Bioscience Reports, Vol. 12, No. 6, 1992
Effect of Reactive Oxygen Species on the Erythrocyte Calcium-Pump Function in Protein-Energy Malnutrition W. G. Okunade L2 and O. O. Olorunsogo I Received June 6, 1992 The presence of detectable amounts of non-heme iron in erythrocyte ghost membranes have been postulated to lead to the initiation of membrane lipid peroxidation and the attendant perturbation of membrane functions. We have investigated the presence of non-heme iron and endogenous products of lipid peroxidation in erythrocyte membranes of normal and kwashiorkor (KWA) subjects and asse,ssed the susceptibility of the membranes to exogenously generated reactive oxygen species. The modulation of the basal and calmodulin-stimulated calcium-pumping activity of these membranes by reactive oxygen species was also assessed. The results show the presence of significant amounts of non-heme iron and endogenous free radical reaction products in the red cell membranes of KWA subjects compared with that of normal children. Estimation of the extent of lipid peroxidation in the presence of exogenously generated reactive oxygen species further revealed that erythrocyte ghost membranes of KWA subjects are more susceptible to oxidative stress than those of normal individuals. Although both the basal and calmodulin-stimulated activities of the membrane-bound Ca2+-pump enzyme in normal and KWA subjects were inhibited by oxygen-free radicals, the erythrocyte enzyme in KWA subjects showed higher susceptibility to inhibition by oxygen free radicals than that of normal individuals. We propose that the reduced erythrocyte calcium-pump function in K W A is not unconnected with excessive generation of reactive oxygen species. KEY WORDS: Reactive oxygen species; non-heme iron; CA2+-ATPase; erythrocyte; kwashiokor. ABBREVIATIONS: PMSF; phenylmethylsulfonylfuloride; TLCK; N-0~-p-tosyl-L-lysine chloromethylketone; EGTA ethyleneglycol-bis (B-aminoethylether) N,N'-tetraacetic acid; EDTA; ethylene diamine tetraacetic acid; ATP; Adenosine 5-triphosphate; Hepes; 4-(2-hydroxyethyl)-l-piperazine ethanesulphonic acid; Tris-HCl, Tris (hydroxymethyl) aminomethane-hydrochloride; SDS; Sodium dodecyl sulphate; TBAR; thiobarbituric acid-reactive products TBA, thiobarbituric acid; TCA; trichloroacetic acid.
INTRODUCTION C o n s i d e r a b l e a t t e n t i o n is p r e s e n t l y b e i n g f o c u s s e d o n t h e i n f l u e n c e o f o x y g e n d e r i v e d f r e e r a d i c a l s o n t h e i n t e g r i t y a n d f u n c t i o n s o f cell m e m b r a n e s in d i s e a s e Laboratory for Biomembrane Research, Department of Biochemistry, University of Ibadan, Ibadan, Nigeria. a To whom correspondence should be addressed. 433
0144-8463/92/1200-0433506.50/0~ 1992PlenumPublishingCorporation
434
Okunade and Olorunsogo
states in view of the cytotoxic effects of reactive oxygen species on normal cells (1). In this connection, activated oxygen species have been implicated in certain pathological conditions such as cancer (2), rheumatoid arthritis (3) haemoglobinopathies (4), parasitic infection (5) and a variety of ischaemia (6, 7). It is now well established that when reactive oxygen species evade or overwhelm the protective mechanisms of cells and tissues, they elicit metabolic and cellular disturbances as a result of peroxidation of membrane lipid (8). Although severe dietary protein deficiency in the presence of relative carbohydrate excess is generally accepted as the cause of kwashiokor (KWA) (9), a well-defined clinical syndrome at one end of protein-energy malnutrition (PEM) spectrum, the various clinical and biochemical features associated with the disease have made several investigators to put forward a number of hypotheses in order to explain the biochemical basis of the condition (10, 11, 12, 13, 14). Indeed, the reports of Gopalan (1968) (11) and later Golden (1985) (14) that there are no essential differences in the protein: energy ratios of the diets of children who developed marasmus and those with kwashiokor made the concept of lack of protein in the diet to fall out of favour. Based on the demonstration in PEM of an increased flux of products of free radical action, a decrease in all the protective pathways for free radicals and an increase in tissue iron (14) Golden and Ramdath (1987) (15) suggested that kwashiokor probably results from an imbalance between the production of free radicals and the protective mechanisms that remove them in the malnourished state. Since one of the clinical features in kwashiokor is anaemia, we have investigated in this study, erythrocytes from kwashiokor children because of erythrocyte membrane susceptibility to reactive oxygen species and the effect of these species on cell membrane calcium pumping.
MATERIALS AND METHODS Thiobarbituric acid, TLCK, PMSF, Ferrozine, Soybean trypsin inhibitor, disodium salt of ATP, and thiourea were purchased from Sigma Chem. Co., London. Calmodulin was purchased from Calbiochemical, La Jolla, California. All other reagents used were of analytical grade and were purchased from British Drug Houses, Poole, England.
Preparation of Erythrocyte Ghost Membrane for Ca2+-ATPase Studies Haemoglobin-free ghost membranes deficient in calmodulin were prepared essentially according to the procedure of Niggli et al. (1981) (16) which is based on the principle of hypotonic lysis developed by Dodge et al. (1963) (17). All steps of the membrane preparation were carried out at 4~ Blood samples were obtained from children who were newly identified at the University College, Oni Memorial, and Adeoyo State Hospitals, Ibadan, as having severe kwashiokor. As
Erythrocyte Calcium-Pump Function
435
a rule, none of the subjects received any dietary therapy or medication prior to the: time blood was collected. Normal human blood was collected from healthy children in the same age bracket and was used as control. All blood samples were collected in acid-citrate-dextrose buffer, stored at 4~ and used within 24 hours. Whole blood samples were centrifuged at 5800 rpm for 10 mins, the plasma and "buffy" layers were removed by aspiration to obtain packed erythrocytes. The erythrocytes were washed thrice in 10 volumes of 130 mM KC1, 10 mM Tris, phi 7.4 to remove plasma constituents. Each time, the cell suspension was centrifuged at 5800 rpm and the supernatant removed by aspiration. Washed erythrocytes were lysed in 10 volumes of 1 mM Na-EDTA, 10 mM Tris pH 7.4 containing 1 #M soybean trypsin inhibitor, 10/~M TLCK and 200/~M PMSF and centrifuged at 18,000 rpm for 20 rains. Membranes were washed four times in the hemolysis solution and eight times in 10 mM HEPES pH 7.4 containing 200 gM PMSF in order to remove residual calmodulin, haemoglobin and EDTA. The haemoglobin-free ghosts were finally resuspended and centrifuged at 18,000 rpm for 20 mins in a 20 mM HEPES, pH 7.4 storage buffer containing 130 mM KCI, 500/~M MgCI2 and 50/~M CaCI2. The membrane pellet was suspended and stored in the same buffer at -60~
Preparation of Erythrocyte Ghost Membranes for Studies on Membrane Peroxidation and Non-heme Iron Determination Haemoglobin-free ghost membranes were prepared by a slight modification of the above procedure. Blood samples were centrifuged at 2500 rpm for 10 mins followed by removal of the plasma and buffy layers by aspiration. The packed cells were suspended in 10 volumes of cold isotonic phosphate buffer pH 7.4 and centrifuged at same speed for 10 minutes. This was repeated once and any remaining buffy coat was carefully aspirated. The packed cells were then resuspended in 10 volumes of cold hypotonic 5raM phosphate buffer pH 7.4 and allowed to stand on ice for 30 mins. Centrifugation was later carried out at 16,000rpm for 10 mins at 4~ The supernatant as well as the dense pellet at the bottom of the tubes were carefully aspirated. The membranes obtained were washed four more times in 10 vols. of the same buffer and later resealed by resuspension in 10 volumes of the isotonic phosphate buffer pH 7.4 followed by centrifugation at 16,000 rpm for 10 mins. The membranes were stored in the same buffer at 4~ and used within 24 hours.
