ARCHIVES
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 297, No. 2, September, pp. 265-270, 1992
Effect of Hydrogen Peroxide on Calcium Homeostasis in Smooth Muscle Cells Antonella Roveri, Mariagrazia Coassin, Matilde Maiorino, Adriana Zamburlini, Frank Th.M. van Amsterdam,* Emiliangelo Ratti,* and Fulvio Ursinitrl Department of Biological Chemistry, University of Padova, Padova, *Glaxo Research Laboratories, and TDepartment of Chemical Science and Technology, University of Udine, Italy
Received February
Verona,
26, 1992
One of the major biological targets of free radical oxidations, prone, for anatomical reasons, to oxidative challenges, is the cardiovascular system. In the present paper the effect of hydrogen peroxide on intracellular ionized calcium ([Ca2’]J homeostasis in smooth muscle cells (SMC) is studied, the major aim of the study being a better understanding of the protective effect of antioxidants and Ca2+ channel blockers. The exposure of SMC to 300 pM H202 induced a rapid increase of [Ca2+lj, followed by a decrease to a new constant level, higher than the basal before the oxidative challenge. When incubation medium was Ca2+ free, the pattern of [Ca2+]i change was different. The rapid increase was still observed, but it was followed by a rapid decrease to a level only slightly above the basal before the oxidative challenge. The involvement of intracellular Ca2+ stores was tested by using vasopressin, a hormone able to induce discharge of inositoll,4,5-triphosphate-sensitive Ca2+ stores. When H202 was added after vasopressin no [Ca2+]i increase was observed. Treatment of cells, in which the stable increase of [Ca2+], was induced by H202, with disulfide reducing compounds, induced a progressive decrease of [Ca’+]i toward the level observed before the oxidative challenge. Calcium channel blockers and antioxidants, on the other hand, effectively prevented the stabilization of [Ca2’]i at the high steady-state, after the internal Ca2+ release phase. Dihydropyridine Ca2+ channel blockers were by far more active than verapamil and among those the most active was lacidipine. Also the antioxidants trolox and N,N’-diphenyl1,4-phenylenediamine both prevented the [Ca2+]i unbalance. These results suggest that Ca+ channel blockers and antioxidants, although inactive on oxidative stress-induced Ca2’ release from intracellular stores, prevent the increased influx apparently related to a membrane thiol oxidation. 0 lee2 Academic PMS, IIN. ‘To whom correspondence should be addressed at Department of Biological Chemistry, University of Padova, v. Trieste 75, I-35121, Padova, Italy. Fax: +39-49-807-3310. 0003-9861/92 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
Hydrogen peroxide can be generated, enzymatically or nonenzymatically, in aerobic cells, by two-electron reduction of oxygen. Hydrogen peroxide is also produced by dismutation of superoxide and can generate hydroxyl radical by metal ion-catalyzed heterolytic cleavage (1). Due to its particular position in the univalent oxygen reduction pathway, hydrogen peroxide, although nonradical in nature, plays a major role among those attributed to “free oxygen radicals” (2). In fact, by challenging cells with hydrogen peroxide, the experimental condition of an unbalance between damaging oxidations and defense systems, usually referred to as “oxidative stress” (3), is generated. One of the major biological targets of free radical oxidations, which is, for anatomical reasons, prone to an oxidative challenge, is the cardiovascular system. Vascular cells are, indeed, exposed to oxidation during inflammation, endotoxic shock, ischemia-reperfusion, and possibly hypertension (4). Furthermore, the identification of peroxidized LDL’ as a possible toxic component leading to atherosclerosis (5) highlights the biological and pathological relevance of oxidations inflicted on the arterial wall. Several adverse biological effects of hydrogen peroxide on endothelial cells have been described, as well as a reactivity of the arterial wall, which is, in some respects, mimetic of physiological responses (4). Although reported results are not completely unambiguous, the pattern of oxidants causing artery relaxation by stimulating the production of endothelium-derived relaxation factor (EDRF) (6), which has been recently identified with nitric
’ Abbreviations used: BSA, bovine serum albumin; [Ca2+li, intracellular ionized calcium; DMEM, Dulbecco’s modified Eagle’s medium; DPPD, NJ’-diphenyl-1,4-phenylenediamine; DTE, dithioerythritol; DTT, dithiothreitol; EDRF, endothelium-derived relaxation factor; EGTA; ethylene glycol tetraacetate; GSH, reduced glutathione; IPB, inositol 1,4,5 triphosphate; LDL, low density lipoprotein; SMC, smooth muscle cells. 265
Inc. reserved.
