- Email: [email protected]
Effect of Aromatic Nitroxides on Hemolysis of Human Erythrocytes Entrapped With Isolated Hemoglobin Chains
Free Radical Biology & Medicine, Vol. 23, No. 2, pp. 278–284, 1997 Copyright q 1997 Elsevier Science Inc. Printed in the USA. All rights reserved 0891-5849/97 $17.00 / .00
PII S0891-5849(96)00611-9
Original Contribution EFFECT OF AROMATIC NITROXIDES ON HEMOLYSIS OF HUMAN ERYTHROCYTES ENTRAPPED WITH ISOLATED HEMOGLOBIN CHAINS ROSITA GABBIANELLI,* GIANCARLO FALCIONI,* ANNA MARIA SANTRONI,* GIANCARLO CAULINI,* LUCEDIO GRECI, † and ELISABETTA DAMIANI † *Dipartimento di Biologia Molecolare Cellulare Animale, Universita`, Via Camerini, 62032 Camerino, Italy; † Dipartimento di Scienze dei Materiali e della Terra, Universita`, Via Brecce Bianche, I-60131 Ancona, Italy (Received 5 January 1996; Accepted 5 December 1996)
Abstract—An in vitro model of thalassemia was produced by entrapment of isolated hemoglobin chains in human erythrocytes, thus subjecting the loaded cells to oxidative stress. The presence of these unpaired chains induced physico-chemical modifications at the membrane level as studied by laurdan fluorescence. The polarity of the lipid bilayer was shown to decrease with a concomitant shift towards a gel phase in a-loaded erythrocytes. The determination of conjugated dienes before the hemolytic event was used as an oxidation index; the results obtained demonstrate that b thalassemia is associated with oxidative stress. Furthermore, the presence of indolinic and quinolinic nitroxide radicals, a new class of antioxidants, in suspensions of a-loaded erythrocytes protected the erythrocytes from the hemolytic event. However, the protective effect exerted by the nitroxide radicals is not related to effects on membrane polarity and lipid peroxidation, even though a chemiluminescence study has demonstrated the superoxide scavenging activity of these nitroxide radicals. q 1997 Elsevier Science Inc. Keywords—RBCs, Hemoglobin-chains entrapment, Oxidative stress, Fluorescence, Hemolysis, Nitroxides, Chemiluminescence
cies (ROS) is altered due to autoxidation of hemoglobin and the release of the superoxide radical.5 Moreover, the RBCs antioxidative system can be impaired; in fact, we have recently demonstrated inactivation of glutathione peroxidase.6 A simple in vitro method of reproducing thalassemia is to entrap isolated a or b hemoglobin chains in human erythrocytes.4,6 – 8 This model represents a valid system for studying the cell alterations determined by the presence of free a or b chains, because erythrocytes from thalassemic patients have already undergone in vivo damage. The aim of this work was to evaluate in human a or b chains encapsulated in erythrocytes the physicochemical modifications at the membrane level. For this study, laurdan [2-dimethylamino-(6-lauroyl)-naphtalene] was selected to probe the changes in membrane polarity before the hemolytic event. This probe incorporates at the hydrophilic–hydrophobic interface of the membrane with the lauric acid tail anchored in the hydrophobic region of the bilayer. The probe displays a
INTRODUCTION
Thalassemia is related to an a or b chain imbalance in erythrocytes. These free hemoglobin chains are unstable and are readily oxidized. Their denaturation and binding to the erythrocyte membrane give rise to cytoskeleton modifications and lipid peroxidation, which are the cause of the many morphological, rheological, and biochemical defects of thalassemic red blood cells.1 – 4 In fact, several studies have shown that isolated a or b hemoglobin chains are less stable than tetrameric hemoglobin and that they rapidly autooxidise with the subsequent formation of precipitates known as Heinz bodies according to the following scheme: oxy-chains r met-chains r hemichromes r precipitates (Heinz bodies). During the process that leads to the formation of Heinz bodies, the production of reactive oxygen speAddress correspondence to: Prof. Lucedio Greci, Dipartimento di Scienze dei Materiali e della Terra, Universita`, Via Brecce Bianche, I-60131 Ancona, Italy. 278
/ 2b2b 2465 Mp 278 Wednesday May 28 10:41 PM EL–FRB 2465
Hb-chains entrapment and nitroxides
spectral sensitivity to the polarity of its surroundings, which is related to the physical state and dynamics of the surrounding phospholipid polar head group. Measurement of conjugated dienes was also performed to evaluate the extent of lipid peroxidation. Recently, the antioxidative activity of nitroxide radicals in various biological systems has been established.9 Therefore, indolinic and quinolinic nitroxide radicals, known to possess free radical scavenging activity, 10,11 were included in the study to determine their possible protective role in mitigating hemolysis in this in vitro model of thalassemia. In addition, by using the chemiluminescence technique, the superoxide scavenging activity of indolinic and quinolinic nitroxides was determined and compared to superoxide dismutase because nitroxide radicals are reported to possess SOD-mimic activity.12,13 MATERIALS AND METHODS
All reagents were of pure analytical grade. Laurdan was purchased from Molecular Probes (Eugene, OR). Xanthine, xanthine oxidase (E.C.1.1.3.22), superoxide dismutase (E.C.1.15.1.1) from bovine erythrocytes and lucigenin were obtained from Sigma Chemical Co., St. Louis, MO. Nitroxide radicals IAL (1,2-dihydro-2ethyl-2-phenyl-3H-indole-3-phenylimino-1-oxyl) and QAL (1,2-dihydro-2,2-diphenyl-4-ethoxy-quinoline-1oxyl) were prepared according to Berti et al. (Scheme 1).14,15 a and b chains were prepared from human hemoglobin using p-hydroxymercuribenzoate according to Bucci and Fronticelli.16 The regeneration of sulphydryl groups was carried out by treatment with mercaptoethanol on a chromatografic column as described previously.17 The purified chains containing regenerated sulphydryl groups, were encapsulated in human erythrocytes by a dialysis technique involving transient hypotonic hemolysis followed by isotonic resealing.6,18 The suspensions of RBCs in which a or b chains were entrapped were incubated in appropriate isotonic buffer at 377C and sampling was performed at the beginning (0 h) and after 7 h of incubation, before the hemolytic event. The degree of hemolysis was determined as (100 1 A/10 1 A*100%), where A is the optical density of Hb present in the supernatant after centrifugation of red cell suspension, and A*100% is the optical density of the red cell suspension after complete lysis with 10 vol distilled water at zero time incubation. To compare data obtained from cells in which a or b chains were encapsulated, two different controls were also prepared, one with resealed erythrocytes without chains (unloaded but opened and resealed) and the other one with unloaded erythrocytes (normal control).
279
Scheme 1. IAL: 1,2-dihydro-2-ethyl-2-phenyl-3H-indole-3-phenylimino-1-oxyl; QAL: 1,2-dihydro-2,2-diphenyl-4-ethoxy-quinoline-1oxyl.