Protein Determination Membrane concentration was assessed .by determining protein concentration according to the method of Lowry et aL (i95t) (I8) using bovine serum albumin as standard.
436
Okunade and Olorunsogo
Incubation of Erythrocyte Ghost Membranes with Oxygen Free Radicals. Calmodulin-deficient, haemoglobin-free erythrocyte ghost membranes were pelleted by centrifugation at 20,000 rpm for 30 mins and resuspended in isotonic phosphate buffer pH 7.4. i mg/ml final concentration of the membranes were incubated for 30 mins in phosphate-buffered saline pH 7.4 containing premixed ferrous sulphate with ATP to form an ATP-Fe z+ complex at a final concentration of 1 mM ATP, 0.1 mM FeSO4. Hydrogen peroxide was then added to a final concentration of 5 mM to initiate the free radical reaction. Control incubation was also carried out by incubating erythrocyte ghost membrane in the phosphatebuffered saline pH 7.4 without any of the additives. The reaction was stopped by diluting the incubation mixture with 5 vols. of ice-cold isotonic Tris buffer (130 mM KC1 + 20 mM Tris pH 7.4) containing 5 mM Mannitol. The ghost membranes were wash-centrifuged twice with the same buffer but without 5 mM-Mannitol and once with 130mM KC1, 20mM K § HEPES pH 7.4, 50 #M CaClz and 500 ~tM MgCI2. The ghost membranes were finally suspended in the latter buffer and used to estimate Ca2+-pumping ATPase activity.
Assay of Ca2+-ATPase Activity Ca2+-ATPase activity was assayed by measuring the rate of release of inorganic phosphate from the -position of ATP. The phosphate liberated to the medium was determined colori-metrically using the method of Raess and Vincenzi (1980) (19) based on the procedure of Fiske and Subbarow (1925) (20). Calmodulin-depleted membranes (100-200 #g) were incubated at 37~ for five minutes in a final total volume of 1 ml. The assay medium contained 100 mM KCI, 50 mM K-HEPES pH 7.4, 2 mM MgCI2, 20 ~M CaC12, 2 mM EGTA (when added), and 2 ~g/ml calmodulin (when added). The reaction was started by adding 2 mM ATP (final concentration). After 30 minutes of incubation at 37~ with constant shaking, the reaction was stopped with 1 ml of 10% solution of sodium dodecyl sulphate (SDS) in distilled water. All assays were run in duplicate with and without calmodulin.
Incubation of Erythrocytes with Oxygen Free Radicals Exposure of red cells to activated oxygen species was carried out essentially according to the procedure described by Hebbel et al. (1986) (22). Packed erythrocytes were washed three times with 150 mM NaCI solution and suspended to 20% haematocrit in Hank's balanced salt solution containing 5.6 mM glucose, 136.9 mM NaCI, 5.4mM KCI, 0.3 mM NazHPO4, 0.3 mM KH PO4, 0.4 mM MgSO4, 13 mM CaC12, 0.5 mM MgCI: and 4.2 mM NaHCO3 (pH 7.4). They were incubated in a shaking water bath at 37~ for 40 rains with or without premixed ferrous sulphate with ATP to form an ATP-Fe 2+ complex at a final concentration of l mM ATP, 0.1mM FeSO#, and 5raM hydrogen peroxide to initiate the reaction. After 40 minutes, 5 vols of ice-cold 150 mM NaCI was added to the
Erythrocyte Calcium-Pump Function
437
reaction mixture followed by centrifugation at 5800 rpm for 10 mins. The packed red cells were suspended to 10% Hct in 175 mM Tris-HCl pH 7.4 and admixed wffh saponin at a ratio of 10 ~g saponin per 10 ~1 erythrocytes. After 15 mins incubation at room temperature, any unlysed red cells were removed by low-speed centrifugation. 0.1 ml red cells lysate was whirlmixed with 0.2rnl of 175 mM Tris-HC1 pH 7.4 and incubated for 30 mins at 37~ in a final volume of 1 ml containing (in final concentrations) 20 mM imidazole, 60 mM KC1, 10 mM NaC1, 3 mM MgCI2, 0.2mM Ouabain, 20 laM CaCI2 (or 200/zM EGTA) and 4 mM ATP. The reaction which was run in duplicate was stopped by addition of 0.1 ml 50% trichloroacetic acid and centrifuged to obtain a clear supernatant. The inorganic phosphate in the supernatant was determined by adding 1 ml of 1.25% ammonium molybdate in 6.5% H2804, followed by 1 ml of 9% ascorbic acid. The absorbance of the blue colour developed was measured at 660 nm. CaZ+-ATPase activity was calculated by substracting the Pi content of the RGTA-containing sample (i.e. Mg2+-ATPase activity) from that of the calcium containing sample (i~ Ca 2+ + M f § activities). Determination of Membrane-bound non-haeme Iron Membrane-bound non-heme iron was determined according to a modified method of Ceriotti and Ceriotti (1980). (21) Final concentration of reagents used were 38 mM ascorbate, 0.67% SDS, 10 mM phosphate buffer pH 7.4, 150 mM NaC1 and 1.28 mM ferrozine in a total volume of 3.0 ml. The intensity of the pink-coloured complex formed was determined by measuring absorbance at 562 nm using distilled water as blank. Standards were simultaneously run with a stock solution of 25 mg/L iron. Assessment of Lipid Peroxidation Lipid peroxidation was assayed by measuring the thiobarbituric acid-reactive (TBAR) products using the procedure of Walls et al. (1976) (23) as previously described (24). 1 mg/ml final concentration of membranes in isotonic phosphate buffer pH 7.4, were incubated for six hours at 37~ in a shaking water bath with oi without 0.1 mM FeSOa, 1 mM EDTA, 1 mM ascorbate and 0.2 mM H202 (final concentrations). 0.5 ml of 0.75% TBA in 0.1 M HCI was added to 0.5 ml of the incubation mixture already quenched with 0.5 ml of 10% TCA. The mixture was heated at 90-95~ for 20 mins and after cooling, centrifuged for 10 mins at 2,500r.p.m. The supernatant was transferred into acid resistant tubes and centrifuged at 16,000 rpm for 10 mins. The absorbance of the resulting clear solution was determined at 532 nm with phosphate buffer as blank. RESULTS Estimation of Membrane-bound Non-heme Iron and Susceptibifity to Lipid Peroxidation The procedure used in this study to determine non-heme iron is based on the reduction of iron by ascorbic acid (21). Addition of ferrozine to this mixture
438
Okunade and Olorunsogo 10"
cO7
e-
o.c_ n=6 ~
0
g ~
E
e"
a~
0 I N
! KWA
Fig. 1. Erythrocyte ghost membrane-bound nonheine Iron in healthy (N) and kwashiokor (KWA) children. Non-heme Iron was estimated by Ferrozine assay as described in Materials and Methods and expressed as nmotesnon-hemeIron per mg protein.
