- Email: [email protected]
Oxidation of hemoglobin to methemoglobin in intact erythrocyte by a hydroperoxide induces formation of glutathionyl hemoglobin and binding of α-hemoglobin to membrane
ABB Archives of Biochemistry and Biophysics 417 (2003) 244–250 www.elsevier.com/locate/yabbi
Oxidation of hemoglobin to methemoglobin in intact erythrocyte by a hydroperoxide induces formation of glutathionyl hemoglobin and binding of a-hemoglobin to membrane Kaori Murakami and Shiro Mawatari* Department of Nutrition and Health Science, Faculty of Human Environmental Science, Fukuoka WomenÕs University, 1-1-1 Kasumigaoka, Higashi-ku, Fukuoka, Japan Received 6 June 2003, and in revised form 22 July 2003
Abstract Biochemical consequences of oxidation of hemoglobin (Hb) in intact human erythrocytes were studied. The incubation of washed erythrocyte with 1 mM tert-butylhydroperoxide induced an increase in glutathionyl-Hb (G-Hb). The formation of G-Hb occurred linearly until 10 min in parallel with the formation of methemoglobin (metHb) after exhaustion of reduced glutathione. The results show that metHb, but not normal Hb, reacts with oxidized glutathione to form G-Hb. G-Hb was confirmed by immunoblotting with anti-glutathione antibody and the formation of G-Hb was accompanied by parallel decrease in b-globin detected with a reversed phase HPLC. Electrophoretic studies showed time-dependent increase in membrane-associated a-Hb until 10 min, indicating that a part of unpaired a-Hb bound to the membrane. Pre-b-globin increased despite the decrease in b-globin and a part of the increase was independent of the decrease in b-globin. Pre-b-globin reacted with anti-glutathione antibody, but it differs from G-Hb in many features. Ó 2003 Elsevier Inc. All rights reserved. Keywords: Intact erythrocytes; Peroxidation; Hydroperoxide; Hemoglobin; Methemoglobin; a-Hemoglobin; b-Hemoglobin; Pre-b-globin; Glutathionyl hemoglobin
Formation of methemoglobin (metHb)1 in intact erythrocytes causes dysfunction of erythrocytes including peroxidative changes in erythrocyte membranes [1,2] and is implicated in the pathology of unstable hemoglobin diseases [3–8]. Biological functions of nitric oxide may also relate to oxidation of Hb in intact erythrocytes [9–12]. However, biochemical consequences of formation of metHb in intact erythrocytes do not seem to be fully understood. tert-Butylhydroperoxide (tBHP), a cell membrane permeable hydroperoxide, is known to induce peroxi* Corresponding author. Fax: +92-671-7843. E-mail address: [email protected] (S. Mawatari). 1 Abbreviations used: G-Hb, glutathionyl-Hb; metHb, methemoglobin; tBHP, tert-Butylhydroperoxide; a-Toc, a-tocopherol; PL, phospholipids; HPLC, high-performance liquid chromatography; PBS, phosphate-buffered saline; DTNB, 5,50 -dithiobis(2-benzoic acid); SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; PVDF, polyvinylidene difluoride.
0003-9861/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0003-9861(03)00389-8
dative changes in intact human erythrocytes including peroxidation of Hb [13–17]. Our previous study showed that, by the incubation of erythrocytes with 1 mM tBHP, reduced glutathione (GSH) of the erythrocytes exhausted within 1 min followed by the changes in membrane a-tocopherol (a-Toc) and phospholipids (PL). a-Toc and PL decreased maximally within a few minutes, whereas the formation of metHb occurred linearly until 15 min [17]. Dithiothreitol (DTT), a sulfhydryl reductant, protected the erythrocytes completely against the formation of metHb by tBHP, indicating the importance of sulfhydryls in the erythrocytes [17]. Actually, the erythrocytes contain a large amount of GSH, which protects the erythrocytes from peroxidative changes. During the experiment of peroxidation of the intact human erythrocytes with tBHP, we observed time dependent decrease in b-globin until 10 min by a reversed phase high-performance liquid chromatography (HPLC). An ion exchange HPLC of Hb revealed that
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
the decrease in b-globin was associated with a parallel increase in glutathionyl-Hb (G-Hb). Electrophoretic studies showed time dependent increase in membranebound a-Hb until 10 min. These changes in Hb occurred in parallel with formation of metHb. Pre-b-globin increased despite the decrease of b-globin. Both G-Hb and pre-b-globin reacted with anti-glutathione antibody, however, pre-b-globin differed from G-Hb in many features. These biochemical consequences of peroxidation of Hb in intact erythrocytes by a hydroperoxide have not been reported hitherto, and the results indicate that metHb, but not normal Hb, reacts with oxidized glutathione (GSSG) to form G-Hb.