For fluorescence experiments, erythrocyte membranes were prepared by hypotonic hemolysis.19 All samples were normalized by Lowry’s method.20 Steady-state fluorescence measurements on phospholipids extracted from erythrocyte membranes were performed on an Hitachi 4500 spectrofluorometer. Liposomes were prepared according to the method of Folch; 21 briefly, lipids were resuspended in CHCl3 / CH3OH 3:1 and mixed with the probe and dried in a gentle stream of N2 . The final probe:lipid molar ratio was 600:1. The phospholipid mixture was rehydrated with 10 mM Tris containing 154 mM NaCl, pH 7.4. Generalized polarization ( GP ) for laurdan, described by Parasassi et al., 22 was calculated by exciting laurdan at 340 nm ( GP340 ) according to the following equation: GP Å
IB0IR IB/ IR
(1)
where IB and IR are the emission intensities at the blue (440 nm) and red (490 nm) edges of the emission spectrum and correspond to the fluorescence emission maxima in the gel and liquid–crystalline phases, respectively, of the bilayer.23 Fluorescence measurements were performed at 377C. The absorbance ratio A233nm /A215nm , defined as the ‘‘oxidation index’’ was used as a relative measurement for conjugated dienes according to Konings.24 UV measurements were carried out using a Carry 1 Varian spectrophotometer at 257C. Because it has been demonstrated that a chains entrapped in erythrocytes give rise to an increase in hemolysis, 4,6 – 8 and having observed in this study that modifications are also present at the membrane level as studied by GP and conjugated diene, the effect of the nitroxides on these parameters was evaluated. The degree of hemolysis, the GP value, and the oxidation index of erythrocytes entrapped with a chains were determined on samples in which nitroxide radicals were
/ 2b2b 2465 Mp 279 Wednesday May 28 10:41 PM EL–FRB 2465
280
R. GABBIANELLI et al.
Fig. 1. Laurdan GP340 measured on lipids extracted from unloaded erythrocytes (normal control) ( h ); resealed erythrocytes without chains (control-resealed) ( j ); and erythrocytes loaded with a ( n ) or b ( L ) Hb chains as described in Materials and Methods. Data represent the means { SD. n Å 6.
included in the RBCs suspension medium. Experiments with nitroxides dissolved in ethanol were performed in Krebs buffer (143 mM NaCl, 5.7 mM KCl, 1.4 mM MgCl2 , 18 mM phosphate buffer, pH 7.4). In this case 50 mM (final concentration) of nitroxides were included in the erythrocyte suspensions and incubated at 377C. For all these experiments, control samples were prepared in the presence of the same volume of ethanol used in the suspensions containing nitroxides. Chemiluminescence measurements were performed using lucigenin as chemiluminogenic probe, and superoxide radical was produced by the xanthine/xanthine oxidase system.25 The chemiluminescence (CL) was measured in Autolumat LB 953 (Berthold Co. Wildbad, Germany) in a reaction mixture containing 0.9 U/ml xanthine oxidase, 150 mM lucigenin in physiological solution and different concentrations of nitroxides dissolved in ethanol. The reaction was started by injecting xanthine at a final concentration of 50 mM and followed for 60 s as described previously.26 The values are expressed as counts per second (cps). All the results presented are representative of at least three experiments, and appropriate controls were carried out throughout. The significance of values was obtained by using Student’s t-test. A p-value õ0.05 was considered statistically significant. RESULTS
The results on fluorescence experiments of Generalized Polarization for laurdan embedded in liposomes
consisting of lipids extracted from erythrocyte membranes are shown in Fig. 1. To evaluate the modifications in the lipid bilayer before the hemolytic event, the samples were analyzed both at 0 h and after 7 h of incubation. The 7-h time frame was established because up to this time the extent of hemolysis was irrelevant. The data obtained show a significant (p õ .0001) increase in GP340 values for samples in which a chains were encapsulated. No significant (p ú .05) differences were obtained for the other samples. The same increase in GP observed in samples containing a chains encapsulated in erythrocytes was obtained in those samples that were incubated in the presence of nitroxides (data not shown). Figure 2 shows the ‘‘oxidation index’’ of lipids extracted from erythrocyte membranes obtained from cells entrapped with a or b chains. The ratio A233nm / A215nm , can be used to evaluate the extent of conjugated diene formation.24 In this case, the data show a significant (p õ .0001) increase in ‘‘oxidation index’’ for samples obtained from erythrocytes encapsulated with a chains after 7 h of incubation at 377C. No significant differences (p ú .05) were observed in the other samples after the same incubation period. Figure 3 shows results obtained on a set of experiments in which erythrocytes entrapped with a chains were incubated in the presence of nitroxides IAL and QAL. The results show that neither IAL nor QAL protected conjugated diene formation and that QAL actually increased the forma-
Fig. 2. Oxidation index obtained by the absorbance ratio A233nm / A215nm measured on lipids extracted from unloaded erythrocytes (normal control) ( h ); resealed erythrocytes without chains (control-resealed) ( j ); and erythrocytes loaded with a ( n ) or b ( L ) Hb chains as described in Materials and Methods. Data represent the means { SD. n Å 6.
/ 2b2b 2465 Mp 280 Wednesday May 28 10:41 PM EL–FRB 2465
Hb-chains entrapment and nitroxides
Fig. 3. Oxidation index obtained by the absorbance ratio A233nm / A215nm measured on lipids extracted from resealed erythrocytes without chains (control-resealed) ( j ); erythrocytes loaded with a Hb chains in the presence of ethanol ( n ); erythrocytes loaded with a Hb chains in the presence of 50 mM IAL (X) or 50 mM QAL ( l ) as described in Materials and Methods. Data represent the means { SD. n Å 6.
tion of dienes right from the beginning of the incubation period. The results of experiments for evaluating the degree of hemolysis carried out with nitroxides are reported in Fig. 4. As can be observed, the presence of nitroxides IAL and QAL retarded hemolysis in erythrocytes encapsulated with a chains. In fact, after 18 h, the extent of hemolysis was as follows: control-resealed erythrocytes, 1.7%; a-chain loaded erythrocytes plus IAL in the suspension medium, 12%; a-chain loaded erythrocytes plus QAL in the suspension medium, 34%; achain loaded erythrocytes in the absence of nitroxides in the suspension medium, 50%. Chemiluminescence determinations were performed by using lucigenin as chemiluminogenic probe for superoxide radical produced by xanthine/xanthine oxidase.27 The reaction of lucigenin with superoxide radical gives rise to chemiluminescence and the level of CL indicates the presence of superoxide in the medium under study.28 Figure 5 shows the kinetics of the lucigenin amplified chemiluminescence measured in the presence and in the absence of IAL and QAL. Data obtained with a fixed concentration (100 mM) show that both nitroxides are efficient in scavenging superoxide radical. The maximum value of CL was 1.595 1 10 4 cps for the control in the presence of the same volume of ethanol as that used for the nitroxide samples. For IAL and QAL the CL values reached were, respectively, 4.333 1 10 2 cps and 2.555 1 10 3 . The
281
Fig. 4. Effect of nitroxide radicals on the degree of hemolysis of resealed erythrocytes without chains in the presence of ethanol (control-resealed) ( j ); erythrocytes loaded with a Hb chains in the presence of ethanol ( n ); erythrocytes loaded with a Hb chains in the presence of 50 mM IAL (X) or 50 mM QAL ( l ) as described in Materials and Methods.
area under the curves are as follows: control, 107.9 cm2 ; 0.4 cm2 for IAL, and 18.1 cm2 for QAL. In Fig. 6 are reported the changes in lucigenin CL when superoxide dismutase (SOD), IAL and QAL were added separately to the superoxide radical-generating system
Fig. 5. Kinetics of the lucigenin-amplified CL measured in the presence and in the absence of IAL and QAL. CL was measured in the presence of 0.9 U/ml xanthine oxidase, 150 mM lucigenin, and 100 mM of IAL or QAL; the reaction was started injecting xanthine at a final concentration of 50 mM in physiological solution. Chemiluminescence was measured as counts per seconds (cps). ( n ) control ( / ethanol); ( l ) QAL; (X) IAL.
/ 2b2b 2465 Mp 281 Wednesday May 28 10:41 PM EL–FRB 2465
282
R. GABBIANELLI et al.
Fig. 6. Inhibition (%) of lucigenin-amplified chemiluminescence of the xanthine/xanthine oxidase system in the presence of superoxide dismutase ( n ), IAL ( l ), QAL ( h ). Chemiluminescence was measured in the presence of 0.9 U/ml xanthine oxidase and 150 mM lucigenin; the reaction was initiated by injecting xanthine at a final concentration of 50 mM in physiological solution. Inhibiton values refer to peak heights.
consisting of xanthine/xanthine oxidase. SOD causes a significant inhibition in CL values: a complete inhibition was obtained using less than 1 mM SOD (no superoxide is present to react with lucigenin as a consequence of the dismutation reaction of superoxide mediated by SOD), and this inhibition is concentration dependent. Both IAL and QAL were able to react with superoxide radical; the effect of different concentrations of IAL and QAL (0.1–10 mM) on the chemiluminescence measurements is the same for both compounds. Obviously, the effect of these two compounds is less pronounced with respect to superoxide dismutase; for example, in the presence of 1 mM of IAL and QAL, the CL level is reduced by 50%, while it is completely absent at the same concentration of SOD. However, at higher concentrations (100 mM) almost complete inhibition was observed by both nitroxides (data not shown).