266
ROVER1
oxide, seems to have emerged. On the other hand, superoxide has been shown to block the effect of EDRF, possibly by reacting with nitric oxide (7) and generating peroxynitrite, which is a strong oxidant (8). Therefore, a balance between different free radicals and oxidants seems to cope with the regulation of vascular tone under pathological conditions. Smooth muscle cells (SMC) are also exposed to oxidative challenges, which can elicit cell-specific responses. The effect of oxidants on SMC must, therefore, account for a significant part of their overall effect on the arterial wall. In the present paper we focus on the effect of hydrogen peroxide on intracellular ionized calcium [ Ca2+]i homeostasis in SMC, one of the aims of the study being a better understanding of the dramatic vascular protection effect brought about, in salt-loaded Dahl-S (9) and stroke prone hypertensive (unpublished) rats, by Lacidipine (lo), a new dihydropyridine Ca2’ channel blocker with antioxidant capacity (11). MATERIALS
AND METHODS
Materials. Fura-B/AM was from Boehringer Mannheim GmbH (Germany); Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), and trypsin-EDTA were from ICN-Flow Laboratories (Irvine, Scotland); dl-6,8-thioctic acid (oxidized form), d-biotin, vitamin Bi2 and gelatin were from Sigma Chemical Co. (St. Louis, MO); 75cm2 cell culture flasks were from Costar Europe LTD (Badhoevedorp, The Netherlands). A7r5 smooth muscle cells from rat aorta were from ATCC (Rockville, Maryland). Calcium channel blockers were supplied by Glaxo (Verona, Italy). Cell culture. A7r5 cells were cultured in DMEM supplemented with acid, and 30 nM biotin 10% FBS, 1 pM vitamin Bn, 1 pM dl-6,8-thioctic (DMEM-FBS) and grown to confluency in 75-cm2 flasks at 37°C in a humidified atmosphere of 5% COz in air. Couerslip preparation. Glass coverslips were coated with 0.1% gelatin in PBS at least 3 h before cell plating. When cells reached confluency in 75-cm2 flasks, medium was removed and cells were detached by using 3 ml of 0.05% trypsin, 0.02% EDTA (w/v). After centrifugation at 250g for -10 min, cells were split 1 to 2 and one-half of them were plated at a density of 10s cells/cm2 on gelatin-coated glass coverslips and let grow to confluency, under the same conditions used for flasks. Fura-2/AM loading. Cells were loaded with Fura-S/AM by incubating coverslips for 1 h at room temperature in the following incubation medium (IM): 125 mM NaCl, 5 mM KCl, 1.2 mM MgC&, 5 mM NaH2P04, 5 mM NaHCO,, 10 mM Hepes, 10 mM d-glucose, 1 mM CaCll, pH 7.4, containing 5 pM Fura-B/AM and 1% fatty acid-free BSA. After loading, cells were kept in IM, 1% BSA at room temperature until used. Zntracellular ionized Ca2’ measurements. A fluorescence spectrophotometer (Model RF-5000, Shimadzu Corp., Japan) equipped with a magnetic stirrer and a thermostatically controlled cuvette holder was used. Coverslips were placed into the fluorometer cuvette containing IM at an angle of 30’ to the exciting light beam as described by Di Virgilio et al. (12). [Ca2+li was calculated every 2 s by measuring fluorescence intensity with excitation at 340 and 380 nm and emission at 505 nm according to the following equation described by Grynkiewicz et al. (13): [Ca2+li
= &(R
were Kd is 224 nM and R, R,, intensity of the intracellular
-
f&.in)/(R
-
&mx)
(&2/~b2)
Rti are the ratios between the fluorescence dye at 340 nm and 380 nm in the actual
ET AL. condition, at saturating concentrations of Ca*+ and at 1 nM Ca’+, respectively. Sn and Sb, are fluorescence intensities at 380 nm for free and bound dye, respectively. Membrane depolarization was measured by using the lypophilic anion dye oxonol V, as previously described (14).
RESULTS AND DISCUSSION Smooth muscle cells attached to coverslip and loaded with Fura- were used for [Ca2’]i measurement. Computer processing of fluorescence signals allowed the display of either the fluorescence at each of the two wavelengths or the ratio between these signals. [Ca2+]i was calculated from the ratio and the reliability of measurements by this approach was confirmed since under all experimental conditions, signals recorded at the two single wavelengths were specular. The exposure to H202 of SMC in the incubation medium containing 1 mM Ca2’ induced a rapid increase of [Ca2+]i followed by a decrease to a new stable level higher than basal before the oxidative challenge (Fig. 1). The H202 concentration used was 300 PM. This concentration was used since a very slow effect was observed up to 150 PM, and a saturation (absence of higher or faster effect) was observed above 400 PM. The [Ca2’]i increase induced by the oxidative challenge was apparently due to H202 but not to hydroxyl radical, since 50 mM dimethyl thiourea, or dimethyl sulfoxide or mannitol, as well as 1 mM desferrioxamine did not show any inhibitory effect (not shown). Catalase addition during the rise of [Ca2+]i stopped the rise. On the other hand, the addition of catalase, when the new calcium steadystate was reached, was without effect. This suggested that the oxidative challenge induced a stable modification of [Ca2’]i homeostasis, which could not be repaired by the cellular systems, at least within the usual experimental time (up to 30 min). When the incubation medium was Ca2+ free (substitution of 1 mM EGTA for Ca2’), the pattern of [Ca2’]i change was different. The rapid increase was still observed, but it was followed by a rapid decrease to a level only slightly above basal before the oxidative challenge (Fig. 1). In other words, the phenomenon appeared similar to that observed in the presence of extracellular Ca2+, but the new, higher, steady-state of [Ca2+]i was prevented. This suggested that the sudden rise of [Ca2+]i required intracellular stores, while the sustained steady-state required extracellular Ca2+. The involvement of intracellular Ca2+ stores was confirmed by using vasopressin, a hormone able to induce, in cells containing the specific hormone receptor, a discharge of IP,-sensitive Ca2+ stores. In SMC, incubated in Ca2+-free medium, vasopressin induced a sudden transient increase in [Ca2f]i (Fig. 2). The recovery of the [Ca2+]i basal level was reached in a few minutes. Under these conditions, the IPB-dependent stores were depleted and further vasopressin additions had no effect. When Hz02
HYDROGEN
PEROXIDE
AND
CALCIUM
IN SMOOTH
MUSCLE
267
CELLS
B
A + Ca2+
- Ca2+
300-
300-
150-
150
ZE c r v u
Lb+--tH202
t 1
I
5
I
I
H202 I
I
5
10 TIME
I
I
10
(min)
FIG. 1. Effect of hydrogen peroxide on [Cazfli in SMC incubated in the presence (A) or in the absence (B) of extracellular Ca’+. [Ca2+li was measured by Fura- fluorescence as described under Materials and Methods. In the calcium-free incubation medium 1 mM EGTA substituted for CaCl,. Hydrogen peroxide was 300 FM, catalase was 5 rig/ml.