245
Measurement of metHb A portion (0.4 ml) of the erythrocyte suspension was hemolyzed with 5 ml of solution containing 100 mM phosphate buffer (pH 6.8) and 1% Triton X-100 (4:6, v/v) and the hemolysate was divided into two parts (2 ml each). Absorbance of the first part was read at 630 nm and then read again at 630 nm after addition of 0.02 ml of 5% potassium cyanide. Absorbance of the second part was read at 630 nm with 0.02 ml of 5% potassium ferricyanide and then read again at 630 nm after addition of 0.02 ml of 5% potassium cyanide [18]. The spectrophotometer used was Shimadzu BioSpec-1600 (Shimadzu, Kyoto, Japan). Measurement of erythrocyte GSH
Materials and methods
GSH, GSSG, tBHP, and 5,50 -dithiobis(2-benzoic acid) (DTNB) were obtained from Wako Pure Chem., Osaka, Japan. Rabbit anti-glutathione polyclonal antibody (Chemicon International Temecula, Ca) was purchased from Cosmo Bio, Tokyo, Japan. Other reagent grade chemicals and HPLC grade solvents were purchased from Wako Pure Chem., Osaka, Japan.
Erythrocyte GSH was measured by the recommended method by International Committee for Standardization in Haematology [19]. Briefly, 1 ml of the erythrocyte suspension was hemolyzed with 1 ml of water and then 3 ml of metaphosphoric acid solution, composed of 1.67 g of metaphosphoric acid, 0.2 g of EDTA-2Na, and 30 g of NaCl in 100 ml of water, was added. After filtration with a filter paper (No. 5B, Advantec, Tokyo, Japan), the filtrate (1 ml) was subjected to measurement of GSH using DTNB.
Preparation of erythrocytes
Preparation of Hb and erythrocyte membranes
Human venous blood from healthy adult volunteers who had fasted overnight, under informed consent, was drawn into tubes containing EDTA-2Na and processed immediately. Serum and buffy coats were removed after centrifugation at 1000g for 5 min, and the erythrocytes were washed three times in cold phosphate-buffered saline (PBS) at 1000g for 5 min. A small portion of the top layer was removed at each washing. The washed erythrocytes were re-suspended in PBS so that the final concentration of the erythrocytes was a hematocrit (Hct) of 10%. Hct was measured by using an automated blood counter (Sysmex F-520, Kobe, Japan). PBS, pH 7.4, was composed of 138 mM NaCl, 5 mM KCl, 6.1 mM Na2 HPO4 , 1.4 mM NaHPO4 , 1 mM MgCl2 , and 5 mM glucose.
The packed erythrocytes from 3 ml of the erythrocyte suspension were hemolyzed by 1.5 ml of cooled water and centrifuged at 20,000g for 15 min at 4 °C. To 0.9 ml of the supernatant, 0.3 ml of CCl4 was added [20]. The Hb solution was stored at )80 °C until use. Erythrocyte membranes (the sediment of 20,000g centrifugation after hemolyzation) were washed with 8 ml of 10 mM Tris– HCl, pH 7.4, by centrifugation at 20,000g for 15 min at 4 °C. The washing was repeated three or four times until the supernatant became clear. The membranes were stored at )80 °C until use.
Materials
Peroxidation of erythrocytes The total volume of the reaction mixture was brought up to 5 ml by the addition of PBS and the final concentration of the erythrocytes was 5% Hct. The final concentration of tBHP was 1 mM. After pre-incubation at 37 °C for 5 min, the reactions were started by addition of tBHP. The reactions were terminated putting the tubes in an ice bath. The erythrocytes were pelleted by centrifugation at 1000g for 5 min at 4 °C and washed twice with PBS. The erythrocytes were re-suspended in 5 ml PBS.