DISCUSSION
Several reports in the literature regarding in vitro erythrocyte models of thalassemia where free a or b chains are entrapped in normal RBCs have shown that this pathologic condition is characterized by a decrease in protein concentrations and reduced thiol contents of spectrin and ankyrin.7,8,29 The instability and autoxidation of these unpaired free chains play a key role in
the modifications at the protein level, due to an increase in the flux of oxygen radicals.5,8 In addition, a-hemoglobin chain loading results in a direct decrease in catalase 30 and glutathione peroxidase activity.6 At the membrane level, a decrease in cellular and membrane deformability in a-loaded erythrocytes was observed as a consequence of a-chain membrane deposition.7,8 In this study, we have demonstrated that modifications are also present in the lipid component of erythrocyte membranes loaded with a-chains. Fluorescence measurements were performed using laurdan as probe, which is inserted into membranes by its lauric acid tail, with the naphtalene fluorescent moiety located at the level of the phospholipid glycerol backbone.31 Generalized polarization is correlated with the polarity of the environment surrounding the probe, as reported by Parasassi et al.22,23 The data reveal that the GP340 value of these membranes increases in a-loaded erythrocytes (see Fig. 1), and this implies a reduction in the polarity of the environment tested by the probe. In particular, the GP value offers several advantages in the measurement and analysis of laurdan fluorescence data as a high value is related to a gel phase, while a lower one is related to a liquid–crystalline one.32 Laurdan fluorescence has thus demonstrated that the physico-chemical state of the membrane phospholipids are modified in this model of thalassemia. More specifically, the physical state of the phospholipids in samples loaded with a chains shifts towards a gel phase. It is now widely accepted that an increment in conjugated dienes is a valid parameter to evaluate the extent of lipid peroxidation.33 The determination of conjugated dienes was, therefore, used as an oxidation index in this study. The results obtained (Fig. 2) have revealed an increase in the formation of conjugated dienes, and this further confirms that b thalassemia is associated with oxidative stress because the lipid component of erythrocyte membranes is also modified. In fact, an increase in these dienes was observed over time in a-chain–loaded erythrocytes before hemolysis. In thalassemia, erythrocytes are characterized by a short life-time correlated with an increment in the oxidative status of the cell; therefore, an increase in the antioxidant components of the erythrocyte should lengthen the cell’s life cycle.5,8 In this in vitro model of thalassemia, the extent of hemolysis was chosen as a parameter for determining the response to oxidative stress in the presence of antioxidants. Because several studies carried out on indolinic and quinolinic nitroxides have been reported to protect lipoproteins, proteins, and microsomes against free radical-induced damage, 10,11,34,35 it was of interest to determine the capacity of these compounds to protect model thalassemic cells against oxidative damage. The results re-
/ 2b2b 2465 Mp 282 Wednesday May 28 10:41 PM EL–FRB 2465
Hb-chains entrapment and nitroxides
ported in Fig. 4 show that hemolysis was reduced in those suspensions of erythrocytes in which nitroxides IAL and QAL were included. In particular, this protective effect was more enhanced in the presence of IAL. This trend, in which IAL is more efficient than QAL as antioxidant, is in line with the CL measurements, where IAL was shown to scavenge superoxide radical better than QAL (Fig. 5). The protective effect of indolinic and quinolinic nitroxides is in agreement with other reported studies that have shown that the nitroxide TEMPO was capable of mitigating hemolysis induced oxidatively by the drug primaquine.36 However, the protective effect of nitroxide radicals on the hemolysis of a-loaded erythrocytes seems not to be related to protection at the membrane level of these RBCs. From the results obtained by evaluating the laurdan GP of liposomes of a-loaded erythrocytes in which nitroxides were present in the medium, it was observed that there were no differences between these and samples incubated without nitroxides, suggesting that the presence of the latter does not influence the shift of membranes towards a gel phase. Secondly, the oxidation index of suspensions in which the nitroxides are present (Fig. 3) shows that there is no protection of conjugated diene formation, and that conjugated dienes are actually enhanced in the presence of the nitroxide QAL. These results were rather surprising, as it was expected that the nitroxides would protect against conjugated diene formation and the shift of membranes towards a gel phase in the a-loaded erythrocytes. At present, we have no explanation for these phenomena except for the fact that the answer to the protection against hemolysis of a-loaded erythrocytes in the presence of nitroxides IAL and QAL is not at the membrane level. Even though these nitroxides are scavengers of superoxide radical, as demonstrated by the chemiluminescence experiments (Figs. 5 and 6), hemolytic protection might not be necessarily linked entirely to the free radical scavenging activity of nitroxide radicals because these compounds in the system under study do not protect conjugated diene formation. In conclusion, the results obtained in this study underline the fact that this in vitro model of thalassemia represents a valid method for following the physicochemical modifications that characterize this pathologic state, and that the nitroxide radicals IAL and QAL protect against hemolysis induced by this imbalance in hemoglobin chains. Many commonly used nitroxide radical spin probes, such as piperidines and doxylstearates, have no adverse effects on cell survival and are relatively noncytotoxic.37 Preliminary studies in our laboratory on the cytotoxicity and mutagenicity of indolinic and quinolinic nitroxides, soon to be reported, have also confirmed the relative noncytotoxicity and
283
nonmutagenicity of these compounds. This is of interest from the biomedical point of view, because nitroxides could represent an alternative approach in interventions aimed to ameliorate biological injury as a consequence of oxidative stress. Acknowledgements — The authors thank the Italian Ministero dell’Universita` e della Ricerca Scientifica e Tecnologica (MURST) for its financial support.