was added after vasopressin, no [Ca2+]i increase was observed (Fig. 2). It is noteworthy that when HeLa cells, lacking the receptors for vasopressin, were used instead of SMC, vasopressin was ineffective and H202 still caused the [Ca2+]i increase. On the other hand, when intracellular stores were depleted in HeLa cells by ionomycin the effect of H20z was also prevented (not shown). The stable higher [ Ca2+]i steady-state, induced by H202, following the phase depending on Ca2+ release from intracellular stores, was apparently due to an inbalance between Ca2+ inflow and outflow. The new [Ca2+]i steadystate, indeed, no longer required the presence of H202 and was generated only in the presence of extracellular Ca2+. Although a possible role of an inhibition of Ca’+-dependent ATPase was not ruled out, experimental evidence (see below) suggested that the inflow was positively affected by a H20,-dependent oxidation.
- Ca2+
450 z .- 300 T 1\
CVP
I
I 5 TIME
I 1lO (min)
FIG. 2. Prevention by vasopressin of hydrogen peroxide-induced calcium release from intracellular stores. SMC were incubated in calciumfree incubation medium. Vasopressin was 10 nM, hydrogen peroxide was 300 /.LM.
In fact, treatment of cells in which the stable increase of [Ca2+]i was induced by Hz02 with disulfide reducing compounds, such as DTT and DTE, but not GSH, induced a progressive decrease of [Ca’+]i toward the level observed before the oxidative challenge (Fig. 3). This suggested that the basic chemical mechanism of the [Ca2+]i homeostasis impairment was an oxidation of thiols which apparently could not be repaired from the inside of the cell. Neither antioxidants nor Ca2+ channel blockers could substitute for DTT or DTE in this effect, being ineffective once the new [Ca2’]i steady-state was reached. Calcium channel blockers and antioxidants, on the other hand, effectively prevented the stabilization of [Ca2+]i at the high steady-state after the internal Ca2+ release phase, without affecting the latter (Fig. 4A). This effect was observed only if cells were preincubated with these compounds before the challenge with H202. A doseeffect relationship was identified for all tested compounds, a lower steady-state of [Ca2+]i being produced, after the oxidative challenge, when higher concentrations of the tested compounds were used. The effect of lacidipine is reported in Fig. 4 as an example. The dose-effect relationship for several Ca2+ channel blockers and antioxidants was calculated as IDS0 by plotting the log of the concentration of the analyzed compound vs the log of a - b/b, where a is the difference in [Ca2+]i between the level before the oxidative challenge and when the stable steady-state was reached in the absence of the protective agent and b the value at the same time, but in the presence of different amounts of the protective agent. The IDS,, is the concentration of the analyzed compound to which the zero value on the y axis corresponds. The results reported in Table I indicate that dihydropyridine Ca2+ channel blockers were by far more active than verapamil and that among those the most active was lacidipine. The antioxidants trolox, a soluble vitamin E analog, and DPPD, both prevented the [Ca2+]i unbalance.
268
ROVER1
ET AL.
B
I I 1 10 5 TIME (min)
I
FIG. 3. Effect of disulfide reducing agents on [Ca2+li in SMC in which an increase of [Ca2+li was induced by 300 FM hydrogen peroxide. Incubation conditions and [Ca2+li measurements as in previous figures. Catalase was 5 pg/mg, DTE (A), DTT (B), and GSH (C) were 1 mM.
These results indicated that Ca2+ channel blockers and antioxidants are inactive on the Ca2+ release from intracellular stores but actively prevent an increased influx generated by thiol oxidation from the outside of the cells.
The most likely targets for this oxidation, as suggested by the observed protection by Ca2+ antagonists, are voltage-dependent calcium channels (16), specifically inhibited by these drugs.
A
B
I
1,
,
300
I
I
I
I
5 TIME (min)
I
10
-2 0
I 1 log [ Lacidi pine]
2
Lacidi pine concent ration a=0 b = 12.5 nM b,=25 nM b,=50 nM FIG. 4. Effect of lacidipine on [Ca2+li in SMC. Cells were incubated in the presence of different concentrations of lacidipine and challenged with 300 pM hydrogen peroxide. The recordings of [Ca*‘& in the presence of some lacidipine concentrations are reported (A). The IDSO was calculated from the plot (B) (ID6o corresponds to the value on the x axis when y = 0) where a is the difference of [Ca2+li between the level before the oxidative challenge and when the stable steady-state was reached in the absence of the protective agent and b (b,, bz . . in A) the value at the same time, but in the presence of different amounts of the protective agent. Similar plots were obtained with all tested compounds (see Table I).
HYDROGEN
PEROXIDE
;:$
AND
CALCIUM
IN SMOOTH
MUSCLE
269
CELLS
::;!;r $1j I
I
I
I
I
2.5
5
75
2.5
5
TIME
70-
2 2
KC1
2 2
I
I
7.5
I
I
50
100 150 KC1 mM
[ 1
(mid
FIG. 5. Comparison between the effects of hydrogen peroxide and KC1 on [Ca2+li in SMC. The effect of 300 pM hydrogen peroxide is reported in A and the effect of 150 mM KCl, and subsequently 300 pM hydrogen peroxide, in B. The effect of increasing KC1 concentrations on [Ca2+li is reported in C.