Separation of Hb chains Separation of hemoglobin chains (Hb-chains) was done using the reversed phase HPLC method [20,21]. The HPLC system consisted of a Hewlett Packard HP 1100 series (Yokogawa Analytic. Systems, Tokyo, Japan) equipped with a four solvent delivery system, a degassing unit, an automatic injector, a visual-ultraviolet detector, and a HP-Vectra computer with HPLC ChemStation software. The column was a Vydac C4, 250 4.6 mm (Agilent Technology, Tokyo, Japan). Mobile phase A was 20% acetonitrile in water containing 0.1% trifluoroacetic acid (TFA), and mobile phase B was 60% acetonitrile in water containing 0.1% TFA. Gradient elution was done with 52% B at 0 min, 59% B at 50 min, 92% B at 51 min,
246
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
92% B at 55 min, and 52% B at 60 min. The turnaround time was 70 min. The flow rate was 1 ml/min and the column temperature was 25 °C. Globins were detected at 220 nm. Some globin chain fractions were collected and concentrated under a stream of N2 gas followed by centrifugal ultrafiltration method using Centricon YM-10 (Millipore, Bedford, MA, USA). Determination of glutathionyl-Hb Separation of G-Hb was done with an ion-exchange HPLC [21,22]. The column was a TSK-GEL CM-2SW, 4.6 250 mm (Tosoh, Tokyo, Japan). Mobile phase A consisted of 30 mM Bis–Tris, 1.5 mM potassium cyanide, and 150 mM sodium acetate, pH 6.40, and mobile phase B was 30 mM Bis–Tris and 1.5 mM potassium cyanide, pH 6.4. The mobile phase A was increased linearly from 40 to 95% in 40 min. The flow rate was 0.8 ml/min and the column temperature was 30 °C. Hb was detected at 415 nm. For identification of G-Hb, GHb fraction was collected from the HPLC, concentrated by the centrifugal ultrafiltration method using Centriplus YM-10 (Millipore), and subjected to immunoblotting study with anti-glutathione antibody. Some of the Hb of the erythrocytes incubated with and without tBHP were incubated further with GSH or GSSG for 60 min at 37 °C. Immunoblotting of G-Hb and pre-b-globin The concentrated G-Hb and pre-b-globin were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) with 14% gel in which 2mercaptoethanol was omitted from the sample buffer and electrically blotted on polyvinylidene difluoride (PVDF) membrane. PVDF membrane was blocked in PBS containing 5% skimmed milk and 0.1% Tween 20 and subsequently incubated overnight with anti-glutathione antibody at 5 °C. After washing, anti-rabbit IgG linked to horseradish peroxidase (Amersham Bioscience, Tokyo, Japan) was added and incubated for 1 h. Detection was performed using ECL Plus (Amersham Bioscience) according to the manufacturerÕs instructions.
used [19]. The gels were cast in a slab gel cell (AE-6401 M, ATTO, Tokyo, Japan), with gel size being 80 mm long, 90 mm wide, and 1 mm thick. One part by volume of the erythrocyte membrane solution (protein concentration was approximately 2 mg/ml) was dissolved in three parts by volume of sample buffer. The sample buffer consisted of 4 ml of 8 M urea, 0.4 ml of acetic acid, 0.4 ml of 2-mercaptoethanol, and 0.04 ml Triton X-100. Electrophoresis was done at 10 mA/gel for 4 h and Hb chains were stained with Coomassie brilliant blue.
Results Formation of metHb and no hemolysis The incubation of the erythrocytes with 1 mM tBHP induced a lineal increase in metHb until 10 min (Fig. 1). The control incubation without tBHP for 15 min showed no formation of metHb. None of the supernatants in the reaction mixtures after the incubation of the erythrocytes with tBHP showed coloration by monitoring at 540 nm, which indicated that no hemolysis had occurred under any of the conditions created in the present study. Rapid exhaustion of erythrocyte GSH GSH levels in the control reaction mixture containing 5% erythrocytes from six healthy adult volunteers were 90–160 lM. GSH was exhausted within 1 min after the incubation with 1 mM tBHP (Fig. 1).
Electrophoresis of erythrocyte membrane proteins The membranes were heated at 100 °C for 3 min in the sample buffer containing 0.125 M Tris–HCl, pH, 6.8, 10% sucrose, 4% SDS, and 10% of 2-mercaptoethanol, and an aliquot of the membrane solution was applied to SDS–PAGE with 12% gel. The protein bands were stained with Coomassie brilliant blue. To identify the Hb chains bound to the membranes, Triton acid urea gel electrophoresis was applied to the membrane [23,24]. Polyacrylamide gel (12%) containing 6 M urea and 2% Triton X-100 in 5% acetic acid was
Fig. 1. Time courses of the formation of metHb and the change in GSH after the incubation of erythrocytes with 1 mM tBHP. MetHb increased almost linearly until 10 min. GSH was exhausted within 1 min and began to recover after 10 min. Data are means SD of triplicate determinations of the erythrocytes from the same volunteer.
