1G. Celed6n*, ~V. Lips, 1C. Alvarado, 1M. Cort6s, 2E. A. Lissi and 3G. Gonzfilez

Transcription

1 Vol. 43, No. 5, December 1997 Pages ] 27 PROTEIN DEGRADATION IN RED CELLS EXPOSED TO 2,2'-AZO-BIS(2-AMIDINOPROPANE) DERIVED RADICALS 1G. Celed6n*, ~V. Lips, 1C. Alvarado, 1M. Cort6s, 2E. A. Lissi and 3G. Gonzfilez 1Departamento de Fisiologia, Facultad de Ciencias, Universidad de Valparaiso, CasiUa 5030, Valparaiso, Chile. 2 Departamento de Quimica, Facultad de Quimica y Biologia, Universidad de Santiago, Chile. 3 Instituto de Quimica, Universidad Catolica de Valparaiso, Chile Received March 18, 1997 Received after revision July 2, 1997 SUMMARY Extensive proteolysis is observed when red blood cells are exposed to free radicals produced in the thermotysis of 2,2'-azo-bis(2-amidinopropane). It is evaluated that nearly one amino terminal group is produced by each free radical introduced into the system. These groups are considered to arise mainly from band 3 fragmentation due to the action of red cell proteinases. Protein fragmentation takes place prior to significant hemolysis or lipid peroxidation, as evaluated by thiobarbituric acid-reactive substances measurements. Key Words: oxidative damage, proteolysis, human membrane erythrocyte, azo compounds, band 3, lipoperoxidation. INTRODUCTION Oxy radicals have been suggested to play a role in mediating the tissue injury associated with the metabolism of certain drugs and xenobiotics (1). Although the mechanisms by which reactive metabolites of oxygen injure cells and tissues remain rather speculative, it has been suggested that oxidation &membrane associated polyunsaturated fatty acids represents a major pathway by which free radicals mediate their cytotoxic effect. However, recent studies indicate that proteins may represent important targets for oxidative tissue injury (2-5). The use of the Abbreviations: AAPH: 2,2'-azo-bis(2-amidinopropane) dihydrochloride TBARS: PMSF: thiobarbituric acid-reactive substance phenylmethylsulphonyl fluoride To whom correspondence should be addressed (FAX ) /97/05 l l /0 Copyright 1997 by Academic Press A ustralia. All rights of reproduction in any form reserved.

2 water soluble azocompound 2,2'-azo-bis(2-amidinopropane)dihydrochloride (AAPH) as a thermolabile peroxyl radical generator, allows studies of membrane events underlying hemolysis (6-8). As radicals are generated in the extracellular medium, this experimental model presents the advantage that oxidative damage is centered in plasma membrane with no or very little influence on intracellutar constituents (4). Moreover, it has been reported that AAPH in erythrocytes induces lipoperoxidation and protein damage independently (9) and, recently, a competitive reaction between lipoperoxidation and band 3 oxidation has been proposed (8). It is known that oxidatively modified proteins are more susceptible to proteolysis (10-13) and several proteolytic enzymes of erythrocytes could be involved in the degradation of proteins oxidized by oxyge n radicals, e.g. the multicatalytic proteinase complex (12), membrane bound proteinases (14) and calpain I (15). The objective of the present study is to establish the extent and mechanism of protein modification using human erythrocytes as target cells. The results discussed here show that peroxyl radicals promote proteolysis before the onset of hemolysis or lipid peroxidation. METtlODS Blood was collected from human volunteers by venepuncture in the presence ofheparine and then centrifuged to remove plasma and huffy coats. Erythrocytes were washed three times with ice-cold Krebs-HEPES, ph 7.4 (mmoles/l : NaCI, 120; KCI, 4.8; Ca CIz, 1.3; MgSO4, t.2; HEPES, 16) and then resuspended in the same buffer to a 16% hematocryt. Protein oxidation. For the oxidation experiments, 1 volume of erythrocyte suspension was incubated with an equal volume of 150 mm AAPH (Wako) at 37~ with gentle stirring in the dark and for varying lengths of time (7). Control experiments were performed by incubating erythrocytes in the absence of AAPH. The cell pellet was washed twice with ice-cold Krebs- HEPES, ph 7.4 and resuspended in the same buffer for further experiments. Lipid oxidation. Thiobarbituric acid-reactive substances (TBARS) were determined by the method described by Trotta et al. (16). Erythrocyte hemolysis. Percent hemolysis was assessed by measuring hemoglobin release from cells, relative to the total cellular hemoglobin content. Following incubation, the cells were diluted 1:40 with ice-cold 0.15 M NaC1, centrifuged and oxy- plus met-hemoglobin concentration of the supernatant was determined spectrophotometrically as in Lips et al. (17). Total hemolysis was achieved by replacing the 0.15 M NaCI solution by distilled water (7). Protein degradation. The hydrolysis of cell proteins was followed by measuring the appearance of free amino groups assayed with fluorescamine (Sigma). To 0.5 ml of erythrocyte suspension, 1.0 ml of cold 0.8 M perchloric acid was added with vortexing. After 4 h at 4~ the suspension was centrifuged for 15 min at 1000 x g. Next, 1.0 ml of supernatant was brought to ph 10 with 0.2 ml of KOH and buffered by the addition of 0.8 ml of 0.5 M HEPES, ph 10. After 10 min at 4~ supernatant amino groups were measured by diluting 0.4 ml aliquots with 2.6 ml of 50 mm phosphate buffer ph 10 and, with vortexing, 1 ml of 0.3 mg/ml fluorescamine solution in dioxane. The fluorescence of the sample was measured at 390 nm excitation and 475 nm emission in comparison with standards (9). 1122