These results raised the question of whether H202 affects the channel directly or by inducing a depolarization. The second hypothesis was ruled out by the following evidence: (i) The stable increased [Ca’+]i steady-state induced by HzOz is higher than that caused by depolarization and the effect of H202, after KC1 depolarization, is still apparent (Fig. 5). (ii) If the membrane potential was clamped by substituting N-methyl glucamine, a nonpermeable cation, for Na+ and addition of the ionophore gramicidine (17), the effect of HzOz was still apparent (Fig. 6). (iii) The lipophilic anionic probe oxonol V (14) was taken up by cells, indicating a depolarization, after the addition of KC1 but not after the addition of H,O, (Fig. 7).
this paper suggest the involvement of at least two mechanisms: the release of intracellular IP3-sensitive stores and the alteration of the voltage-dependent channel by thiol oxidation. At present, it is not possible to discriminate which level of the cascade from receptor activation to Ca2+ release is affected by H202. Nevertheless, the possible conclusion that an oxidant mimics a cascade of physiological events which either are mediated by redox reactions, also under physiological conditions, or are just very sensitive to oxidations seems relevant. The effect of H202 on the generation of the higher [Ca2+]i steady-state, after the release of intracellular stores, seems to be related to an oxidative modification of the voltage-dependent channel presumably leading to an increased probability of the open configuration. This
CONCLUSIONS
The oxidative challenge by HzOz causes in SMC a profound unbalance of [Ca2+]i homeostasis. Data reported in
VOLTAGE CLAMPED 300
TABLE
Efficiency of Calcium Channel Blockers in Preventing the Stable Higher Steady-State Intracellular Ionized Calcium Concentration after an Oxidative Challenge” Compound
1%
Lacidipine Amlodipine Nifedipine Teludipine Nicardipine Verapamil Trolox DPPD a For incubation
conditions
z
I
(nM) 14 99 151 252 347 854 62 64
and graphical calculation
of IDr,, see Fig. 4.
l-5 +
YTJ 150 v I
I
I
I
I
5 TIME (mid
I
10
FIG. 6. Voltage clamp does not affect the perturbation of [Ca2+li induced by hydrogen peroxide. The voltage clamp was obtained by substituting in the incubation medium 140 mM N-methylglucamine for Na+ and adding 500 nM gramicidine.
270
ROVER1
ET AL.
B
I
KC1
I 1
I 2 TIME
I
“ZO, 1
I 2
(min)
FIG. 7. Evidence that KCl, but not hydrogen peroxide, induces a membrane depolarization. Depolarization was measured in SMC in a dualwavelength spectrophotometer by recording the absorbance at 630-603 nm in the presence of oxonol V, an ionic lipophilic dye, which shifts the absorbance maximum in a hydrophobic environment [see Ref. (14)]. The effect of depolarization by KC1 is reported in A and the absence of depolarization in the presence of 300 fiM hydrogen peroxide in B.
could be an important mechanism of cellular damage under conditions of oxidative stress generated from the outside, being apart from the intracellular reducing and repairing processes. In this respect the protection could be obtained pharmacologically with antioxidant, disulfide reducing agents or Ca’+ channel blockers. The mechanism of the effect of the latter seems to be the occupation of the receptor site which apparently prevents its oxidation. Furthermore, the antioxidant effect of dihydropyridine Ca2+ channel blockers, recently described, could play a specific role in preventing the oxidation of thiols involved in the suggested stabilization of the open configuration of the channel. This would depict a new cell protecting mechanism of these drugs. ACKNOWLEDGMENT The authors are grateful to Dr. Tullio Pozzan for the useful discussion and criticism during this work.
REFERENCES 1. Chance, B., Sies, H., and Boveris, A. (1979) Physiol. Reu. 49, 527605. 2. Halliwell, B., and Gutteridge, J. M. C. (1990) in Methods in Enzymology (Packer, L., and Glazer, A. N., Eds.), Vol. 186, pp. l-85, Academic Press, San Diego. 3. Sies, H. (1985) in Oxidative Press, London.
Stress (Sies, H., Ed.), pp. l-8, Academic
4. Rubanyi, G. M. (1988) Free. Radicals Biol. Med. 4, 107-120. 5. Steinberg, D. (1987) in Hypercholesterolemia and Atherosclerosis. Pathogenesis and Prevention, pp. 5-30, Churchill/Livingston, New York. 6. Rubanyi, G. M., and Vanhoutte, P. M. (1986) Am. J. Physiol. 260, H815-H821. 7. Palmer, R. M. J., Ferrige, A. G., and Moncada, S. (1987) Nature 327,524-526. 8. Radi, R., Beckman, J. S. Bush, K. M., and Freeman, B. (1991) Arch. Biochem. Biophys. 288, 481-487. 9. Cristofori, P., Terron, A., Micheli, D., Bertolini, G., Gaviraghi, G., and Carpi, C. (1991). J. Cardiouasc. Pharmacol. 17(S4), S75. 10. Micheli, D., Collodel, A., Semeraro, C., Gaviraghi, G., and Carpi, C. (1990) J. Cardiouasc. Pharmacol. 16.666-675. 11. van Amsterdam, F. Th. M., Roveri, A., Maiorino, M., Ratti, E., and Ursini, F. (1992) Free. Radicals Biol. Med., 12, 183-187. 12. Di Virgilio, F., Meyer, B. C., Greenberg, S., and Silverstein, S. C. (1988) J. Cell Biol. 106, 657-666. 13. Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985) J. Biol. Chem. 260, 3440-3450. 14. Forman, H. J., and Kim, E. (1989) Arch. Biochem. Biophys. 274, 443-452. 15. Volpe, P., Di Virgilio, F., Bruschi, G., Regolisti, G., and Pozzan, T. (1989) in Inositol Lipids in Cell Signalling (Michell, R. H., Drummond, A. R., and Downes, C. T., Eds.), pp. 377403, Academic Press, New York. 16. Pietrobon, D., Di Virgilio, F., and Pozzan, T. (1990) Eur. J. B&hem.
193,599-622. 17. Fasolato, 19636.