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
247
Decrease in b-globin and formation of G-Hb Each globin was clearly separated by the reversed phase HPLC within 50 min (Fig. 2). b-Globin decreased linearly from a 46% to a 35% per total globin (sum of pre-b-, b-, a-, and other minor globins) during the initial 10 min of the incubation of the erythrocytes with 1 mM tBHP (Figs. 2 and 3). The decrease in b-globin was about 10% of the initial level (Fig. 3). The relative composition of pre-b-globin increased despite the decrease in b-globin (Fig. 3), and a part of the increase of pre-b-globin was independent of the decrease in b-globin (Figs. 2 and 3). The ion exchange HPLC of Hb revealed an increase in G-Hb after the incubation of erythrocyte with tBHP (Fig. 4A). The increase in G-Hb was almost linear until 10 min (Fig. 6) and was associated with parallel decrease Fig. 3. Time courses of the changes in each Hb chain after the incubation with 1 mM tBHP. Hb chains were separated by a reversed phase HPLC as shown in Fig. 2. The increase in a-globin was relative to the decrease in b-globin; however, a part of the increase in pre-b-globin was independent of the decrease in b-globin. Each data point represents the mean SD of the erythrocytes from three different volunteers.
in HbA0 (data not shown). The amount of G-Hb at 10 min was about 9% of total Hb peaks (sum of HbA1c , G-Hb, HbA1d HbA0 , and HbA2 ) (Fig. 6). The peak of G-Hb decreased by the further incubation of Hb with 1 mM DTT for 60 min (Fig. 5). The Hb of the erythrocytes incubated without tBHP formed only trace amount of G-Hb by the further incubation with 1 mM GSSG (Fig. 4B-1). On the other hand, the Hb of the erythrocytes incubated with tBHP showed marked increase in G-Hb by the further incubation with 1 mM GSSG (Fig. 4B-2), but the Hb did not increase G-Hb by the incubation with GSH (data are not shown). SDS–PAGE of the concentrated G-Hb peak showed about 16 kDa protein and the electrophoretic mobility of G-Hb was almost same as that of b-globin (Fig. 7A). Immunoblotting study using anti-glutathione antibody confirmed the presence of glutathione in G-Hb (Fig. 7B). Pre-b-globin also reacted with anti-glutathione antibody (Fig. 7B).
Fig. 2. Representative HPLC chromatograms of Hb chains after the incubation of the erythrocytes with tBHP. (A) The incubation of the erythrocytes without tBHP. (B and C) b-Globin decreased time dependently after the incubation of the erythrocytes with tBHP for 5 min and 10 min, respectively. The chromatograms also show an increase in pre-b-globin at 5 min after the incubation with tBHP (B), which is independent of the decrease in b-globin. All of the chromatograms show the results from the same batch of erythrocytes for which incubations were done simultaneously. Heme, heme; a-gl, a-globin; b-gl, b-globin; and Pre-b-gl, and pre-b-globin.
Time dependent increase in binding of a-Hb to erythrocyte membranes The results of SDS–PAGE of the erythrocyte membranes show a time dependent increase in monomeric Hb until 10 min by the incubation of the erythrocytes with tBHP (Fig. 8A). The Triton acid urea gel electrophoresis of the membranes at 10 min incubation of the erythrocyte with 1 mM tBHP indicates that most of the Hb-chains bound to the membrane are a-Hb (Fig. 8B).
248
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
Fig. 4. Representative HPLC chromatograms of hemoglobin. (A-1) Control incubation of the erythrocyte without tBHP showing only trace of glutathionyl-Hb (G-Hb). (A-2) G-Hb increased after the incubation of erythrocyte with tBHP for 10 min. (B-1) The Hb of the erythrocytes of the control incubation formed only a little amount of G-Hb by the further incubation with 1 mM GSSG for 60 min (compare A-1 and B-1). (B-2) The Hb of the erythrocytes incubated with tBHP markedly increased G-Hb by the further incubation with 1 mM GSSG for 60 min (compare A-2 and B-2). All of the chromatograms show the results from the same batch of erythrocytes.