results in the formation of a complex with iron that is intensely coloured. SDS was included in the assay to ensure solubilisation and denaturation of membrane proteins. Although the absorbance of the pink chromophore was read at 6 hrs and 18 hrs after addition of ferrozine, the results (Fig. 1) obtained did not differ markedly thus indicating that all the non-heme iron have been released into solution. The results as presented in Fig. 1 indicate the absence of non-heme iron in haemoglobin-free erythrocyte ghost membranes (EGM) prepared from red cells of healthy children. In contrast, non-heme iron is present in significant amounts in ghost membranes prepared from red cells obtained from kwashiokor children even though there was considerable variation in the amounts observed among the ghost membranes of these children. The extent of endogenous lipid peroxidation was determined in these membranes because of the significant amounts of non-heme iron associated with the ghost membranes. The results show that the extent of lipid peroxidation in the EGM of KWA subjects was about 3-fold that determined for red cell membranes of normal individuals (Fig. 2). Furthermore, exposure of the membranes to exogenously generated reactive oxygen species results in a significant increase in the amounts of lipid peroxidation products in both membranes. A 3-fold increase in the amount of Thiobar-bituric acid-reactive (TBAR) products was observed for EGM membranes of normal individuals. This is about half of the value obtained for similarly treated red cell membranes of KWA subjects and about 80% of the value obtained for untreated EGM of KWA subjects. Moreover, exogenously generated reactive oxygen species produced an almost 2-fold increase over the basal value of Thiobarbituric acid-reactive (TBAR) products in EGM of KWA subjects. These results reflect a state of oxidative
Erythrocyte Calcium-Pump Function
439
3.0n=lO
I c
m
Q_ iIII
E
n=lO
r~
n=10
r
l-IJ1
1.0-
0
E
I
n=10
H I
N
I
KWA
I
N§
I
KV~I&§ ROS
Fig. 2. Degree of lipid peroxidation of erythrocyte ghost membrane of healthy (N) and kwashiokor (KWA) children in the presence and absence of exogenously generated reactive oxygen species (ROS). Lipid peroxidation was estimated as described in Material and Methods and expressed as nmoles Thiobarbituric reactive (TBAR) products per mg protein.
stress in the KWA subjects and also provide an indication that red cells in KWA individuals are highly susceptible to peroxidation by activated oxygen species.
Ca2+-ATPase Studies The CaZ+-ATPase assay used in this study for intact red cells utilized the red cells own cytosol as a source of calmodulin so that the activity measured is the calmodulin-stimulated CA2*-ATPase activity. Table 1 shows that the activity of the pump in red cells of KWA subjects is about 57% lower than that of the enzyme in red cells of normal individuals. Exposure of the washed erythrocytes fiom KWA and normal subjects to reactive oxygen species results in a significant inhibition of the CaM-activated CaZ+-pumping ATPase activity even though the degree of inhibition differs. While the CaZ+-ATPase activity of red cells from Table 1. Effect of Exogenously-generated free radical on the activity of CaZ+-ATPase in intact Red cells from normal and kwashiokor individuals Erythrocyte Samples from:
CaZ+-ATPase activity umole Pi release glib -1 hr -1
% of Control
82-56 + 6.2 49.40 -t- 9.3 35.60 5:4,7
59.8
11.20 + 3.8
31.5
Normal Normal + Free Radicals Kwashiokor Kwashiokor + Free Radicals Each value is a mean of four experiments.
Okunade and Olorunsogo
440
Table 2. Effect of Exogenously-generatedfree radicals on the cae+-ATPase activity
of Erythrocyte ghost membranes in kwashiokor and Normal Subjects. Ghost Membranes Samples from erythrocytes of: Normal Normal + Free Radicals Kwashiokor Kwashiokor + Free Radicals
Mga+-ATPase
Caz+ + MgZ+-ATPase activity umole Pimg protein -1 hr -1 - CaM + CaM
0.36 5:0.02 0.28 5:0.03 0.18 5:0.03
0.70 5:0.05 0.43 5:0.07 0.38 + 0.04
2.89 5:0.17 1.71 5:0.08 0.88 :t: 0.06
0.11 + 0.04
0.15 5:0.03
0.31 5:0.04
Each value is a mean of four experiments. normal individuals was reduced to about 60% of control, (40% inhibition) the cae§ activity of erythrocytes from K W A subjects was inhibited to about 30% of their controls (70% inhibition). The results obtained with erythrocyte ghost membranes followed the same trend as that observed with intact erythrocytes (Table 2). The Ca2§ ATPase activity of erythrocyte plasma m e m b r a n e of K W A subjects show, in the absence of calmodulin, a lower specific activity (46%) than the enzyme of the erythrocyte membrane of normal individuals. Although the addition of calmodulin increased the specific activity of the calcium pump of erythrocyte ghost membrane of K W A and normal individuals by at least 2-fold and 4-fold, respectively, the stimulation of the pump protein in membranes from K W A individuals is about 70% lower than that of the enzyme of membranes of normal individuals. Both the basal and calmodulin-stimulated activities of the membranebound enzyme in normal and K W A subjects were inhibited by oxygen free radicals. While the membrane-bound enzyme in normal subjects showed about 40% inhibition in the presence and absence of calmodulin, the membrane-bound enzyme in K W A subjects showed higher susceptibility to oxygen free radicals ( > 6 0 % inhibition) than the membrane-bound enzyme of erythrocytes in normal children even in the presence of calmodulin.
DISCUSSION
The current implication of oxygen free radicals in the pathophysiology of kwashiokor by Golden (1985) (14) suggests that many tissues in kwashiokor individuals could show features of oxidant stress. Indeed, Golden and his team have provided evidence to suggest (i) reduced activity of the protective mechanisms against oxygen free radicals including organic and enzymatic antioxidants and (ii) increase availability of iron particularly in liver samples of children dying from kwashiokor. In this study, we have investigated (i) the possibility that erythrocytes in kwashiokor individuals are under oxidative stress and (ii) the implications of this in relation to red cell CaZ+-ion homeostasis. Although Golden and Ramdath (1987) (15) have given the impression that erythrocytes may not be representative of the more vulnerable cells in the body
Erythrocyte Calcium-PumpFunction
441
during kwashiokor, the significant levels of non-heme iron and membrane lipid peroxidation observed in erythrocyte ghost membranes obtained from erythrocytes in children suffering from kwashiokor in this study give an indication of the susceptibility of erythrocytes to damage, just like other vulnerable cells in the body. The significant level of endogenous lipid peroxidation observed in the erythrocyte membranes may be attributed to the availability of iron, in view of the well established fact that iron could catalyze the generation of toxic reactive oxygen species (25, 26, 27). Previous studies on children dying from kwashiokor revealed high hepatic iron concentration compared with the concentration of other trace metals (15). Increased availability of iron in kwashiokor children is further substantiated by the finding that the bone marrow of these children contain abundant iron (28). The latter could possibly be the source of the non-heme iron observed in the erythrocyte ghost membrane obtained from erythrocytes in kwashiokor. Moreover, in areas of the world in which kwashiokor is prevalent, there is a high incidence of malaria, viral and bacterial infections as a result of poor sanitation and unhygienic conditions. It is now well established that such diseases or infections elicit the stimulation of polymorphonuclear leukocytes whose activities results in the production of superoxide radicals (O~) (29). The superoxide radicals could be dismutated to H202 by superoxide dismutases and both molecules can interact to produce hydroxyl radicals in the presence of iron via the Fenton and Haber-Weiss reactions (30, 31). The demonstration by Golden and his colleagues of a decrease in all the protective pathways for free radicals thus ensures the availability of the toxic oxygen species for damage. It has been suggested that modification of the physico-chemical properties of biomembranes due to the action of toxic oxygen species and lipid peroxidation products could lead to severe disturbances of Ca2+-homeostasis (2) especially because peroxidation renders the cell membrane more permeable to ions. In erythrocytes, the major mechanism involved in the regulation of intracellular CaZ+-concentration is the calmodulin-activated Ca 2§ + MgZ+-ATPase (31). The finding that the loss of Ca 2§ + Mg2+-ATPase activity following hydrolysis of erythrocyte ghost membrane phospholipids by phospholipase A2 or purified phospholipase (E) (33, 34) could be restored by the addition of phosphatidyl serine and phosphatidyl inositol (35, 36) or the unsaturated fatty acids (33) suggests that acidic phospholipids with their intact unsaturated fatty acids are important for the normal function of the enzyme. Clemens and Einsele (1987) (37) have shown that, in red cells, predominantly fatty acids with four or more double bonds seem to be involved in membrane peroxidation reactions; such fatty acids constitute about 15% of total membrane fatty acids and are mostly associated with acidic phospholipids (38). It appears therefore, that in situations where there is significant production of free radicals and where the antioxidant defenses of the red blood cells are overwhelmed by these radicals, the subsequent membrane damage involves the impairment of Ca2+-pump function. Previous studies in this laboratory (39) have shown that the membrane-bound Ca 2§ pumping ATPase of erythrocytes in kwashiokor exhibit lower affinity for ATP and calmodulin and does not exhibit high CaZ+-affinity sites in the presence of calmodulin, an indication that a significant percentage of the enzyme in situ
442
Okunade and Olorunsogo
appears not to be in the right conformational state for Ca 2+ translocation and ATP hydrolysis. Since it is not clear whether the decrease in calmodulin and Ca2+-afiinity of the erythrocyte Ca2+-pump enzyme in KWA is due to the nature of the interaction of the pump protein with membrane phospholipids, attempt was made in this study to determine the level of endogenous lipid peroxidation products in the erythrocyte membranes and more importantly the susceptibility of the Ca2§ ATPase in the erythrocyte ghost membranes and intact erythrocytes from KWA children to exogenously generated free radicals. It seems probable that the significantly high level of lipid peroxidation products observed in EGM of KWA subjects (Fig. 2) could result from a destruction of acidic phospholipids of the membrane. Moreover, since these phospholipids are associated with, and activate the CaZ+-pump in situ, (40) their destruction could affect the functional status of the enzyme. Although the exact mechanism of modulation of Ca2§ function by oxygen free radicals has not been elucidated, it appears that modulation by cellular redox status involving the degree of thiol oxidation and lipid peroxidation may have some bearing, especially because the Ca2§ traverses the membrane eight times with interspersing phospholipid molecules associated with each bend (40, 41). Based on the structural model proposed by these workers, it seems likely that oxidative damage to the erythrocyte membrane will involve the peroxidation of the phospholipid molecules associated with the different transmembrane domains of the enzyme. This alteration in the structural integrity of the phospholipids in the immediate environment of the enzyme could affect the transition of the enzyme from one conformational state to the other, an event that involves a major refolding of the polypeptide chain.