REFERENCES 1. Schrier, S. L.; Rachmilewitz, E. A.; Mohandas, N. Cellular and membrane properties of a and b thalassemic erythrocytes are different: Implications for differences in clinical manifestations. Blood 74:2194–2202; 1989. 2. Rouyer-Fessard, P.; Scott, M. D.; Garrel, M. C., Bachir, D.; Galacteros, F.; Beuzard, Y. Fate of a hemoglobin chains and erythrocyte defects in b thalassemia. In: Bank, A., ed. Sixth Cooley’s anemia symposium. New York: Annals of the New York Academy of Sciences; 1990:612;106–117. 3. Snyder, L. M.; Fortier, N. L.; Leb, L.; McKenny, J.; Trainor, J.; Sheerin, H.; Mohandas, N. The role of membrane sulphydryl groups in hydrogen peroxide-mediated membrane damage in human erythrocytes. Biochim. Biophys. Acta 937:229–240; 1988. 4. Scott, M. D.; Rouyer-Fessard, P.; Lubin, B. H.; Beuzard, Y. Entrapment of purified a-hemoglobin chains in normal erythrocytes: A model for b thalassemia. J. Biol. Chem. 265:17953– 17959; 1990. 5. Brunori, M.; Falcioni, G.; Fioretti, E.; Giardina, B.; Rotilio, G. Formation of superoxide in the autoxidation of isolated a and b chains of human hemoglobin and its involvement in the hemicromes precipitation. Eur. J. Biochem. 53:99–104; 1975. 6. Grelloni, F.; Gabbianelli, R.; Santroni, A. M.; Falcioni, G. Inactivation of glutathione peroxidase following entrapment of purified a or b hemoglobin chains in human erythrocytes. Clin. Chim. Acta 217:187–192; 1993. 7. Scott, M. D.; Rouyer-Fessard, P.; Soda Ba, M.; Lubin, B. H.; Beuzard, Y. a and b hemoglobin chain induced changes in normal erythrocyte deformability: Comparison to b thalassemia intermedia and Hb H disease. Br.J. Haematol. 80:519–526; 1992. 8. Scott, M. D. Entrapment of purified a-hemoglobin chains in normal erythrocytes as a model for human b thalassemia. In: Magnani, M.; DeLoach, J. R., eds. The use of resealed erythrocytes as carriers and bioreactors. New York: Plenum Press; 1992:139–148. 9. Krishna, M. C.; Samuni, A. Nitroxides as antioxidants. Methods Enzymol. 234:580–589; 1994. 10. Damiani, E.; Paganga, G.; Greci, L.; Rice-Evans, C. Inhibition of copper-mediated low density lipoprotein peroxidation by quinoline and indolinone nitroxide radicals. Biochem. Pharmacol. 48:1155–1161; 1994. 11. Antosiewicz, J.; Popinigis, J.; Wozniak, M.; Damiani, E.; Carloni, P.; Greci, L. Effects of indolinic and quinolinic aminoxyls on protein and lipid peroxidation of rat liver microsomes. Free Radic. Biol. Med. 18:913–917; 1995. 12. Mitchell, J. B.; Samuni, A.; Krishna, M. C.; De Graff, W. G.; Ahn, M. S.; Samuni, U.; Russo, A. Biologically active metalindependent superoxide dismutase mimics. Biochemistry 29:2802–2807; 1990. 13. Krishna, M. C.; Grahame, D. A.; Samuni, A.; Mitchell, J. B.; Russo, A. Oxoammonium cation intermediate in the nitroxidecatalyzed dismutation of superoxide. Proc. Natl. Acad. Sci. USA 89:5537–5541; 1992. 14. Berti, C.; Colonna, M.; Greci, L.; Marchetti, L. Stable nitroxide radicals from phenylisatogen and arylimino derivatives with organometallic compounds. Tetrahedron 31:1745–1753; 1975. 15. Berti, C.; Colonna, M.; Greci, L.; Marchetti, L. Nitroxide radicals
/ 2b2b 2465 Mp 283 Wednesday May 28 10:41 PM EL–FRB 2465
284
16. 17. 18. 19.
20. 21. 22. 23.
24. 25.
26. 27. 28.
R. GABBIANELLI et al. from 2-substituted quinolin-N-oxides with organometallic compounds. Tetrahedron 32:2147–2151; 1976. Bucci, E.; Fronticelli, C. A. A new method for the preparation of a and b subunits of human hemoglobin. J. Biol. Chem. 240:551–552; 1965. Geraci, G.; Parkhurst, L. J.; Gibson, Q. H. Preparation and properties of a and b chains from human hemoglobin. J. Biol. Chem. 244:4664–4667; 1969. DeLoach, J. R.; Ihler, J. A dialysis procedure for loading erythrocytes with enzymes and lipids. Biochim. Biophys. Acta 496:136–145;1977. Bramley, T. A.; Coleman, R.; Finean, J. B. Chemical, enzymological and permeability properties of human erythrocyte ghosts prepared by hypotonic lysis in media of different osmolarities. Biochim. Biophys. Acta 241:752–769; 1971. Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. L. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265–275; 1951. Folch, J.; Less, M.; Sloane-Stanley, G. H. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:466–468; 1957. Parasassi, T.; De Stasio, G.; D’Ubaldo, A.; Gratton, E. Phase fluctuation in phospholipid membranes revealed by Laurdan fluorescence. Biophys. J. 57:1179–1186; 1990. Parasassi, T.; De Stasio, G.; Ravagnan, G.; Rusch, R. M.; Gratton, E. Quantitation of lipid phases in phospholipid vesicles by the generalized polarization of laurdan fluorescence. Biophys. J. 60:120–132; 1991. Konings, A. W. T. Lipid peroxidation in liposomes. In: Gregoriadis, G., ed. Liposomes technology, vol. 1. Boca Raton, FL: CRC Press Inc.; 1984:139–161. Gabbianelli, R.; Santroni, A. M.; Kantar, A.; Falcioni, G. Superoxide anion handling by erythrocytes loaded with alfa and beta hemoglobin chains. A chemiluminescence study. In: Campbell, A. K.; Kricka, L.J.; Sanley, P.E., eds. Bioluminescence and chemiluminescence fundaments and applied aspects. Chichester: John Wiley and Sons; 1994:227–230. Kantar, A.; Oggiano, N.; Gabbianelli, R.; Giorgi, P. L.; Biraghi, M. Effect of imidazole salicylate on the respiratory burst of polymorphonuclear leukocytes. Curr. Ther. Res. 54:1–7; 1993. Cotella, N.; Berneir, J. L.; Henichart, J. P.; Gaydou, E.; Wallet, J. C. Scavenger and antioxidant properties of ten synthetic flavones. Free Radic. Biol. Med. 13:211–219; 1992. Murphy, M. E.; Sies, H. Visible range low level chemiluminescence in the biological system. Methods Enzymol. 186:595–610; 1990.
29. Rouyer-Fessard, P.; Garel, M. C.; Domenget, C.; Guetarnis, D.; Bachir, D.; Colonna, P.; Beuzard, Y. A study of membrane protein defects and a hemoglobin chains of red blood cells in human b thalassemia. J. Biol. Chem. 264:19092–19098; 1989. 30. Scott, M. D.; Van Den Berg, J. J.; Reptka, T.; Rouyer-Fessard, P.; Hebbel, R. P.; Beuzard, Y.; Lubin, B. H. Effect of excess a hemoglobin chains on cellular and membrane oxidation in model b thalassemic erythrocytes. J. Clin. Invest. 91:1706–1712; 1993. 31. Parasassi, T.; Conti, F.; Gratton, E. Time-resolved fluorescence emission spectra of laurdan in phospholipid vesicles by multifrequency phase and modulation fluorometry. Cell Mol. Biol. 32:103–108; 1986. 32. Parasassi, T.; Di Stefano, M.; Loiero, M.; Ravagnan, G.; Gratton, E. Influence of cholesterol on phospholipid bilayer phase domains as detected by laurdan fluorescence. Biophys. J. 66:120– 132; 1994. 33. Tanfani, F.; Curatola, G.; Bertoli, E. Steady-state fluorescence anisotropy and multifrequency phase fluorometry on oxidized phosphatidylcholine vesicles. Chem. Phys. Lipids 50:1–9; 1989. 34. Tanfani, F.; Carloni, P.; Damiani, E.; Greci, L.; Wozniak, M.; Kulawiak, D.; Jankowski, K.; Kaczor, J.; Matuskiewics, A. Quinolinic aminoxyl protects albumin against peroxyl radical mediated damage. Free Radic. Res. 21:309–315; 1994. 35. Antosiewicz, J.; Bertoli, E.; Damiani, E.; Greci, L.; Popinigis, J.; Przybylski, S.; Tanfani, F.; Wozniak, M. Indolinonic and quinolinic aminoxyls as protectants against oxidative stress. Free Radic. Biol. Med. 15:203–208; 1993. 36. Grinberg, L. N.; Samuni, A. Nitroxide stable radical prevents primaquine-induced lysis of red blood cell. Biochim. Biophys. Acta 1201:284–288; 1994. 37. Ankel, E. G.; Lai, C.-S.; Hopwood, L. E.; Zivkovic, Z. Cytotoxicity of commonly used nitroxide spin probes. Life Sci. 40:495– 498; 1987.