3 Membrane isolation. Membranes of oxidized or control e rythrocytes were obtained by cell hemolysis with the addition of a 15-fold volume of ice-cold 1 mm EDTA, 10 mm Tris, ph 7.4 buffer. The membranes were washed several times with 10 mm Tris, ph 7.4 by centrifugation (23000 x g, 12 min, 0~ until the supernatant buffer became colorless. Cell membranes were resuspended in the same buffer and stored at -70~ Membrane proteins were assessed according to Peterson (18). (}el electrophoresis. Sodium dodecyl sulfate-polyacrylamide electrophoresis (SDS-PAGE) was performed in the discontinuous buffer system of Laemmli (19). Protein bands were visualized by stainning with Coomassie Brilliant Blue R-250. Densitometric patterns of stained SDS-PAGE gels were obtained with a Genius scanner and analyzed according to Bozzo el al, (20). Statistical analysis. Wilcoxon test was applied to the obtained data, Data are expressed as mean value ~-standard deviation. RESULTS AND DISCUSSION Incubation of erythrocytes with AAPH showed an increase in fluorescamine-reactive substances in the free-cell soluble acid fraction. This increase can be used as a marker of proteolysis (Fig 1). Protein modification upon exposure to AAPH is due to the action of oxygenated flee radicals, since the same results were obtained in oxygen saturated solutions (21). It is known that oxidatively modified proteins are more susceptible to proteinases (10-13). Therefore, AAPH induced proteolysis can be explained by the direct action of free radicals and/or by the action of proteinases, a process known to be important for the removal of inactive oxidized proteins (22). Proteolysis occurred at times preceding hemolysis (Fig 1) (p<0.001, n=5 at 180 min.) and significantly affects proteins associated at or spanning the lipid bilayer, since 40% of the total increase in fluorescamine-reactive substances appears in the erythrocyte suspension supernatant. This is probably due to the fact that peroxyl radicals derived from AAPH were either unable to gain access to the cytosolic face of the erythrocyte membrane or were scavenged by endogenous antioxidants. In agreement with these considerations, hemoglobin oxidation was below 2% and erythrocyte glutathione concentration was not affected by exposure to AAPH (data not shown), suggesting a lack of accessibility of AAPH into the intact cells, as proposed by Sandhu et al. (23). Cell membranes isolated from the AAPH-oxidized erythrocytes were subjected to SDS- PAGE in the presence and absence of a reducing agent. Under reductive conditions (Fig. 2A), gel staining showed that band 3 protein from the oxidized cells was significantly decreased when compared with that from untreated cells (Fig. 2B). It was found that band 3 protein was the most susceptible to degradation. Densitometric measurements of protein bands in the stained gel of AAPH treated eryhtroeytes membranes indicated that band 3 protein staining intensity was 1123

4 VOI. 43, NO. 5, 1997 BIOCHEMISTRYend MOLECULAR BIOLOGY INTERNATIONAL w _I 300 x ul _= 200 o= 100 s =,o o Time (rnin) I-- Figure I. Time course of increase in hemolysis, protein degradation and lipid oxidation of human erythrocytes (Hto: 8 %) exposed to 75 mm AAPH. (,,x ) % oflysis (n=8), ( o ) NH2 (n=6), ( 9 ) TBARS (n=7). Control values after 300 min incubation are (n=6), 65.7 _ (n=5) and 0.06 i 0.05 (n=4) respectively. importantly reduced, as compared to other main membrane proteins. The observed decrease was 27% (p<0.05, n=5) at 180 min. On the other hand, spectrin was not modified, and an increase of low-molecular weight peptides was not observed (Fig 2A and 2B). On the contrary, our results showed a decrease in the amount of low-molecular weight peptides (lower than 45 kd), in part due to the membrane detachment of glyceraldehyde-3-phosphate dehydrogenase (band 6) which in the erythrocyte is associated to band 3. A modified band 3 will no longer sustain an interaction with band 6 and, therefore, this protein is translocated to the cytoplasm (24). It must be beared in mind that membrane electrophoresis only detects membrane-associated peptides and therefore the increase in fluorescamine-reactive substances in the free-cell acid fraction (Fig. 1) indicates the presence of small peptides not detected in electrophoresis. To test the involvement of proteolytic enzymes, a serine proteinase inhibitor, PMSF (lmm) and a metalloproteinase inhibitor, EDTA (1.5 mm) were used. When they were employed together, a 78% inhibition (n=2) in the increment of-nh2 groups in cell-free extract was found after 180 minutes of exposure. It has been described that erythrocytes exposed to oxygen radicals show a preferential degradation of band 3 (though spectrin is also oxydized ) by membrane-bound proteinases, along 1124