CONCLUSION Nevertheless, an obvious implication of the results obtained in this study is that in situ, the red cell membrane calcium-pumping activity and hence Ca2+-homeostasis may be altered by the activities of endogenously generated reactive oxygen species. ACKNOWLEDGEMENT
We are especially grateful to Dr. O. Shodeinde, Department of Pediatrics, University of Ibadan, Ibadan, for ensuring supply of blood samples from kwashiokor subjects. REFERENCES 1. Dykens, J. A., Sullivan, S. G. and Stern, A. (1987) In: Free Radicals, Oxidant Stress and Drug Action (C. Rice-Evans ed.) pp. 347. 2. Cerrutti, P. (1985) Science 227:375.
Erythrocyte Calcium-Pump Function
443
3. Blake, D. R., Gallagher, P. J., Potter, A. R., Bell, M. J. and Bacon, P. A. (1984) Arthritis Rheum 27: 495. 4. Hebbel, R. P. and Eaton, J. W. (1989). Semin Haematol: 26:136. 5. Buffinton, G. D., Hunt, N. H., Cowden, W. B. and Clark, I. A. (1986). In: Free Radicals, Cell Damage and Disease (Rice-Evans C. ed.) p. 201. 6. Hess, M. L. and Mason, N. H. (1984). J. Mol. Cell Cardiol. 16:969. 7. Mccord, J. M. (1984). New Eng. J. Med. 312:159. 8. Slater, T. F. (1988). In: "Free Radicals, Methodology and Concepts" (Rice-Evans, C. and Halliwell, B. eds.) p. 17. 9. Williams, C. D. (1933). Arch. Dis. Childh. 8:423-433. 10. Srikantia, S. G. (1958). Lancet i:667. 11. Gopalan, G. (1968). In: Calorie deficiencies and Protein deficiency. (McChance, R. A. and Widdowson E. M. eds) London Churchill p. 49. 12. Whitehead, R. G. (1979). Zeitschriftfur Emah rungswissen schaft, 23 (Suppl.) 72. 13. Hendriekse, R. G. (1984). Trans. Roy, Soc. Trop. Med. Hyg. 78:427. 14. Golden, M. H. N. (1985). In: NutritionalAdaptation in Man (Blaxter, K. B. and Waterlow, J. C. eds) John Libbey, London. p. 169. 15. Golden, M. H. N. and Ramdath, D. (1987). Proceedings of the Nutrition Society: 46:53. 16. Niggli, V., Penniston, J. T. and Carafoli, E. (1979). J. Biol. Chem. 254, 9955. 17. Dodge, J. T., Mitchell, C. and Hanahan, D. J. (1963). Arch Biochem. Biophys. 100:119. 18. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951). J. Biol. Chem. 193:265. 19. Raess, B. U. and Vincenzi, F. F. (1980). J. Pharm. Methods 4:273-283. 20. Fiske, C. H. and Subbarow, Y. (1925) J. Biol. Chem. 66:375. 21. Ceriotti, F. and Cerriotti, G. (1980). Clin. Chem. 26:327. 22. Hebbel, R. P., Shalev, O., Foker, W. and Rank, B. H. (1986). Biochirn. Biophys. Acta 862:8. 23. Walls, R., Kurnar, K. S. and Hochstein, P. (1976). Arch. Biochem. Biophys. 100:119. 24. Rice-Evans C., Okunade, W. G. and Khan R. (1989). Free Rad Res. Comms. 7:45-54. 25. O'Brien, P. (1969). Can J. Biochem. 47:485. 26. Aust, S. D., Morehouse, L. A. and Thomas, C. E. (1985). J. Free Rad. Biol. Med. 1:3. 27. Rice-Evans, C. (1987). In: Free Radicals, Oxidant stress and Drug action. (Rice-Evans, C. W.). Richelieu Press, London p. 30. 28. Adams, E. B. and Scragg, J. N. (1965). Br. J. Haematol. 11:676. 29. Babior, B. M. (1981). J. Clin. Invest. 73:599. 30. Fenton, H. J. H. (1894). J. Chem. Soc. 899. 31. Haber, F. and Weiss, J. (1934) Proc. Roy-Soc. Lond. Series A147:332. 32. Schatzmann, H. J. (1966). Experentia 22:364. 33. Ronner, P., Gazzotti, P. and Carafoli, E. (1977) Arch. Biochem. Biophy. 179:578. 34. Roelofsen, B. and Schatzmann, H. J. (1977) Biochim. Biophys. Acta. 464:17. 35. Niggli, V., Adunyah, E. S. and Carafoli, E. (1981) J. Biol. Chem. 256:395. 36. AI-Jobore, A. and Roufogalis, B. D. (1981) Can. J. Biochem. 59:880. 37. Clemens, M. R. and Einsele, H. (1987). In: Free Radicals, Cell Damage and Disease. (Rice-Evans C. ed.) Richelieu Press London p. 127. 38. Ways, P. and Hanahan, D. J. (1963). J. Lipid Res. 5:318. 39. Olorunsogo, O. O. (1989). Biosci. Rep. 9:359-368. 40. Carafoli, E. (1991). PhysioL Rev. 71: 129. 41. Verma, A. K., Filoteo, A. G., Stanford, D. R., Wieben, E. D., Penniston, J. T., Strehler, E. E., Fischer, R., Heim, R., Vogel, G., Mathews, S., Strehler-Page, M. A., James, P., Vorherr, T., Krebs, J. and Carafoli, E. (1988). J. Biol. Chem. 263:14152.