ABBREVIATIONS
CL—chemiluminescence GP—generalized polarization Hb—hemoglobin RBCs—red blood cells ROS—reactive oxygen species TEMPO—2,2,4,4-tetramethylpiperidine-N-oxyl
/ 2b2b 2465 Mp 284 Wednesday May 28 10:41 PM EL–FRB 2465
PII S0891-5849(96)00611-9
Original Contribution EFFECT OF AROMATIC NITROXIDES ON HEMOLYSIS OF HUMAN ERYTHROCYTES ENTRAPPED WITH ISOLATED HEMOGLOBIN CHAINS ROSITA GABBIANELLI,* GIANCARLO FALCIONI,* ANNA MARIA SANTRONI,* GIANCARLO CAULINI,* LUCEDIO GRECI, † and ELISABETTA DAMIANI † *Dipartimento di Biologia Molecolare Cellulare Animale, Universita`, Via Camerini, 62032 Camerino, Italy; † Dipartimento di Scienze dei Materiali e della Terra, Universita`, Via Brecce Bianche, I-60131 Ancona, Italy (Received 5 January 1996; Accepted 5 December 1996)
Abstract—An in vitro model of thalassemia was produced by entrapment of isolated hemoglobin chains in human erythrocytes, thus subjecting the loaded cells to oxidative stress. The presence of these unpaired chains induced physico-chemical modifications at the membrane level as studied by laurdan fluorescence. The polarity of the lipid bilayer was shown to decrease with a concomitant shift towards a gel phase in a-loaded erythrocytes. The determination of conjugated dienes before the hemolytic event was used as an oxidation index; the results obtained demonstrate that b thalassemia is associated with oxidative stress. Furthermore, the presence of indolinic and quinolinic nitroxide radicals, a new class of antioxidants, in suspensions of a-loaded erythrocytes protected the erythrocytes from the hemolytic event. However, the protective effect exerted by the nitroxide radicals is not related to effects on membrane polarity and lipid peroxidation, even though a chemiluminescence study has demonstrated the superoxide scavenging activity of these nitroxide radicals. q 1997 Elsevier Science Inc. Keywords—RBCs, Hemoglobin-chains entrapment, Oxidative stress, Fluorescence, Hemolysis, Nitroxides, Chemiluminescence
cies (ROS) is altered due to autoxidation of hemoglobin and the release of the superoxide radical.5 Moreover, the RBCs antioxidative system can be impaired; in fact, we have recently demonstrated inactivation of glutathione peroxidase.6 A simple in vitro method of reproducing thalassemia is to entrap isolated a or b hemoglobin chains in human erythrocytes.4,6 – 8 This model represents a valid system for studying the cell alterations determined by the presence of free a or b chains, because erythrocytes from thalassemic patients have already undergone in vivo damage. The aim of this work was to evaluate in human a or b chains encapsulated in erythrocytes the physicochemical modifications at the membrane level. For this study, laurdan [2-dimethylamino-(6-lauroyl)-naphtalene] was selected to probe the changes in membrane polarity before the hemolytic event. This probe incorporates at the hydrophilic–hydrophobic interface of the membrane with the lauric acid tail anchored in the hydrophobic region of the bilayer. The probe displays a
INTRODUCTION
Thalassemia is related to an a or b chain imbalance in erythrocytes. These free hemoglobin chains are unstable and are readily oxidized. Their denaturation and binding to the erythrocyte membrane give rise to cytoskeleton modifications and lipid peroxidation, which are the cause of the many morphological, rheological, and biochemical defects of thalassemic red blood cells.1 – 4 In fact, several studies have shown that isolated a or b hemoglobin chains are less stable than tetrameric hemoglobin and that they rapidly autooxidise with the subsequent formation of precipitates known as Heinz bodies according to the following scheme: oxy-chains r met-chains r hemichromes r precipitates (Heinz bodies). During the process that leads to the formation of Heinz bodies, the production of reactive oxygen speAddress correspondence to: Prof. Lucedio Greci, Dipartimento di Scienze dei Materiali e della Terra, Universita`, Via Brecce Bianche, I-60131 Ancona, Italy. 278
/ 2b2b 2465 Mp 278 Wednesday May 28 10:41 PM EL–FRB 2465
Hb-chains entrapment and nitroxides
spectral sensitivity to the polarity of its surroundings, which is related to the physical state and dynamics of the surrounding phospholipid polar head group. Measurement of conjugated dienes was also performed to evaluate the extent of lipid peroxidation. Recently, the antioxidative activity of nitroxide radicals in various biological systems has been established.9 Therefore, indolinic and quinolinic nitroxide radicals, known to possess free radical scavenging activity, 10,11 were included in the study to determine their possible protective role in mitigating hemolysis in this in vitro model of thalassemia. In addition, by using the chemiluminescence technique, the superoxide scavenging activity of indolinic and quinolinic nitroxides was determined and compared to superoxide dismutase because nitroxide radicals are reported to possess SOD-mimic activity.12,13 MATERIALS AND METHODS
All reagents were of pure analytical grade. Laurdan was purchased from Molecular Probes (Eugene, OR). Xanthine, xanthine oxidase (E.C.1.1.3.22), superoxide dismutase (E.C.1.15.1.1) from bovine erythrocytes and lucigenin were obtained from Sigma Chemical Co., St. Louis, MO. Nitroxide radicals IAL (1,2-dihydro-2ethyl-2-phenyl-3H-indole-3-phenylimino-1-oxyl) and QAL (1,2-dihydro-2,2-diphenyl-4-ethoxy-quinoline-1oxyl) were prepared according to Berti et al. (Scheme 1).14,15 a and b chains were prepared from human hemoglobin using p-hydroxymercuribenzoate according to Bucci and Fronticelli.16 The regeneration of sulphydryl groups was carried out by treatment with mercaptoethanol on a chromatografic column as described previously.17 The purified chains containing regenerated sulphydryl groups, were encapsulated in human erythrocytes by a dialysis technique involving transient hypotonic hemolysis followed by isotonic resealing.6,18 The suspensions of RBCs in which a or b chains were entrapped were incubated in appropriate isotonic buffer at 377C and sampling was performed at the beginning (0 h) and after 7 h of incubation, before the hemolytic event. The degree of hemolysis was determined as (100 1 A/10 1 A*100%), where A is the optical density of Hb present in the supernatant after centrifugation of red cell suspension, and A*100% is the optical density of the red cell suspension after complete lysis with 10 vol distilled water at zero time incubation. To compare data obtained from cells in which a or b chains were encapsulated, two different controls were also prepared, one with resealed erythrocytes without chains (unloaded but opened and resealed) and the other one with unloaded erythrocytes (normal control).
279
Scheme 1. IAL: 1,2-dihydro-2-ethyl-2-phenyl-3H-indole-3-phenylimino-1-oxyl; QAL: 1,2-dihydro-2,2-diphenyl-4-ethoxy-quinoline-1oxyl.
For fluorescence experiments, erythrocyte membranes were prepared by hypotonic hemolysis.19 All samples were normalized by Lowry’s method.20 Steady-state fluorescence measurements on phospholipids extracted from erythrocyte membranes were performed on an Hitachi 4500 spectrofluorometer. Liposomes were prepared according to the method of Folch; 21 briefly, lipids were resuspended in CHCl3 / CH3OH 3:1 and mixed with the probe and dried in a gentle stream of N2 . The final probe:lipid molar ratio was 600:1. The phospholipid mixture was rehydrated with 10 mM Tris containing 154 mM NaCl, pH 7.4. Generalized polarization ( GP ) for laurdan, described by Parasassi et al., 22 was calculated by exciting laurdan at 340 nm ( GP340 ) according to the following equation: GP Å
IB0IR IB/ IR
(1)
where IB and IR are the emission intensities at the blue (440 nm) and red (490 nm) edges of the emission spectrum and correspond to the fluorescence emission maxima in the gel and liquid–crystalline phases, respectively, of the bilayer.23 Fluorescence measurements were performed at 377C. The absorbance ratio A233nm /A215nm , defined as the ‘‘oxidation index’’ was used as a relative measurement for conjugated dienes according to Konings.24 UV measurements were carried out using a Carry 1 Varian spectrophotometer at 257C. Because it has been demonstrated that a chains entrapped in erythrocytes give rise to an increase in hemolysis, 4,6 – 8 and having observed in this study that modifications are also present at the membrane level as studied by GP and conjugated diene, the effect of the nitroxides on these parameters was evaluated. The degree of hemolysis, the GP value, and the oxidation index of erythrocytes entrapped with a chains were determined on samples in which nitroxide radicals were
/ 2b2b 2465 Mp 279 Wednesday May 28 10:41 PM EL–FRB 2465
280
R. GABBIANELLI et al.
Fig. 1. Laurdan GP340 measured on lipids extracted from unloaded erythrocytes (normal control) ( h ); resealed erythrocytes without chains (control-resealed) ( j ); and erythrocytes loaded with a ( n ) or b ( L ) Hb chains as described in Materials and Methods. Data represent the means { SD. n Å 6.