C., and Pozzan, T. (1989) J. Biol. Chem. 264,
19630-
OF BIOCHEMISTRY
AND
BIOPHYSICS
Vol. 297, No. 2, September, pp. 265-270, 1992
Effect of Hydrogen Peroxide on Calcium Homeostasis in Smooth Muscle Cells Antonella Roveri, Mariagrazia Coassin, Matilde Maiorino, Adriana Zamburlini, Frank Th.M. van Amsterdam,* Emiliangelo Ratti,* and Fulvio Ursinitrl Department of Biological Chemistry, University of Padova, Padova, *Glaxo Research Laboratories, and TDepartment of Chemical Science and Technology, University of Udine, Italy
Received February
Verona,
26, 1992
One of the major biological targets of free radical oxidations, prone, for anatomical reasons, to oxidative challenges, is the cardiovascular system. In the present paper the effect of hydrogen peroxide on intracellular ionized calcium ([Ca2’]J homeostasis in smooth muscle cells (SMC) is studied, the major aim of the study being a better understanding of the protective effect of antioxidants and Ca2+ channel blockers. The exposure of SMC to 300 pM H202 induced a rapid increase of [Ca2+lj, followed by a decrease to a new constant level, higher than the basal before the oxidative challenge. When incubation medium was Ca2+ free, the pattern of [Ca2+]i change was different. The rapid increase was still observed, but it was followed by a rapid decrease to a level only slightly above the basal before the oxidative challenge. The involvement of intracellular Ca2+ stores was tested by using vasopressin, a hormone able to induce discharge of inositoll,4,5-triphosphate-sensitive Ca2+ stores. When H202 was added after vasopressin no [Ca2+]i increase was observed. Treatment of cells, in which the stable increase of [Ca2+], was induced by H202, with disulfide reducing compounds, induced a progressive decrease of [Ca’+]i toward the level observed before the oxidative challenge. Calcium channel blockers and antioxidants, on the other hand, effectively prevented the stabilization of [Ca2’]i at the high steady-state, after the internal Ca2+ release phase. Dihydropyridine Ca2+ channel blockers were by far more active than verapamil and among those the most active was lacidipine. Also the antioxidants trolox and N,N’-diphenyl1,4-phenylenediamine both prevented the [Ca2+]i unbalance. These results suggest that Ca+ channel blockers and antioxidants, although inactive on oxidative stress-induced Ca2’ release from intracellular stores, prevent the increased influx apparently related to a membrane thiol oxidation. 0 lee2 Academic PMS, IIN. ‘To whom correspondence should be addressed at Department of Biological Chemistry, University of Padova, v. Trieste 75, I-35121, Padova, Italy. Fax: +39-49-807-3310. 0003-9861/92 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form
Hydrogen peroxide can be generated, enzymatically or nonenzymatically, in aerobic cells, by two-electron reduction of oxygen. Hydrogen peroxide is also produced by dismutation of superoxide and can generate hydroxyl radical by metal ion-catalyzed heterolytic cleavage (1). Due to its particular position in the univalent oxygen reduction pathway, hydrogen peroxide, although nonradical in nature, plays a major role among those attributed to “free oxygen radicals” (2). In fact, by challenging cells with hydrogen peroxide, the experimental condition of an unbalance between damaging oxidations and defense systems, usually referred to as “oxidative stress” (3), is generated. One of the major biological targets of free radical oxidations, which is, for anatomical reasons, prone to an oxidative challenge, is the cardiovascular system. Vascular cells are, indeed, exposed to oxidation during inflammation, endotoxic shock, ischemia-reperfusion, and possibly hypertension (4). Furthermore, the identification of peroxidized LDL’ as a possible toxic component leading to atherosclerosis (5) highlights the biological and pathological relevance of oxidations inflicted on the arterial wall. Several adverse biological effects of hydrogen peroxide on endothelial cells have been described, as well as a reactivity of the arterial wall, which is, in some respects, mimetic of physiological responses (4). Although reported results are not completely unambiguous, the pattern of oxidants causing artery relaxation by stimulating the production of endothelium-derived relaxation factor (EDRF) (6), which has been recently identified with nitric
’ Abbreviations used: BSA, bovine serum albumin; [Ca2+li, intracellular ionized calcium; DMEM, Dulbecco’s modified Eagle’s medium; DPPD, NJ’-diphenyl-1,4-phenylenediamine; DTE, dithioerythritol; DTT, dithiothreitol; EDRF, endothelium-derived relaxation factor; EGTA; ethylene glycol tetraacetate; GSH, reduced glutathione; IPB, inositol 1,4,5 triphosphate; LDL, low density lipoprotein; SMC, smooth muscle cells. 265
Inc. reserved.