Effect of Reactive Oxygen Species on the Erythrocyte Calcium-Pump Function in Protein-Energy Malnutrition W. G. Okunade L2 and O. O. Olorunsogo I Received June 6, 1992 The presence of detectable amounts of non-heme iron in erythrocyte ghost membranes have been postulated to lead to the initiation of membrane lipid peroxidation and the attendant perturbation of membrane functions. We have investigated the presence of non-heme iron and endogenous products of lipid peroxidation in erythrocyte membranes of normal and kwashiorkor (KWA) subjects and asse,ssed the susceptibility of the membranes to exogenously generated reactive oxygen species. The modulation of the basal and calmodulin-stimulated calcium-pumping activity of these membranes by reactive oxygen species was also assessed. The results show the presence of significant amounts of non-heme iron and endogenous free radical reaction products in the red cell membranes of KWA subjects compared with that of normal children. Estimation of the extent of lipid peroxidation in the presence of exogenously generated reactive oxygen species further revealed that erythrocyte ghost membranes of KWA subjects are more susceptible to oxidative stress than those of normal individuals. Although both the basal and calmodulin-stimulated activities of the membrane-bound Ca2+-pump enzyme in normal and KWA subjects were inhibited by oxygen-free radicals, the erythrocyte enzyme in KWA subjects showed higher susceptibility to inhibition by oxygen free radicals than that of normal individuals. We propose that the reduced erythrocyte calcium-pump function in K W A is not unconnected with excessive generation of reactive oxygen species. KEY WORDS: Reactive oxygen species; non-heme iron; CA2+-ATPase; erythrocyte; kwashiokor. ABBREVIATIONS: PMSF; phenylmethylsulfonylfuloride; TLCK; N-0~-p-tosyl-L-lysine chloromethylketone; EGTA ethyleneglycol-bis (B-aminoethylether) N,N'-tetraacetic acid; EDTA; ethylene diamine tetraacetic acid; ATP; Adenosine 5-triphosphate; Hepes; 4-(2-hydroxyethyl)-l-piperazine ethanesulphonic acid; Tris-HCl, Tris (hydroxymethyl) aminomethane-hydrochloride; SDS; Sodium dodecyl sulphate; TBAR; thiobarbituric acid-reactive products TBA, thiobarbituric acid; TCA; trichloroacetic acid.
INTRODUCTION C o n s i d e r a b l e a t t e n t i o n is p r e s e n t l y b e i n g f o c u s s e d o n t h e i n f l u e n c e o f o x y g e n d e r i v e d f r e e r a d i c a l s o n t h e i n t e g r i t y a n d f u n c t i o n s o f cell m e m b r a n e s in d i s e a s e Laboratory for Biomembrane Research, Department of Biochemistry, University of Ibadan, Ibadan, Nigeria. a To whom correspondence should be addressed. 433
0144-8463/92/1200-0433506.50/0~ 1992PlenumPublishingCorporation
434
Okunade and Olorunsogo
states in view of the cytotoxic effects of reactive oxygen species on normal cells (1). In this connection, activated oxygen species have been implicated in certain pathological conditions such as cancer (2), rheumatoid arthritis (3) haemoglobinopathies (4), parasitic infection (5) and a variety of ischaemia (6, 7). It is now well established that when reactive oxygen species evade or overwhelm the protective mechanisms of cells and tissues, they elicit metabolic and cellular disturbances as a result of peroxidation of membrane lipid (8). Although severe dietary protein deficiency in the presence of relative carbohydrate excess is generally accepted as the cause of kwashiokor (KWA) (9), a well-defined clinical syndrome at one end of protein-energy malnutrition (PEM) spectrum, the various clinical and biochemical features associated with the disease have made several investigators to put forward a number of hypotheses in order to explain the biochemical basis of the condition (10, 11, 12, 13, 14). Indeed, the reports of Gopalan (1968) (11) and later Golden (1985) (14) that there are no essential differences in the protein: energy ratios of the diets of children who developed marasmus and those with kwashiokor made the concept of lack of protein in the diet to fall out of favour. Based on the demonstration in PEM of an increased flux of products of free radical action, a decrease in all the protective pathways for free radicals and an increase in tissue iron (14) Golden and Ramdath (1987) (15) suggested that kwashiokor probably results from an imbalance between the production of free radicals and the protective mechanisms that remove them in the malnourished state. Since one of the clinical features in kwashiokor is anaemia, we have investigated in this study, erythrocytes from kwashiokor children because of erythrocyte membrane susceptibility to reactive oxygen species and the effect of these species on cell membrane calcium pumping.
MATERIALS AND METHODS Thiobarbituric acid, TLCK, PMSF, Ferrozine, Soybean trypsin inhibitor, disodium salt of ATP, and thiourea were purchased from Sigma Chem. Co., London. Calmodulin was purchased from Calbiochemical, La Jolla, California. All other reagents used were of analytical grade and were purchased from British Drug Houses, Poole, England.
Preparation of Erythrocyte Ghost Membrane for Ca2+-ATPase Studies Haemoglobin-free ghost membranes deficient in calmodulin were prepared essentially according to the procedure of Niggli et al. (1981) (16) which is based on the principle of hypotonic lysis developed by Dodge et al. (1963) (17). All steps of the membrane preparation were carried out at 4~ Blood samples were obtained from children who were newly identified at the University College, Oni Memorial, and Adeoyo State Hospitals, Ibadan, as having severe kwashiokor. As
Erythrocyte Calcium-Pump Function
435
a rule, none of the subjects received any dietary therapy or medication prior to the: time blood was collected. Normal human blood was collected from healthy children in the same age bracket and was used as control. All blood samples were collected in acid-citrate-dextrose buffer, stored at 4~ and used within 24 hours. Whole blood samples were centrifuged at 5800 rpm for 10 mins, the plasma and "buffy" layers were removed by aspiration to obtain packed erythrocytes. The erythrocytes were washed thrice in 10 volumes of 130 mM KC1, 10 mM Tris, phi 7.4 to remove plasma constituents. Each time, the cell suspension was centrifuged at 5800 rpm and the supernatant removed by aspiration. Washed erythrocytes were lysed in 10 volumes of 1 mM Na-EDTA, 10 mM Tris pH 7.4 containing 1 #M soybean trypsin inhibitor, 10/~M TLCK and 200/~M PMSF and centrifuged at 18,000 rpm for 20 rains. Membranes were washed four times in the hemolysis solution and eight times in 10 mM HEPES pH 7.4 containing 200 gM PMSF in order to remove residual calmodulin, haemoglobin and EDTA. The haemoglobin-free ghosts were finally resuspended and centrifuged at 18,000 rpm for 20 mins in a 20 mM HEPES, pH 7.4 storage buffer containing 130 mM KCI, 500/~M MgCI2 and 50/~M CaCI2. The membrane pellet was suspended and stored in the same buffer at -60~
Preparation of Erythrocyte Ghost Membranes for Studies on Membrane Peroxidation and Non-heme Iron Determination Haemoglobin-free ghost membranes were prepared by a slight modification of the above procedure. Blood samples were centrifuged at 2500 rpm for 10 mins followed by removal of the plasma and buffy layers by aspiration. The packed cells were suspended in 10 volumes of cold isotonic phosphate buffer pH 7.4 and centrifuged at same speed for 10 minutes. This was repeated once and any remaining buffy coat was carefully aspirated. The packed cells were then resuspended in 10 volumes of cold hypotonic 5raM phosphate buffer pH 7.4 and allowed to stand on ice for 30 mins. Centrifugation was later carried out at 16,000rpm for 10 mins at 4~ The supernatant as well as the dense pellet at the bottom of the tubes were carefully aspirated. The membranes obtained were washed four more times in 10 vols. of the same buffer and later resealed by resuspension in 10 volumes of the isotonic phosphate buffer pH 7.4 followed by centrifugation at 16,000 rpm for 10 mins. The membranes were stored in the same buffer at 4~ and used within 24 hours.
