 channels in isolated rat hepatocytes. Jpn. J. Physiol. 46:311–318; 1996. Garner, M. H.; Garner, W. H.; Spector, A. Kinetic cooperativity change after H2O2 modification of (Na,K)-ATPase. J. Biol. Chem. 259:7712–7718; 1984. Matsuoka, T.; Kato, M.; Kako, K. J. Effect of oxidants on Na, K,ATPase and its reversal. Basic Res. Cardiol. 85:330 –341; 1990. Yamamoto, K.; Tsukidate, K.; Farber, J. Differing effects of the inhibition of poly(ADP-ribose)polymerase on the course of oxidative cell injury in hepatocytes and fibroblasts. Biochem. Pharmacol. 46:483– 491; 1993. Horvat, B.; Shahram, T.; Salihagic, A. Tumour cell proliferation is abolished by inhibitors of Na1/H1 and HCO32 exchange. Eur. J. Cancer 29A:132–137; 1993. Frelin, C.; Vigne, P.; Ladoux, A.; Lazdunski, M. The regulation of the intracellular pH in cells from vertebrates. Eur. J. Biochem. 174:3–14; 1988. Meharg, J. V; McGowan, J. J.; Charles, A.; Parmelee, J. T; Cutaia, M. V.; Rounds, S. Hydrogen peroxide stimulates sodiumpotassium pump activity in cultured pulmonary arterial endothelial cells. Am. J. Physiol. 265:L613–L621; 1993. Carafoli, E. Intracellular calcium homeostasis. Annu. Rev. Biochem. 56:395– 433; 1987. Blaustein, M. P. Goldman, W. F., Fontana, G., Kreuger, B. K., Santiago, E. M., Steele, T. D., Weiss, D. N., Yarowsky, P. J. Physiological roles of sodium-calcium exchanger in nerve and muscle. Ann. N. Y. Acad. Sci. 639:253–273; 1991.
719
[23] Herman, B.; Gores, G. J.; Nieminen, A.-L; Kawanishi, T.; Harman, A:; Lemasters, J. J. Calcium and pH in anoxic and toxic injury. Crit. Rev. Toxicol. 21:127–148; 1990. [24] Haigney, M.C.; Miyata, H.; Lakatta, E.G.; Stern, M.D.; Silverman, H. S. Dependence of hypoxic cellular calcium loading on Na1-Ca21 exchange. Circ. Res. 71:547–557; 1992. [25] Schlief, T; Heinemann, S. H. H2O2-induced chloride currents are indicative of an endogenous Na1-Ca21 exchange mechanism in Xenopus oocytes. J. Physiol. 486:123–130; 1995. [26] Richter C. Pro-oxidants and mitochondrial Ca21: their relationship to apoptosis and oncogenesis. FEBS-Lett. 325:104 –107; 1993. [27] Lidofsky, S. D.; Xie, M. H.; Scharmschmidt, B. F. Na1-Ca21 exchange in cultured rat hepatocytes: evidence against a role in cytosolic Ca21 regulation or signaling. Am. J. Physiol. 259:G56 – G61; 1990. [28] Nicotera, P.; Bellomo, G.; Orrenius, G. The role of Ca21 in cell killing. Chem. Res. Toxicol. 3:484 – 494; 1990. [29] Beal, M. F. Mechanisms of excitotoxicity in neurologic diseases. FASEB J. 6:3338 –3344; 1992.
ABBREVIATIONS
H2O2— hydrogen peroxide SBFI—sodium-binding benzofuran isophthalate EDTA— ethylenediaminetetraacetic acid SITS— 4-acetamido-49-isothio-cyanatostilbene-2,29disulfonic acid Hoe 694 —Hoechst 694 MEM—minimal essential medium Hepes— 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid DMSO— dimethylsulphoxide LDH—lactate dehydrogenase EGTA—(2-aminoethoxyethane)-N,N,N9,N9-tetraacetic acid