included in the RBCs suspension medium. Experiments with nitroxides dissolved in ethanol were performed in Krebs buffer (143 mM NaCl, 5.7 mM KCl, 1.4 mM MgCl2 , 18 mM phosphate buffer, pH 7.4). In this case 50 mM (final concentration) of nitroxides were included in the erythrocyte suspensions and incubated at 377C. For all these experiments, control samples were prepared in the presence of the same volume of ethanol used in the suspensions containing nitroxides. Chemiluminescence measurements were performed using lucigenin as chemiluminogenic probe, and superoxide radical was produced by the xanthine/xanthine oxidase system.25 The chemiluminescence (CL) was measured in Autolumat LB 953 (Berthold Co. Wildbad, Germany) in a reaction mixture containing 0.9 U/ml xanthine oxidase, 150 mM lucigenin in physiological solution and different concentrations of nitroxides dissolved in ethanol. The reaction was started by injecting xanthine at a final concentration of 50 mM and followed for 60 s as described previously.26 The values are expressed as counts per second (cps). All the results presented are representative of at least three experiments, and appropriate controls were carried out throughout. The significance of values was obtained by using Student’s t-test. A p-value õ0.05 was considered statistically significant. RESULTS
The results on fluorescence experiments of Generalized Polarization for laurdan embedded in liposomes
consisting of lipids extracted from erythrocyte membranes are shown in Fig. 1. To evaluate the modifications in the lipid bilayer before the hemolytic event, the samples were analyzed both at 0 h and after 7 h of incubation. The 7-h time frame was established because up to this time the extent of hemolysis was irrelevant. The data obtained show a significant (p õ .0001) increase in GP340 values for samples in which a chains were encapsulated. No significant (p ú .05) differences were obtained for the other samples. The same increase in GP observed in samples containing a chains encapsulated in erythrocytes was obtained in those samples that were incubated in the presence of nitroxides (data not shown). Figure 2 shows the ‘‘oxidation index’’ of lipids extracted from erythrocyte membranes obtained from cells entrapped with a or b chains. The ratio A233nm / A215nm , can be used to evaluate the extent of conjugated diene formation.24 In this case, the data show a significant (p õ .0001) increase in ‘‘oxidation index’’ for samples obtained from erythrocytes encapsulated with a chains after 7 h of incubation at 377C. No significant differences (p ú .05) were observed in the other samples after the same incubation period. Figure 3 shows results obtained on a set of experiments in which erythrocytes entrapped with a chains were incubated in the presence of nitroxides IAL and QAL. The results show that neither IAL nor QAL protected conjugated diene formation and that QAL actually increased the forma-
Fig. 2. Oxidation index obtained by the absorbance ratio A233nm / A215nm measured on lipids extracted from unloaded erythrocytes (normal control) ( h ); resealed erythrocytes without chains (control-resealed) ( j ); and erythrocytes loaded with a ( n ) or b ( L ) Hb chains as described in Materials and Methods. Data represent the means { SD. n Å 6.
/ 2b2b 2465 Mp 280 Wednesday May 28 10:41 PM EL–FRB 2465
Hb-chains entrapment and nitroxides
Fig. 3. Oxidation index obtained by the absorbance ratio A233nm / A215nm measured on lipids extracted from resealed erythrocytes without chains (control-resealed) ( j ); erythrocytes loaded with a Hb chains in the presence of ethanol ( n ); erythrocytes loaded with a Hb chains in the presence of 50 mM IAL (X) or 50 mM QAL ( l ) as described in Materials and Methods. Data represent the means { SD. n Å 6.
tion of dienes right from the beginning of the incubation period. The results of experiments for evaluating the degree of hemolysis carried out with nitroxides are reported in Fig. 4. As can be observed, the presence of nitroxides IAL and QAL retarded hemolysis in erythrocytes encapsulated with a chains. In fact, after 18 h, the extent of hemolysis was as follows: control-resealed erythrocytes, 1.7%; a-chain loaded erythrocytes plus IAL in the suspension medium, 12%; a-chain loaded erythrocytes plus QAL in the suspension medium, 34%; achain loaded erythrocytes in the absence of nitroxides in the suspension medium, 50%. Chemiluminescence determinations were performed by using lucigenin as chemiluminogenic probe for superoxide radical produced by xanthine/xanthine oxidase.27 The reaction of lucigenin with superoxide radical gives rise to chemiluminescence and the level of CL indicates the presence of superoxide in the medium under study.28 Figure 5 shows the kinetics of the lucigenin amplified chemiluminescence measured in the presence and in the absence of IAL and QAL. Data obtained with a fixed concentration (100 mM) show that both nitroxides are efficient in scavenging superoxide radical. The maximum value of CL was 1.595 1 10 4 cps for the control in the presence of the same volume of ethanol as that used for the nitroxide samples. For IAL and QAL the CL values reached were, respectively, 4.333 1 10 2 cps and 2.555 1 10 3 . The
281
Fig. 4. Effect of nitroxide radicals on the degree of hemolysis of resealed erythrocytes without chains in the presence of ethanol (control-resealed) ( j ); erythrocytes loaded with a Hb chains in the presence of ethanol ( n ); erythrocytes loaded with a Hb chains in the presence of 50 mM IAL (X) or 50 mM QAL ( l ) as described in Materials and Methods.
area under the curves are as follows: control, 107.9 cm2 ; 0.4 cm2 for IAL, and 18.1 cm2 for QAL. In Fig. 6 are reported the changes in lucigenin CL when superoxide dismutase (SOD), IAL and QAL were added separately to the superoxide radical-generating system
Fig. 5. Kinetics of the lucigenin-amplified CL measured in the presence and in the absence of IAL and QAL. CL was measured in the presence of 0.9 U/ml xanthine oxidase, 150 mM lucigenin, and 100 mM of IAL or QAL; the reaction was started injecting xanthine at a final concentration of 50 mM in physiological solution. Chemiluminescence was measured as counts per seconds (cps). ( n ) control ( / ethanol); ( l ) QAL; (X) IAL.
/ 2b2b 2465 Mp 281 Wednesday May 28 10:41 PM EL–FRB 2465
282
R. GABBIANELLI et al.
Fig. 6. Inhibition (%) of lucigenin-amplified chemiluminescence of the xanthine/xanthine oxidase system in the presence of superoxide dismutase ( n ), IAL ( l ), QAL ( h ). Chemiluminescence was measured in the presence of 0.9 U/ml xanthine oxidase and 150 mM lucigenin; the reaction was initiated by injecting xanthine at a final concentration of 50 mM in physiological solution. Inhibiton values refer to peak heights.
consisting of xanthine/xanthine oxidase. SOD causes a significant inhibition in CL values: a complete inhibition was obtained using less than 1 mM SOD (no superoxide is present to react with lucigenin as a consequence of the dismutation reaction of superoxide mediated by SOD), and this inhibition is concentration dependent. Both IAL and QAL were able to react with superoxide radical; the effect of different concentrations of IAL and QAL (0.1–10 mM) on the chemiluminescence measurements is the same for both compounds. Obviously, the effect of these two compounds is less pronounced with respect to superoxide dismutase; for example, in the presence of 1 mM of IAL and QAL, the CL level is reduced by 50%, while it is completely absent at the same concentration of SOD. However, at higher concentrations (100 mM) almost complete inhibition was observed by both nitroxides (data not shown).