Protein Determination Membrane concentration was assessed .by determining protein concentration according to the method of Lowry et aL (i95t) (I8) using bovine serum albumin as standard.
436
Okunade and Olorunsogo
Incubation of Erythrocyte Ghost Membranes with Oxygen Free Radicals. Calmodulin-deficient, haemoglobin-free erythrocyte ghost membranes were pelleted by centrifugation at 20,000 rpm for 30 mins and resuspended in isotonic phosphate buffer pH 7.4. i mg/ml final concentration of the membranes were incubated for 30 mins in phosphate-buffered saline pH 7.4 containing premixed ferrous sulphate with ATP to form an ATP-Fe z+ complex at a final concentration of 1 mM ATP, 0.1 mM FeSO4. Hydrogen peroxide was then added to a final concentration of 5 mM to initiate the free radical reaction. Control incubation was also carried out by incubating erythrocyte ghost membrane in the phosphatebuffered saline pH 7.4 without any of the additives. The reaction was stopped by diluting the incubation mixture with 5 vols. of ice-cold isotonic Tris buffer (130 mM KC1 + 20 mM Tris pH 7.4) containing 5 mM Mannitol. The ghost membranes were wash-centrifuged twice with the same buffer but without 5 mM-Mannitol and once with 130mM KC1, 20mM K § HEPES pH 7.4, 50 #M CaClz and 500 ~tM MgCI2. The ghost membranes were finally suspended in the latter buffer and used to estimate Ca2+-pumping ATPase activity.
Assay of Ca2+-ATPase Activity Ca2+-ATPase activity was assayed by measuring the rate of release of inorganic phosphate from the -position of ATP. The phosphate liberated to the medium was determined colori-metrically using the method of Raess and Vincenzi (1980) (19) based on the procedure of Fiske and Subbarow (1925) (20). Calmodulin-depleted membranes (100-200 #g) were incubated at 37~ for five minutes in a final total volume of 1 ml. The assay medium contained 100 mM KCI, 50 mM K-HEPES pH 7.4, 2 mM MgCI2, 20 ~M CaC12, 2 mM EGTA (when added), and 2 ~g/ml calmodulin (when added). The reaction was started by adding 2 mM ATP (final concentration). After 30 minutes of incubation at 37~ with constant shaking, the reaction was stopped with 1 ml of 10% solution of sodium dodecyl sulphate (SDS) in distilled water. All assays were run in duplicate with and without calmodulin.
Incubation of Erythrocytes with Oxygen Free Radicals Exposure of red cells to activated oxygen species was carried out essentially according to the procedure described by Hebbel et al. (1986) (22). Packed erythrocytes were washed three times with 150 mM NaCI solution and suspended to 20% haematocrit in Hank's balanced salt solution containing 5.6 mM glucose, 136.9 mM NaCI, 5.4mM KCI, 0.3 mM NazHPO4, 0.3 mM KH PO4, 0.4 mM MgSO4, 13 mM CaC12, 0.5 mM MgCI: and 4.2 mM NaHCO3 (pH 7.4). They were incubated in a shaking water bath at 37~ for 40 rains with or without premixed ferrous sulphate with ATP to form an ATP-Fe 2+ complex at a final concentration of l mM ATP, 0.1mM FeSO#, and 5raM hydrogen peroxide to initiate the reaction. After 40 minutes, 5 vols of ice-cold 150 mM NaCI was added to the
Erythrocyte Calcium-Pump Function
437
reaction mixture followed by centrifugation at 5800 rpm for 10 mins. The packed red cells were suspended to 10% Hct in 175 mM Tris-HCl pH 7.4 and admixed wffh saponin at a ratio of 10 ~g saponin per 10 ~1 erythrocytes. After 15 mins incubation at room temperature, any unlysed red cells were removed by low-speed centrifugation. 0.1 ml red cells lysate was whirlmixed with 0.2rnl of 175 mM Tris-HC1 pH 7.4 and incubated for 30 mins at 37~ in a final volume of 1 ml containing (in final concentrations) 20 mM imidazole, 60 mM KC1, 10 mM NaC1, 3 mM MgCI2, 0.2mM Ouabain, 20 laM CaCI2 (or 200/zM EGTA) and 4 mM ATP. The reaction which was run in duplicate was stopped by addition of 0.1 ml 50% trichloroacetic acid and centrifuged to obtain a clear supernatant. The inorganic phosphate in the supernatant was determined by adding 1 ml of 1.25% ammonium molybdate in 6.5% H2804, followed by 1 ml of 9% ascorbic acid. The absorbance of the blue colour developed was measured at 660 nm. CaZ+-ATPase activity was calculated by substracting the Pi content of the RGTA-containing sample (i.e. Mg2+-ATPase activity) from that of the calcium containing sample (i~ Ca 2+ + M f § activities). Determination of Membrane-bound non-haeme Iron Membrane-bound non-heme iron was determined according to a modified method of Ceriotti and Ceriotti (1980). (21) Final concentration of reagents used were 38 mM ascorbate, 0.67% SDS, 10 mM phosphate buffer pH 7.4, 150 mM NaC1 and 1.28 mM ferrozine in a total volume of 3.0 ml. The intensity of the pink-coloured complex formed was determined by measuring absorbance at 562 nm using distilled water as blank. Standards were simultaneously run with a stock solution of 25 mg/L iron. Assessment of Lipid Peroxidation Lipid peroxidation was assayed by measuring the thiobarbituric acid-reactive (TBAR) products using the procedure of Walls et al. (1976) (23) as previously described (24). 1 mg/ml final concentration of membranes in isotonic phosphate buffer pH 7.4, were incubated for six hours at 37~ in a shaking water bath with oi without 0.1 mM FeSOa, 1 mM EDTA, 1 mM ascorbate and 0.2 mM H202 (final concentrations). 0.5 ml of 0.75% TBA in 0.1 M HCI was added to 0.5 ml of the incubation mixture already quenched with 0.5 ml of 10% TCA. The mixture was heated at 90-95~ for 20 mins and after cooling, centrifuged for 10 mins at 2,500r.p.m. The supernatant was transferred into acid resistant tubes and centrifuged at 16,000 rpm for 10 mins. The absorbance of the resulting clear solution was determined at 532 nm with phosphate buffer as blank. RESULTS Estimation of Membrane-bound Non-heme Iron and Susceptibifity to Lipid Peroxidation The procedure used in this study to determine non-heme iron is based on the reduction of iron by ascorbic acid (21). Addition of ferrozine to this mixture
438
Okunade and Olorunsogo 10"
cO7
e-
o.c_ n=6 ~
0
g ~
E
e"
a~
0 I N
! KWA
Fig. 1. Erythrocyte ghost membrane-bound nonheine Iron in healthy (N) and kwashiokor (KWA) children. Non-heme Iron was estimated by Ferrozine assay as described in Materials and Methods and expressed as nmotesnon-hemeIron per mg protein.