266
ROVER1
oxide, seems to have emerged. On the other hand, superoxide has been shown to block the effect of EDRF, possibly by reacting with nitric oxide (7) and generating peroxynitrite, which is a strong oxidant (8). Therefore, a balance between different free radicals and oxidants seems to cope with the regulation of vascular tone under pathological conditions. Smooth muscle cells (SMC) are also exposed to oxidative challenges, which can elicit cell-specific responses. The effect of oxidants on SMC must, therefore, account for a significant part of their overall effect on the arterial wall. In the present paper we focus on the effect of hydrogen peroxide on intracellular ionized calcium [ Ca2+]i homeostasis in SMC, one of the aims of the study being a better understanding of the dramatic vascular protection effect brought about, in salt-loaded Dahl-S (9) and stroke prone hypertensive (unpublished) rats, by Lacidipine (lo), a new dihydropyridine Ca2’ channel blocker with antioxidant capacity (11). MATERIALS
AND METHODS
Materials. Fura-B/AM was from Boehringer Mannheim GmbH (Germany); Dulbecco’s modified Eagle’s medium (DMEM), fetal bovine serum (FBS), and trypsin-EDTA were from ICN-Flow Laboratories (Irvine, Scotland); dl-6,8-thioctic acid (oxidized form), d-biotin, vitamin Bi2 and gelatin were from Sigma Chemical Co. (St. Louis, MO); 75cm2 cell culture flasks were from Costar Europe LTD (Badhoevedorp, The Netherlands). A7r5 smooth muscle cells from rat aorta were from ATCC (Rockville, Maryland). Calcium channel blockers were supplied by Glaxo (Verona, Italy). Cell culture. A7r5 cells were cultured in DMEM supplemented with acid, and 30 nM biotin 10% FBS, 1 pM vitamin Bn, 1 pM dl-6,8-thioctic (DMEM-FBS) and grown to confluency in 75-cm2 flasks at 37°C in a humidified atmosphere of 5% COz in air. Couerslip preparation. Glass coverslips were coated with 0.1% gelatin in PBS at least 3 h before cell plating. When cells reached confluency in 75-cm2 flasks, medium was removed and cells were detached by using 3 ml of 0.05% trypsin, 0.02% EDTA (w/v). After centrifugation at 250g for -10 min, cells were split 1 to 2 and one-half of them were plated at a density of 10s cells/cm2 on gelatin-coated glass coverslips and let grow to confluency, under the same conditions used for flasks. Fura-2/AM loading. Cells were loaded with Fura-S/AM by incubating coverslips for 1 h at room temperature in the following incubation medium (IM): 125 mM NaCl, 5 mM KCl, 1.2 mM MgC&, 5 mM NaH2P04, 5 mM NaHCO,, 10 mM Hepes, 10 mM d-glucose, 1 mM CaCll, pH 7.4, containing 5 pM Fura-B/AM and 1% fatty acid-free BSA. After loading, cells were kept in IM, 1% BSA at room temperature until used. Zntracellular ionized Ca2’ measurements. A fluorescence spectrophotometer (Model RF-5000, Shimadzu Corp., Japan) equipped with a magnetic stirrer and a thermostatically controlled cuvette holder was used. Coverslips were placed into the fluorometer cuvette containing IM at an angle of 30’ to the exciting light beam as described by Di Virgilio et al. (12). [Ca2+li was calculated every 2 s by measuring fluorescence intensity with excitation at 340 and 380 nm and emission at 505 nm according to the following equation described by Grynkiewicz et al. (13): [Ca2+li
= &(R
were Kd is 224 nM and R, R,, intensity of the intracellular
-
f&.in)/(R
-
&mx)
(&2/~b2)
Rti are the ratios between the fluorescence dye at 340 nm and 380 nm in the actual
ET AL. condition, at saturating concentrations of Ca*+ and at 1 nM Ca’+, respectively. Sn and Sb, are fluorescence intensities at 380 nm for free and bound dye, respectively. Membrane depolarization was measured by using the lypophilic anion dye oxonol V, as previously described (14).
RESULTS AND DISCUSSION Smooth muscle cells attached to coverslip and loaded with Fura- were used for [Ca2’]i measurement. Computer processing of fluorescence signals allowed the display of either the fluorescence at each of the two wavelengths or the ratio between these signals. [Ca2+]i was calculated from the ratio and the reliability of measurements by this approach was confirmed since under all experimental conditions, signals recorded at the two single wavelengths were specular. The exposure to H202 of SMC in the incubation medium containing 1 mM Ca2’ induced a rapid increase of [Ca2+]i followed by a decrease to a new stable level higher than basal before the oxidative challenge (Fig. 1). The H202 concentration used was 300 PM. This concentration was used since a very slow effect was observed up to 150 PM, and a saturation (absence of higher or faster effect) was observed above 400 PM. The [Ca2’]i increase induced by the oxidative challenge was apparently due to H202 but not to hydroxyl radical, since 50 mM dimethyl thiourea, or dimethyl sulfoxide or mannitol, as well as 1 mM desferrioxamine did not show any inhibitory effect (not shown). Catalase addition during the rise of [Ca2+]i stopped the rise. On the other hand, the addition of catalase, when the new calcium steadystate was reached, was without effect. This suggested that the oxidative challenge induced a stable modification of [Ca2’]i homeostasis, which could not be repaired by the cellular systems, at least within the usual experimental time (up to 30 min). When the incubation medium was Ca2+ free (substitution of 1 mM EGTA for Ca2’), the pattern of [Ca2’]i change was different. The rapid increase was still observed, but it was followed by a rapid decrease to a level only slightly above basal before the oxidative challenge (Fig. 1). In other words, the phenomenon appeared similar to that observed in the presence of extracellular Ca2+, but the new, higher, steady-state of [Ca2+]i was prevented. This suggested that the sudden rise of [Ca2+]i required intracellular stores, while the sustained steady-state required extracellular Ca2+. The involvement of intracellular Ca2+ stores was confirmed by using vasopressin, a hormone able to induce, in cells containing the specific hormone receptor, a discharge of IP,-sensitive Ca2+ stores. In SMC, incubated in Ca2+-free medium, vasopressin induced a sudden transient increase in [Ca2f]i (Fig. 2). The recovery of the [Ca2+]i basal level was reached in a few minutes. Under these conditions, the IPB-dependent stores were depleted and further vasopressin additions had no effect. When Hz02
HYDROGEN
PEROXIDE
AND
CALCIUM
IN SMOOTH
MUSCLE
267
CELLS
B
A + Ca2+
- Ca2+
300-
300-
150-
150
ZE c r v u
Lb+--tH202
t 1
I
5
I
I
H202 I
I
5
10 TIME
I
I
10
(min)
FIG. 1. Effect of hydrogen peroxide on [Cazfli in SMC incubated in the presence (A) or in the absence (B) of extracellular Ca’+. [Ca2+li was measured by Fura- fluorescence as described under Materials and Methods. In the calcium-free incubation medium 1 mM EGTA substituted for CaCl,. Hydrogen peroxide was 300 FM, catalase was 5 rig/ml.