DISCUSSION
Several reports in the literature regarding in vitro erythrocyte models of thalassemia where free a or b chains are entrapped in normal RBCs have shown that this pathologic condition is characterized by a decrease in protein concentrations and reduced thiol contents of spectrin and ankyrin.7,8,29 The instability and autoxidation of these unpaired free chains play a key role in
the modifications at the protein level, due to an increase in the flux of oxygen radicals.5,8 In addition, a-hemoglobin chain loading results in a direct decrease in catalase 30 and glutathione peroxidase activity.6 At the membrane level, a decrease in cellular and membrane deformability in a-loaded erythrocytes was observed as a consequence of a-chain membrane deposition.7,8 In this study, we have demonstrated that modifications are also present in the lipid component of erythrocyte membranes loaded with a-chains. Fluorescence measurements were performed using laurdan as probe, which is inserted into membranes by its lauric acid tail, with the naphtalene fluorescent moiety located at the level of the phospholipid glycerol backbone.31 Generalized polarization is correlated with the polarity of the environment surrounding the probe, as reported by Parasassi et al.22,23 The data reveal that the GP340 value of these membranes increases in a-loaded erythrocytes (see Fig. 1), and this implies a reduction in the polarity of the environment tested by the probe. In particular, the GP value offers several advantages in the measurement and analysis of laurdan fluorescence data as a high value is related to a gel phase, while a lower one is related to a liquid–crystalline one.32 Laurdan fluorescence has thus demonstrated that the physico-chemical state of the membrane phospholipids are modified in this model of thalassemia. More specifically, the physical state of the phospholipids in samples loaded with a chains shifts towards a gel phase. It is now widely accepted that an increment in conjugated dienes is a valid parameter to evaluate the extent of lipid peroxidation.33 The determination of conjugated dienes was, therefore, used as an oxidation index in this study. The results obtained (Fig. 2) have revealed an increase in the formation of conjugated dienes, and this further confirms that b thalassemia is associated with oxidative stress because the lipid component of erythrocyte membranes is also modified. In fact, an increase in these dienes was observed over time in a-chain–loaded erythrocytes before hemolysis. In thalassemia, erythrocytes are characterized by a short life-time correlated with an increment in the oxidative status of the cell; therefore, an increase in the antioxidant components of the erythrocyte should lengthen the cell’s life cycle.5,8 In this in vitro model of thalassemia, the extent of hemolysis was chosen as a parameter for determining the response to oxidative stress in the presence of antioxidants. Because several studies carried out on indolinic and quinolinic nitroxides have been reported to protect lipoproteins, proteins, and microsomes against free radical-induced damage, 10,11,34,35 it was of interest to determine the capacity of these compounds to protect model thalassemic cells against oxidative damage. The results re-
/ 2b2b 2465 Mp 282 Wednesday May 28 10:41 PM EL–FRB 2465
Hb-chains entrapment and nitroxides
ported in Fig. 4 show that hemolysis was reduced in those suspensions of erythrocytes in which nitroxides IAL and QAL were included. In particular, this protective effect was more enhanced in the presence of IAL. This trend, in which IAL is more efficient than QAL as antioxidant, is in line with the CL measurements, where IAL was shown to scavenge superoxide radical better than QAL (Fig. 5). The protective effect of indolinic and quinolinic nitroxides is in agreement with other reported studies that have shown that the nitroxide TEMPO was capable of mitigating hemolysis induced oxidatively by the drug primaquine.36 However, the protective effect of nitroxide radicals on the hemolysis of a-loaded erythrocytes seems not to be related to protection at the membrane level of these RBCs. From the results obtained by evaluating the laurdan GP of liposomes of a-loaded erythrocytes in which nitroxides were present in the medium, it was observed that there were no differences between these and samples incubated without nitroxides, suggesting that the presence of the latter does not influence the shift of membranes towards a gel phase. Secondly, the oxidation index of suspensions in which the nitroxides are present (Fig. 3) shows that there is no protection of conjugated diene formation, and that conjugated dienes are actually enhanced in the presence of the nitroxide QAL. These results were rather surprising, as it was expected that the nitroxides would protect against conjugated diene formation and the shift of membranes towards a gel phase in the a-loaded erythrocytes. At present, we have no explanation for these phenomena except for the fact that the answer to the protection against hemolysis of a-loaded erythrocytes in the presence of nitroxides IAL and QAL is not at the membrane level. Even though these nitroxides are scavengers of superoxide radical, as demonstrated by the chemiluminescence experiments (Figs. 5 and 6), hemolytic protection might not be necessarily linked entirely to the free radical scavenging activity of nitroxide radicals because these compounds in the system under study do not protect conjugated diene formation. In conclusion, the results obtained in this study underline the fact that this in vitro model of thalassemia represents a valid method for following the physicochemical modifications that characterize this pathologic state, and that the nitroxide radicals IAL and QAL protect against hemolysis induced by this imbalance in hemoglobin chains. Many commonly used nitroxide radical spin probes, such as piperidines and doxylstearates, have no adverse effects on cell survival and are relatively noncytotoxic.37 Preliminary studies in our laboratory on the cytotoxicity and mutagenicity of indolinic and quinolinic nitroxides, soon to be reported, have also confirmed the relative noncytotoxicity and
283
nonmutagenicity of these compounds. This is of interest from the biomedical point of view, because nitroxides could represent an alternative approach in interventions aimed to ameliorate biological injury as a consequence of oxidative stress. Acknowledgements — The authors thank the Italian Ministero dell’Universita` e della Ricerca Scientifica e Tecnologica (MURST) for its financial support.