results in the formation of a complex with iron that is intensely coloured. SDS was included in the assay to ensure solubilisation and denaturation of membrane proteins. Although the absorbance of the pink chromophore was read at 6 hrs and 18 hrs after addition of ferrozine, the results (Fig. 1) obtained did not differ markedly thus indicating that all the non-heme iron have been released into solution. The results as presented in Fig. 1 indicate the absence of non-heme iron in haemoglobin-free erythrocyte ghost membranes (EGM) prepared from red cells of healthy children. In contrast, non-heme iron is present in significant amounts in ghost membranes prepared from red cells obtained from kwashiokor children even though there was considerable variation in the amounts observed among the ghost membranes of these children. The extent of endogenous lipid peroxidation was determined in these membranes because of the significant amounts of non-heme iron associated with the ghost membranes. The results show that the extent of lipid peroxidation in the EGM of KWA subjects was about 3-fold that determined for red cell membranes of normal individuals (Fig. 2). Furthermore, exposure of the membranes to exogenously generated reactive oxygen species results in a significant increase in the amounts of lipid peroxidation products in both membranes. A 3-fold increase in the amount of Thiobar-bituric acid-reactive (TBAR) products was observed for EGM membranes of normal individuals. This is about half of the value obtained for similarly treated red cell membranes of KWA subjects and about 80% of the value obtained for untreated EGM of KWA subjects. Moreover, exogenously generated reactive oxygen species produced an almost 2-fold increase over the basal value of Thiobarbituric acid-reactive (TBAR) products in EGM of KWA subjects. These results reflect a state of oxidative
Erythrocyte Calcium-Pump Function
439
3.0n=lO
I c
m
Q_ iIII
E
n=lO
r~
n=10
r
l-IJ1
1.0-
0
E
I
n=10
H I
N
I
KWA
I
N§
I
KV~I&§ ROS
Fig. 2. Degree of lipid peroxidation of erythrocyte ghost membrane of healthy (N) and kwashiokor (KWA) children in the presence and absence of exogenously generated reactive oxygen species (ROS). Lipid peroxidation was estimated as described in Material and Methods and expressed as nmoles Thiobarbituric reactive (TBAR) products per mg protein.
stress in the KWA subjects and also provide an indication that red cells in KWA individuals are highly susceptible to peroxidation by activated oxygen species.
Ca2+-ATPase Studies The CaZ+-ATPase assay used in this study for intact red cells utilized the red cells own cytosol as a source of calmodulin so that the activity measured is the calmodulin-stimulated CA2*-ATPase activity. Table 1 shows that the activity of the pump in red cells of KWA subjects is about 57% lower than that of the enzyme in red cells of normal individuals. Exposure of the washed erythrocytes fiom KWA and normal subjects to reactive oxygen species results in a significant inhibition of the CaM-activated CaZ+-pumping ATPase activity even though the degree of inhibition differs. While the CaZ+-ATPase activity of red cells from Table 1. Effect of Exogenously-generated free radical on the activity of CaZ+-ATPase in intact Red cells from normal and kwashiokor individuals Erythrocyte Samples from:
CaZ+-ATPase activity umole Pi release glib -1 hr -1
% of Control
82-56 + 6.2 49.40 -t- 9.3 35.60 5:4,7
59.8
11.20 + 3.8
31.5
Normal Normal + Free Radicals Kwashiokor Kwashiokor + Free Radicals Each value is a mean of four experiments.
Okunade and Olorunsogo
440
Table 2. Effect of Exogenously-generatedfree radicals on the cae+-ATPase activity
of Erythrocyte ghost membranes in kwashiokor and Normal Subjects. Ghost Membranes Samples from erythrocytes of: Normal Normal + Free Radicals Kwashiokor Kwashiokor + Free Radicals
Mga+-ATPase
Caz+ + MgZ+-ATPase activity umole Pimg protein -1 hr -1 - CaM + CaM
0.36 5:0.02 0.28 5:0.03 0.18 5:0.03
0.70 5:0.05 0.43 5:0.07 0.38 + 0.04
2.89 5:0.17 1.71 5:0.08 0.88 :t: 0.06
0.11 + 0.04
0.15 5:0.03
0.31 5:0.04
Each value is a mean of four experiments. normal individuals was reduced to about 60% of control, (40% inhibition) the cae§ activity of erythrocytes from K W A subjects was inhibited to about 30% of their controls (70% inhibition). The results obtained with erythrocyte ghost membranes followed the same trend as that observed with intact erythrocytes (Table 2). The Ca2§ ATPase activity of erythrocyte plasma m e m b r a n e of K W A subjects show, in the absence of calmodulin, a lower specific activity (46%) than the enzyme of the erythrocyte membrane of normal individuals. Although the addition of calmodulin increased the specific activity of the calcium pump of erythrocyte ghost membrane of K W A and normal individuals by at least 2-fold and 4-fold, respectively, the stimulation of the pump protein in membranes from K W A individuals is about 70% lower than that of the enzyme of membranes of normal individuals. Both the basal and calmodulin-stimulated activities of the membranebound enzyme in normal and K W A subjects were inhibited by oxygen free radicals. While the membrane-bound enzyme in normal subjects showed about 40% inhibition in the presence and absence of calmodulin, the membrane-bound enzyme in K W A subjects showed higher susceptibility to oxygen free radicals ( > 6 0 % inhibition) than the membrane-bound enzyme of erythrocytes in normal children even in the presence of calmodulin.
DISCUSSION
The current implication of oxygen free radicals in the pathophysiology of kwashiokor by Golden (1985) (14) suggests that many tissues in kwashiokor individuals could show features of oxidant stress. Indeed, Golden and his team have provided evidence to suggest (i) reduced activity of the protective mechanisms against oxygen free radicals including organic and enzymatic antioxidants and (ii) increase availability of iron particularly in liver samples of children dying from kwashiokor. In this study, we have investigated (i) the possibility that erythrocytes in kwashiokor individuals are under oxidative stress and (ii) the implications of this in relation to red cell CaZ+-ion homeostasis. Although Golden and Ramdath (1987) (15) have given the impression that erythrocytes may not be representative of the more vulnerable cells in the body
Erythrocyte Calcium-PumpFunction
441
during kwashiokor, the significant levels of non-heme iron and membrane lipid peroxidation observed in erythrocyte ghost membranes obtained from erythrocytes in children suffering from kwashiokor in this study give an indication of the susceptibility of erythrocytes to damage, just like other vulnerable cells in the body. The significant level of endogenous lipid peroxidation observed in the erythrocyte membranes may be attributed to the availability of iron, in view of the well established fact that iron could catalyze the generation of toxic reactive oxygen species (25, 26, 27). Previous studies on children dying from kwashiokor revealed high hepatic iron concentration compared with the concentration of other trace metals (15). Increased availability of iron in kwashiokor children is further substantiated by the finding that the bone marrow of these children contain abundant iron (28). The latter could possibly be the source of the non-heme iron observed in the erythrocyte ghost membrane obtained from erythrocytes in kwashiokor. Moreover, in areas of the world in which kwashiokor is prevalent, there is a high incidence of malaria, viral and bacterial infections as a result of poor sanitation and unhygienic conditions. It is now well established that such diseases or infections elicit the stimulation of polymorphonuclear leukocytes whose activities results in the production of superoxide radicals (O~) (29). The superoxide radicals could be dismutated to H202 by superoxide dismutases and both molecules can interact to produce hydroxyl radicals in the presence of iron via the Fenton and Haber-Weiss reactions (30, 31). The demonstration by Golden and his colleagues of a decrease in all the protective pathways for free radicals thus ensures the availability of the toxic oxygen species for damage. It has been suggested that modification of the physico-chemical properties of biomembranes due to the action of toxic oxygen species and lipid peroxidation products could lead to severe disturbances of Ca2+-homeostasis (2) especially because peroxidation renders the cell membrane more permeable to ions. In erythrocytes, the major mechanism involved in the regulation of intracellular CaZ+-concentration is the calmodulin-activated Ca 2§ + MgZ+-ATPase (31). The finding that the loss of Ca 2§ + Mg2+-ATPase activity following hydrolysis of erythrocyte ghost membrane phospholipids by phospholipase A2 or purified phospholipase (E) (33, 34) could be restored by the addition of phosphatidyl serine and phosphatidyl inositol (35, 36) or the unsaturated fatty acids (33) suggests that acidic phospholipids with their intact unsaturated fatty acids are important for the normal function of the enzyme. Clemens and Einsele (1987) (37) have shown that, in red cells, predominantly fatty acids with four or more double bonds seem to be involved in membrane peroxidation reactions; such fatty acids constitute about 15% of total membrane fatty acids and are mostly associated with acidic phospholipids (38). It appears therefore, that in situations where there is significant production of free radicals and where the antioxidant defenses of the red blood cells are overwhelmed by these radicals, the subsequent membrane damage involves the impairment of Ca2+-pump function. Previous studies in this laboratory (39) have shown that the membrane-bound Ca 2§ pumping ATPase of erythrocytes in kwashiokor exhibit lower affinity for ATP and calmodulin and does not exhibit high CaZ+-affinity sites in the presence of calmodulin, an indication that a significant percentage of the enzyme in situ
442
Okunade and Olorunsogo
appears not to be in the right conformational state for Ca 2+ translocation and ATP hydrolysis. Since it is not clear whether the decrease in calmodulin and Ca2+-afiinity of the erythrocyte Ca2+-pump enzyme in KWA is due to the nature of the interaction of the pump protein with membrane phospholipids, attempt was made in this study to determine the level of endogenous lipid peroxidation products in the erythrocyte membranes and more importantly the susceptibility of the Ca2§ ATPase in the erythrocyte ghost membranes and intact erythrocytes from KWA children to exogenously generated free radicals. It seems probable that the significantly high level of lipid peroxidation products observed in EGM of KWA subjects (Fig. 2) could result from a destruction of acidic phospholipids of the membrane. Moreover, since these phospholipids are associated with, and activate the CaZ+-pump in situ, (40) their destruction could affect the functional status of the enzyme. Although the exact mechanism of modulation of Ca2§ function by oxygen free radicals has not been elucidated, it appears that modulation by cellular redox status involving the degree of thiol oxidation and lipid peroxidation may have some bearing, especially because the Ca2§ traverses the membrane eight times with interspersing phospholipid molecules associated with each bend (40, 41). Based on the structural model proposed by these workers, it seems likely that oxidative damage to the erythrocyte membrane will involve the peroxidation of the phospholipid molecules associated with the different transmembrane domains of the enzyme. This alteration in the structural integrity of the phospholipids in the immediate environment of the enzyme could affect the transition of the enzyme from one conformational state to the other, an event that involves a major refolding of the polypeptide chain.