was added after vasopressin, no [Ca2+]i increase was observed (Fig. 2). It is noteworthy that when HeLa cells, lacking the receptors for vasopressin, were used instead of SMC, vasopressin was ineffective and H202 still caused the [Ca2+]i increase. On the other hand, when intracellular stores were depleted in HeLa cells by ionomycin the effect of H20z was also prevented (not shown). The stable higher [ Ca2+]i steady-state, induced by H202, following the phase depending on Ca2+ release from intracellular stores, was apparently due to an inbalance between Ca2+ inflow and outflow. The new [Ca2+]i steadystate, indeed, no longer required the presence of H202 and was generated only in the presence of extracellular Ca2+. Although a possible role of an inhibition of Ca’+-dependent ATPase was not ruled out, experimental evidence (see below) suggested that the inflow was positively affected by a H20,-dependent oxidation.
- Ca2+
450 z .- 300 T 1\
CVP
I
I 5 TIME
I 1lO (min)
FIG. 2. Prevention by vasopressin of hydrogen peroxide-induced calcium release from intracellular stores. SMC were incubated in calciumfree incubation medium. Vasopressin was 10 nM, hydrogen peroxide was 300 /.LM.
In fact, treatment of cells in which the stable increase of [Ca2+]i was induced by Hz02 with disulfide reducing compounds, such as DTT and DTE, but not GSH, induced a progressive decrease of [Ca’+]i toward the level observed before the oxidative challenge (Fig. 3). This suggested that the basic chemical mechanism of the [Ca2+]i homeostasis impairment was an oxidation of thiols which apparently could not be repaired from the inside of the cell. Neither antioxidants nor Ca2+ channel blockers could substitute for DTT or DTE in this effect, being ineffective once the new [Ca2’]i steady-state was reached. Calcium channel blockers and antioxidants, on the other hand, effectively prevented the stabilization of [Ca2+]i at the high steady-state after the internal Ca2+ release phase, without affecting the latter (Fig. 4A). This effect was observed only if cells were preincubated with these compounds before the challenge with H202. A doseeffect relationship was identified for all tested compounds, a lower steady-state of [Ca2+]i being produced, after the oxidative challenge, when higher concentrations of the tested compounds were used. The effect of lacidipine is reported in Fig. 4 as an example. The dose-effect relationship for several Ca2+ channel blockers and antioxidants was calculated as IDS0 by plotting the log of the concentration of the analyzed compound vs the log of a - b/b, where a is the difference in [Ca2+]i between the level before the oxidative challenge and when the stable steady-state was reached in the absence of the protective agent and b the value at the same time, but in the presence of different amounts of the protective agent. The IDS,, is the concentration of the analyzed compound to which the zero value on the y axis corresponds. The results reported in Table I indicate that dihydropyridine Ca2+ channel blockers were by far more active than verapamil and that among those the most active was lacidipine. The antioxidants trolox, a soluble vitamin E analog, and DPPD, both prevented the [Ca2+]i unbalance.
268
ROVER1
ET AL.
B
I I 1 10 5 TIME (min)
I
FIG. 3. Effect of disulfide reducing agents on [Ca2+li in SMC in which an increase of [Ca2+li was induced by 300 FM hydrogen peroxide. Incubation conditions and [Ca2+li measurements as in previous figures. Catalase was 5 pg/mg, DTE (A), DTT (B), and GSH (C) were 1 mM.
These results indicated that Ca2+ channel blockers and antioxidants are inactive on the Ca2+ release from intracellular stores but actively prevent an increased influx generated by thiol oxidation from the outside of the cells.
The most likely targets for this oxidation, as suggested by the observed protection by Ca2+ antagonists, are voltage-dependent calcium channels (16), specifically inhibited by these drugs.
A
B
I
1,
,
300
I
I
I
I
5 TIME (min)
I
10
-2 0
I 1 log [ Lacidi pine]
2
Lacidi pine concent ration a=0 b = 12.5 nM b,=25 nM b,=50 nM FIG. 4. Effect of lacidipine on [Ca2+li in SMC. Cells were incubated in the presence of different concentrations of lacidipine and challenged with 300 pM hydrogen peroxide. The recordings of [Ca*‘& in the presence of some lacidipine concentrations are reported (A). The IDSO was calculated from the plot (B) (ID6o corresponds to the value on the x axis when y = 0) where a is the difference of [Ca2+li between the level before the oxidative challenge and when the stable steady-state was reached in the absence of the protective agent and b (b,, bz . . in A) the value at the same time, but in the presence of different amounts of the protective agent. Similar plots were obtained with all tested compounds (see Table I).
HYDROGEN
PEROXIDE
;:$
AND
CALCIUM
IN SMOOTH
MUSCLE
269
CELLS
::;!;r $1j I
I
I
I
I
2.5
5
75
2.5
5
TIME
70-
2 2
KC1
2 2
I
I
7.5
I
I
50
100 150 KC1 mM
[ 1
(mid
FIG. 5. Comparison between the effects of hydrogen peroxide and KC1 on [Ca2+li in SMC. The effect of 300 pM hydrogen peroxide is reported in A and the effect of 150 mM KCl, and subsequently 300 pM hydrogen peroxide, in B. The effect of increasing KC1 concentrations on [Ca2+li is reported in C.