REFERENCES 1. Schrier, S. L.; Rachmilewitz, E. A.; Mohandas, N. Cellular and membrane properties of a and b thalassemic erythrocytes are different: Implications for differences in clinical manifestations. Blood 74:2194–2202; 1989. 2. Rouyer-Fessard, P.; Scott, M. D.; Garrel, M. C., Bachir, D.; Galacteros, F.; Beuzard, Y. Fate of a hemoglobin chains and erythrocyte defects in b thalassemia. In: Bank, A., ed. Sixth Cooley’s anemia symposium. New York: Annals of the New York Academy of Sciences; 1990:612;106–117. 3. Snyder, L. M.; Fortier, N. L.; Leb, L.; McKenny, J.; Trainor, J.; Sheerin, H.; Mohandas, N. The role of membrane sulphydryl groups in hydrogen peroxide-mediated membrane damage in human erythrocytes. Biochim. Biophys. Acta 937:229–240; 1988. 4. Scott, M. D.; Rouyer-Fessard, P.; Lubin, B. H.; Beuzard, Y. Entrapment of purified a-hemoglobin chains in normal erythrocytes: A model for b thalassemia. J. Biol. Chem. 265:17953– 17959; 1990. 5. Brunori, M.; Falcioni, G.; Fioretti, E.; Giardina, B.; Rotilio, G. Formation of superoxide in the autoxidation of isolated a and b chains of human hemoglobin and its involvement in the hemicromes precipitation. Eur. J. Biochem. 53:99–104; 1975. 6. Grelloni, F.; Gabbianelli, R.; Santroni, A. M.; Falcioni, G. Inactivation of glutathione peroxidase following entrapment of purified a or b hemoglobin chains in human erythrocytes. Clin. Chim. Acta 217:187–192; 1993. 7. Scott, M. D.; Rouyer-Fessard, P.; Soda Ba, M.; Lubin, B. H.; Beuzard, Y. a and b hemoglobin chain induced changes in normal erythrocyte deformability: Comparison to b thalassemia intermedia and Hb H disease. Br.J. Haematol. 80:519–526; 1992. 8. Scott, M. D. Entrapment of purified a-hemoglobin chains in normal erythrocytes as a model for human b thalassemia. In: Magnani, M.; DeLoach, J. R., eds. The use of resealed erythrocytes as carriers and bioreactors. New York: Plenum Press; 1992:139–148. 9. Krishna, M. C.; Samuni, A. Nitroxides as antioxidants. Methods Enzymol. 234:580–589; 1994. 10. Damiani, E.; Paganga, G.; Greci, L.; Rice-Evans, C. Inhibition of copper-mediated low density lipoprotein peroxidation by quinoline and indolinone nitroxide radicals. Biochem. Pharmacol. 48:1155–1161; 1994. 11. Antosiewicz, J.; Popinigis, J.; Wozniak, M.; Damiani, E.; Carloni, P.; Greci, L. Effects of indolinic and quinolinic aminoxyls on protein and lipid peroxidation of rat liver microsomes. Free Radic. Biol. Med. 18:913–917; 1995. 12. Mitchell, J. B.; Samuni, A.; Krishna, M. C.; De Graff, W. G.; Ahn, M. S.; Samuni, U.; Russo, A. Biologically active metalindependent superoxide dismutase mimics. Biochemistry 29:2802–2807; 1990. 13. Krishna, M. C.; Grahame, D. A.; Samuni, A.; Mitchell, J. B.; Russo, A. Oxoammonium cation intermediate in the nitroxidecatalyzed dismutation of superoxide. Proc. Natl. Acad. Sci. USA 89:5537–5541; 1992. 14. Berti, C.; Colonna, M.; Greci, L.; Marchetti, L. Stable nitroxide radicals from phenylisatogen and arylimino derivatives with organometallic compounds. Tetrahedron 31:1745–1753; 1975. 15. Berti, C.; Colonna, M.; Greci, L.; Marchetti, L. Nitroxide radicals
/ 2b2b 2465 Mp 283 Wednesday May 28 10:41 PM EL–FRB 2465
284
16. 17. 18. 19.
20. 21. 22. 23.
24. 25.
26. 27. 28.
R. GABBIANELLI et al. from 2-substituted quinolin-N-oxides with organometallic compounds. Tetrahedron 32:2147–2151; 1976. Bucci, E.; Fronticelli, C. A. A new method for the preparation of a and b subunits of human hemoglobin. J. Biol. Chem. 240:551–552; 1965. Geraci, G.; Parkhurst, L. J.; Gibson, Q. H. Preparation and properties of a and b chains from human hemoglobin. J. Biol. Chem. 244:4664–4667; 1969. DeLoach, J. R.; Ihler, J. A dialysis procedure for loading erythrocytes with enzymes and lipids. Biochim. Biophys. Acta 496:136–145;1977. Bramley, T. A.; Coleman, R.; Finean, J. B. Chemical, enzymological and permeability properties of human erythrocyte ghosts prepared by hypotonic lysis in media of different osmolarities. Biochim. Biophys. Acta 241:752–769; 1971. Lowry, O. H.; Rosebrough, N. J.; Farr, A. L.; Randall, R. L. Protein measurement with the Folin phenol reagent. J. Biol. Chem. 193:265–275; 1951. Folch, J.; Less, M.; Sloane-Stanley, G. H. A simple method for the isolation and purification of total lipids from animal tissues. J. Biol. Chem. 226:466–468; 1957. Parasassi, T.; De Stasio, G.; D’Ubaldo, A.; Gratton, E. Phase fluctuation in phospholipid membranes revealed by Laurdan fluorescence. Biophys. J. 57:1179–1186; 1990. Parasassi, T.; De Stasio, G.; Ravagnan, G.; Rusch, R. M.; Gratton, E. Quantitation of lipid phases in phospholipid vesicles by the generalized polarization of laurdan fluorescence. Biophys. J. 60:120–132; 1991. Konings, A. W. T. Lipid peroxidation in liposomes. In: Gregoriadis, G., ed. Liposomes technology, vol. 1. Boca Raton, FL: CRC Press Inc.; 1984:139–161. Gabbianelli, R.; Santroni, A. M.; Kantar, A.; Falcioni, G. Superoxide anion handling by erythrocytes loaded with alfa and beta hemoglobin chains. A chemiluminescence study. In: Campbell, A. K.; Kricka, L.J.; Sanley, P.E., eds. Bioluminescence and chemiluminescence fundaments and applied aspects. Chichester: John Wiley and Sons; 1994:227–230. Kantar, A.; Oggiano, N.; Gabbianelli, R.; Giorgi, P. L.; Biraghi, M. Effect of imidazole salicylate on the respiratory burst of polymorphonuclear leukocytes. Curr. Ther. Res. 54:1–7; 1993. Cotella, N.; Berneir, J. L.; Henichart, J. P.; Gaydou, E.; Wallet, J. C. Scavenger and antioxidant properties of ten synthetic flavones. Free Radic. Biol. Med. 13:211–219; 1992. Murphy, M. E.; Sies, H. Visible range low level chemiluminescence in the biological system. Methods Enzymol. 186:595–610; 1990.
29. Rouyer-Fessard, P.; Garel, M. C.; Domenget, C.; Guetarnis, D.; Bachir, D.; Colonna, P.; Beuzard, Y. A study of membrane protein defects and a hemoglobin chains of red blood cells in human b thalassemia. J. Biol. Chem. 264:19092–19098; 1989. 30. Scott, M. D.; Van Den Berg, J. J.; Reptka, T.; Rouyer-Fessard, P.; Hebbel, R. P.; Beuzard, Y.; Lubin, B. H. Effect of excess a hemoglobin chains on cellular and membrane oxidation in model b thalassemic erythrocytes. J. Clin. Invest. 91:1706–1712; 1993. 31. Parasassi, T.; Conti, F.; Gratton, E. Time-resolved fluorescence emission spectra of laurdan in phospholipid vesicles by multifrequency phase and modulation fluorometry. Cell Mol. Biol. 32:103–108; 1986. 32. Parasassi, T.; Di Stefano, M.; Loiero, M.; Ravagnan, G.; Gratton, E. Influence of cholesterol on phospholipid bilayer phase domains as detected by laurdan fluorescence. Biophys. J. 66:120– 132; 1994. 33. Tanfani, F.; Curatola, G.; Bertoli, E. Steady-state fluorescence anisotropy and multifrequency phase fluorometry on oxidized phosphatidylcholine vesicles. Chem. Phys. Lipids 50:1–9; 1989. 34. Tanfani, F.; Carloni, P.; Damiani, E.; Greci, L.; Wozniak, M.; Kulawiak, D.; Jankowski, K.; Kaczor, J.; Matuskiewics, A. Quinolinic aminoxyl protects albumin against peroxyl radical mediated damage. Free Radic. Res. 21:309–315; 1994. 35. Antosiewicz, J.; Bertoli, E.; Damiani, E.; Greci, L.; Popinigis, J.; Przybylski, S.; Tanfani, F.; Wozniak, M. Indolinonic and quinolinic aminoxyls as protectants against oxidative stress. Free Radic. Biol. Med. 15:203–208; 1993. 36. Grinberg, L. N.; Samuni, A. Nitroxide stable radical prevents primaquine-induced lysis of red blood cell. Biochim. Biophys. Acta 1201:284–288; 1994. 37. Ankel, E. G.; Lai, C.-S.; Hopwood, L. E.; Zivkovic, Z. Cytotoxicity of commonly used nitroxide spin probes. Life Sci. 40:495– 498; 1987.
ABBREVIATIONS
CL—chemiluminescence GP—generalized polarization Hb—hemoglobin RBCs—red blood cells ROS—reactive oxygen species TEMPO—2,2,4,4-tetramethylpiperidine-N-oxyl
/ 2b2b 2465 Mp 284 Wednesday May 28 10:41 PM EL–FRB 2465