CONCLUSION Nevertheless, an obvious implication of the results obtained in this study is that in situ, the red cell membrane calcium-pumping activity and hence Ca2+-homeostasis may be altered by the activities of endogenously generated reactive oxygen species. ACKNOWLEDGEMENT
We are especially grateful to Dr. O. Shodeinde, Department of Pediatrics, University of Ibadan, Ibadan, for ensuring supply of blood samples from kwashiokor subjects. REFERENCES 1. Dykens, J. A., Sullivan, S. G. and Stern, A. (1987) In: Free Radicals, Oxidant Stress and Drug Action (C. Rice-Evans ed.) pp. 347. 2. Cerrutti, P. (1985) Science 227:375.
Erythrocyte Calcium-Pump Function
443
3. Blake, D. R., Gallagher, P. J., Potter, A. R., Bell, M. J. and Bacon, P. A. (1984) Arthritis Rheum 27: 495. 4. Hebbel, R. P. and Eaton, J. W. (1989). Semin Haematol: 26:136. 5. Buffinton, G. D., Hunt, N. H., Cowden, W. B. and Clark, I. A. (1986). In: Free Radicals, Cell Damage and Disease (Rice-Evans C. ed.) p. 201. 6. Hess, M. L. and Mason, N. H. (1984). J. Mol. Cell Cardiol. 16:969. 7. Mccord, J. M. (1984). New Eng. J. Med. 312:159. 8. Slater, T. F. (1988). In: "Free Radicals, Methodology and Concepts" (Rice-Evans, C. and Halliwell, B. eds.) p. 17. 9. Williams, C. D. (1933). Arch. Dis. Childh. 8:423-433. 10. Srikantia, S. G. (1958). Lancet i:667. 11. Gopalan, G. (1968). In: Calorie deficiencies and Protein deficiency. (McChance, R. A. and Widdowson E. M. eds) London Churchill p. 49. 12. Whitehead, R. G. (1979). Zeitschriftfur Emah rungswissen schaft, 23 (Suppl.) 72. 13. Hendriekse, R. G. (1984). Trans. Roy, Soc. Trop. Med. Hyg. 78:427. 14. Golden, M. H. N. (1985). In: NutritionalAdaptation in Man (Blaxter, K. B. and Waterlow, J. C. eds) John Libbey, London. p. 169. 15. Golden, M. H. N. and Ramdath, D. (1987). Proceedings of the Nutrition Society: 46:53. 16. Niggli, V., Penniston, J. T. and Carafoli, E. (1979). J. Biol. Chem. 254, 9955. 17. Dodge, J. T., Mitchell, C. and Hanahan, D. J. (1963). Arch Biochem. Biophys. 100:119. 18. Lowry, O. H., Rosebrough, N. J., Farr, A. L. and Randall, R. J. (1951). J. Biol. Chem. 193:265. 19. Raess, B. U. and Vincenzi, F. F. (1980). J. Pharm. Methods 4:273-283. 20. Fiske, C. H. and Subbarow, Y. (1925) J. Biol. Chem. 66:375. 21. Ceriotti, F. and Cerriotti, G. (1980). Clin. Chem. 26:327. 22. Hebbel, R. P., Shalev, O., Foker, W. and Rank, B. H. (1986). Biochirn. Biophys. Acta 862:8. 23. Walls, R., Kurnar, K. S. and Hochstein, P. (1976). Arch. Biochem. Biophys. 100:119. 24. Rice-Evans C., Okunade, W. G. and Khan R. (1989). Free Rad Res. Comms. 7:45-54. 25. O'Brien, P. (1969). Can J. Biochem. 47:485. 26. Aust, S. D., Morehouse, L. A. and Thomas, C. E. (1985). J. Free Rad. Biol. Med. 1:3. 27. Rice-Evans, C. (1987). In: Free Radicals, Oxidant stress and Drug action. (Rice-Evans, C. W.). Richelieu Press, London p. 30. 28. Adams, E. B. and Scragg, J. N. (1965). Br. J. Haematol. 11:676. 29. Babior, B. M. (1981). J. Clin. Invest. 73:599. 30. Fenton, H. J. H. (1894). J. Chem. Soc. 899. 31. Haber, F. and Weiss, J. (1934) Proc. Roy-Soc. Lond. Series A147:332. 32. Schatzmann, H. J. (1966). Experentia 22:364. 33. Ronner, P., Gazzotti, P. and Carafoli, E. (1977) Arch. Biochem. Biophy. 179:578. 34. Roelofsen, B. and Schatzmann, H. J. (1977) Biochim. Biophys. Acta. 464:17. 35. Niggli, V., Adunyah, E. S. and Carafoli, E. (1981) J. Biol. Chem. 256:395. 36. AI-Jobore, A. and Roufogalis, B. D. (1981) Can. J. Biochem. 59:880. 37. Clemens, M. R. and Einsele, H. (1987). In: Free Radicals, Cell Damage and Disease. (Rice-Evans C. ed.) Richelieu Press London p. 127. 38. Ways, P. and Hanahan, D. J. (1963). J. Lipid Res. 5:318. 39. Olorunsogo, O. O. (1989). Biosci. Rep. 9:359-368. 40. Carafoli, E. (1991). PhysioL Rev. 71: 129. 41. Verma, A. K., Filoteo, A. G., Stanford, D. R., Wieben, E. D., Penniston, J. T., Strehler, E. E., Fischer, R., Heim, R., Vogel, G., Mathews, S., Strehler-Page, M. A., James, P., Vorherr, T., Krebs, J. and Carafoli, E. (1988). J. Biol. Chem. 263:14152.