These results raised the question of whether H202 affects the channel directly or by inducing a depolarization. The second hypothesis was ruled out by the following evidence: (i) The stable increased [Ca’+]i steady-state induced by HzOz is higher than that caused by depolarization and the effect of H202, after KC1 depolarization, is still apparent (Fig. 5). (ii) If the membrane potential was clamped by substituting N-methyl glucamine, a nonpermeable cation, for Na+ and addition of the ionophore gramicidine (17), the effect of HzOz was still apparent (Fig. 6). (iii) The lipophilic anionic probe oxonol V (14) was taken up by cells, indicating a depolarization, after the addition of KC1 but not after the addition of H,O, (Fig. 7).
this paper suggest the involvement of at least two mechanisms: the release of intracellular IP3-sensitive stores and the alteration of the voltage-dependent channel by thiol oxidation. At present, it is not possible to discriminate which level of the cascade from receptor activation to Ca2+ release is affected by H202. Nevertheless, the possible conclusion that an oxidant mimics a cascade of physiological events which either are mediated by redox reactions, also under physiological conditions, or are just very sensitive to oxidations seems relevant. The effect of H202 on the generation of the higher [Ca2+]i steady-state, after the release of intracellular stores, seems to be related to an oxidative modification of the voltage-dependent channel presumably leading to an increased probability of the open configuration. This
CONCLUSIONS
The oxidative challenge by HzOz causes in SMC a profound unbalance of [Ca2+]i homeostasis. Data reported in
VOLTAGE CLAMPED 300
TABLE
Efficiency of Calcium Channel Blockers in Preventing the Stable Higher Steady-State Intracellular Ionized Calcium Concentration after an Oxidative Challenge” Compound
1%
Lacidipine Amlodipine Nifedipine Teludipine Nicardipine Verapamil Trolox DPPD a For incubation
conditions
z
I
(nM) 14 99 151 252 347 854 62 64
and graphical calculation
of IDr,, see Fig. 4.
l-5 +
YTJ 150 v I
I
I
I
I
5 TIME (mid
I
10
FIG. 6. Voltage clamp does not affect the perturbation of [Ca2+li induced by hydrogen peroxide. The voltage clamp was obtained by substituting in the incubation medium 140 mM N-methylglucamine for Na+ and adding 500 nM gramicidine.
270
ROVER1
ET AL.
B
I
KC1
I 1
I 2 TIME
I
“ZO, 1
I 2
(min)
FIG. 7. Evidence that KCl, but not hydrogen peroxide, induces a membrane depolarization. Depolarization was measured in SMC in a dualwavelength spectrophotometer by recording the absorbance at 630-603 nm in the presence of oxonol V, an ionic lipophilic dye, which shifts the absorbance maximum in a hydrophobic environment [see Ref. (14)]. The effect of depolarization by KC1 is reported in A and the absence of depolarization in the presence of 300 fiM hydrogen peroxide in B.
could be an important mechanism of cellular damage under conditions of oxidative stress generated from the outside, being apart from the intracellular reducing and repairing processes. In this respect the protection could be obtained pharmacologically with antioxidant, disulfide reducing agents or Ca’+ channel blockers. The mechanism of the effect of the latter seems to be the occupation of the receptor site which apparently prevents its oxidation. Furthermore, the antioxidant effect of dihydropyridine Ca2+ channel blockers, recently described, could play a specific role in preventing the oxidation of thiols involved in the suggested stabilization of the open configuration of the channel. This would depict a new cell protecting mechanism of these drugs. ACKNOWLEDGMENT The authors are grateful to Dr. Tullio Pozzan for the useful discussion and criticism during this work.
REFERENCES 1. Chance, B., Sies, H., and Boveris, A. (1979) Physiol. Reu. 49, 527605. 2. Halliwell, B., and Gutteridge, J. M. C. (1990) in Methods in Enzymology (Packer, L., and Glazer, A. N., Eds.), Vol. 186, pp. l-85, Academic Press, San Diego. 3. Sies, H. (1985) in Oxidative Press, London.
Stress (Sies, H., Ed.), pp. l-8, Academic
4. Rubanyi, G. M. (1988) Free. Radicals Biol. Med. 4, 107-120. 5. Steinberg, D. (1987) in Hypercholesterolemia and Atherosclerosis. Pathogenesis and Prevention, pp. 5-30, Churchill/Livingston, New York. 6. Rubanyi, G. M., and Vanhoutte, P. M. (1986) Am. J. Physiol. 260, H815-H821. 7. Palmer, R. M. J., Ferrige, A. G., and Moncada, S. (1987) Nature 327,524-526. 8. Radi, R., Beckman, J. S. Bush, K. M., and Freeman, B. (1991) Arch. Biochem. Biophys. 288, 481-487. 9. Cristofori, P., Terron, A., Micheli, D., Bertolini, G., Gaviraghi, G., and Carpi, C. (1991). J. Cardiouasc. Pharmacol. 17(S4), S75. 10. Micheli, D., Collodel, A., Semeraro, C., Gaviraghi, G., and Carpi, C. (1990) J. Cardiouasc. Pharmacol. 16.666-675. 11. van Amsterdam, F. Th. M., Roveri, A., Maiorino, M., Ratti, E., and Ursini, F. (1992) Free. Radicals Biol. Med., 12, 183-187. 12. Di Virgilio, F., Meyer, B. C., Greenberg, S., and Silverstein, S. C. (1988) J. Cell Biol. 106, 657-666. 13. Grynkiewicz, G., Poenie, M., and Tsien, R. Y. (1985) J. Biol. Chem. 260, 3440-3450. 14. Forman, H. J., and Kim, E. (1989) Arch. Biochem. Biophys. 274, 443-452. 15. Volpe, P., Di Virgilio, F., Bruschi, G., Regolisti, G., and Pozzan, T. (1989) in Inositol Lipids in Cell Signalling (Michell, R. H., Drummond, A. R., and Downes, C. T., Eds.), pp. 377403, Academic Press, New York. 16. Pietrobon, D., Di Virgilio, F., and Pozzan, T. (1990) Eur. J. B&hem.
193,599-622. 17. Fasolato, 19636.
C., and Pozzan, T. (1989) J. Biol. Chem. 264,
19630-