5 Vol. 43, No. 5, ] 997 o ) 'q' C~ r" _ -.~,", ~ ~.~..= ~ -,~ o "~ U'.q ~.~ ~.~. 2 ~'~-~ ~ < ~ i- z _. 7,. o 1125

6 with a mild protein oxidation (14). Protein degradation prior to significant hemolysis (at 180 rain incubation) is extensive. If it is assumed that it mostly arises from band 3 degradation, and considering that nearly 25% of the protein is lost, a simple calculation shows that several hundred amino groups are produced by each degraded protein molecule. This extensive degradation would produce small peptides not detectable by gel electrophoresis, explaining the lack of increase in low molecular weight fragments associated to the protein modification (Fig. 2). Total rate of free radical production can be obtained from the AAPH concentration and its thermolysis rate (6). Comparison of this value with the rate of free amino terminal groups increase indicates that for every three radicals introduced into the system nearly one amino group is generated. This high efficiency contrasts with that obtained for TBARS molecules production (ca. one TBARS molecule per 103 radicals). The difference mainly results from the large number of amino groups produced per oxidized molecule which is substrate for proteinases. On the other hand, only a small f?action of the oxidized lipids leads to TBARS type products. The data in Fig. 1 show a decrease in the rate of terminal amino group formation at incubation times longer than 180 rain. Similarly, at this long incubation time the rate of decrease in band 3 protein becomes almost negligible (data not shown). This results could be adscribed to inactivation of the proteolytic mechanisms due to their extensive exposure to the free radicals (10-13). The data obtained in the present work show extensive protein degradation prior to a significant increase in other indexes of red cell damage. This process is mostly due to the action of red cell proteinases since the extent of degradation due to direct free radical processes is less than ca. 20% of the total observed. Acknowledgements. This work was supported by grants FONDECYT and The technical assistance of Miss Carlina Tapia is gratefully acknowledged. REFERENCES 1. B. Halliwell and M.C. Gutteridge (1990) Methods Enzymol. 186, D.M. Richards, R.T. Dean and W. Jessup (1988) Biochim. Biophys. Acta, 946: RE Pacifici and K.J.A. Davies (1990) Methods Enzymol. 186, HR. Petty, M. Zhou, and Z. Zheng (1991) Biochim Biophys. Acta r064, J. Van der Zee, J. van Steveninck, JK. Koster and T.M.AR. Dubbelman (1989) Biochim. Biophys. Acta 980, E. Niki, Y, Yamamoto, E.~ Komuro and K. Sato (1991) J. Clin. Nutr. 53~ 201 S-204S. 1126

7 7. M. Miki, H. Tamai, M. Mino, Y. Yamamoto and E. Niki (1987) Arch. Biochem. Biophys. 258, Y. Sato, S. Kano, T.Takahashi and Y. Suzuki (1995) Biochemistry, 34, K.JA. Davies and A.L. Goldberg (1987) J. Biol. Chem. 262, KJ.A. Davies (1993) Biochem. Soc. Trans. 21, T. Grune, T. Reinheckel, M. Joshi and K.J.A. Davies (1995) J. Biol. Chem. 270, C. Giulivi, R.E Paeifici and K,J.A. Davies (1994) Arch. Biochem, Biophys. 311, C. Giulivi and K.J.A. Davies (1993) J. Biol. Chem. 268, M. Beppu, M. Imoue, T. Ishikawa and K. Kikugawa (1994) Biochim. Biophys, Acta 1196, D.E Croach and G.N. Demartino (1991) Physiol. Rev.71, R.J. Trotta, S.G. Sullivan and A. Stern (1982) Biochem. J. 204, V. Lips, G. Celedon, J. Escobar and E.A. Lissi (1996) Redox Report 2, G.L. Peterson (1977) Anal. Biochem. 83, U.K. Laemmli (1970) Nature 227, S. Bozzo and C. Retamal (1991) Arch. Biol. Med. Exp. 24, R EA. Lissi and N. Clavero (1990) Free Rad. Res. Comm. I0, 177-t K.J.A. Davies (1986) Free Rad. Biol. Med. 2, I.S. Sandhu, K. Ware and M.W. Grisham (1992) Free Rad. Res. Comm. 16, C. MaUozi, A.M.M. Di Stasi and M. Minetti (1995) Arch. Biochem. Biophys. 321,