Discussion It is reported that HbA can be reacted with glutathione with disulfide bond formation between the cysteine 93 of b-Hb and the cysteine of glutathione [25,26] and that G-Hb is not present in significant amounts in normal erythrocytes because GSSG, which could form a mixed disulfide with Hb, is present at a very low con-
Fig. 5. G-Hb of the erythrocytes after the incubation with tBHP for 10 min (A) decreased by the further incubation with 1 mM DTT for 60 min (B).
centration [25]. In this study, the erythrocyte GSH was exhausted within 1 min by the incubation of intact erythrocytes with tBHP indicating rapid increase of GSSG (Fig. 1), but the formation of G-Hb occurred linearly until 10 min in parallel with the formation of metHb (Figs. 1 and 6). The results indicate that metHb, but not normal Hb (oxy-Hb and deoxy-Hb), reacts with GSSG to form G-Hb. Actually, the Hb of the erythrocytes incubated without tBHP formed only trace amount of
Fig. 6. Time dependent increase in glutathionyl-Hb (G-Hb) by the incubation of the erythrocytes with tBHP. G-Hb was calculated on the total Hb peaks (sum of HbA1c , G-Hb, HbA1d , HbA0 , and HbA2 ). Each data point represents the mean of two determinations from two different erythrocytes taken from two volunteers.
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
G-Hb by the further incubation with 1 mM GSSG for 60 min (Fig. 4B-1), whereas the Hb of the erythrocytes incubated with tBHP showed marked increase in G-Hb by the further incubation with GSSG (Fig. 4B-2). The G-Hb peak decreased by the incubation of the Hb with 1 mM DTT, indicating that metHb and GSSG formed disulfide bond (Fig. 5). The decrease in b-globin detected by the reversed phase HPLC is almost symmetrical to the increase in G-Hb detected by the ion exchange HPLC, indicating that the decrease in b-globin was due to the formation of G-Hb (Figs. 3 and 6). The results also indicate that disulfide bonded b-globin with glutathione is not detectable by the reversed phase HPLC method using 0.1% TFA, which has commonly been used for the separation of Hb chains [20,21].
249
The detection of G-Hb by a cation exchange HPLC was recently reported [22]. The mobile phases of HPLC used in this study were different from those in the reported HPLC; however, the appearance of G-Hb on the chromatograph was identical. We collected the G-Hb peak from the HPLC and confirmed for the first time that G-Hb reacts with anti-glutathione antibody (Fig. 7B). Recently, an increase in G-Hb has been reported in patients with diabetes mellitus, hyperlipidemia, uremia, and FriedreichÕs ataxia [31–34], and G-Hb is suggested to be a possible clinical marker of oxidative stress [35]. This study clearly shows that metHb in intact erythrocytes induces formation of G-Hb by reacting with GSSG. In human oxy-Hb, a-Hb is oxidized more rapidly than b-Hb [27,28]. Nitric oxide (NO) preferentially
Fig. 7. (A) SDS–PAGE of glutathionyl-Hb, pre-b-globin, and b-globin stained with Coomassie brilliant blue, indicating that the molecular weight of the three proteins is almost the same. Fourteen percent gel was used. 2-Mercaptoethanol was omitted from the sample buffer. (B) Immunoblotting study of the proteins separated by SDS–PAGE shows that glutathionyl-Hb and pre-b-globin react with anti-glutahione antibody. Each protein was collected repeatedly from HPLC of Hb from three different volunteers and concentrated. For details, see Materials and methods. G-Hb, glutathionylHb; Pre-b-gl, pre-b-globin; and b-gl, and b-globin.
Fig. 8. (A) SDS–PAGE of the membrane proteins shows time dependent increase in membrane associated Hb monomer until 10 min by the incubation of erythrocytes with tBHP. Twelve percent gel was used. (B) Triton acid urea gel electrophoresis of the membrane after the incubation with tBHP shows preferential binding of a-Hb to the membrane. These results were confirmed with erythrocytes from three different volunteers. *, Arrow indicates dimeric hemoglobin; **, arrow indicates monomeric hemoglobin; St, standard proteins; Cont Hb, the Hb of the erythrocytes of the control incubation; tBHP, the membranes of the erythrocytes incubated with tBHP for 10 min; b-Hb, b-hemoglobin; and a-Hb, a-hemoglobin. The numbers below lanes indicate the incubation time (min).
250
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
binds to a-Hb in a low affinity tense state (T-state) as in deoxy-Hb, and NO binds preferentially to cysteine 93 on b-Hb in a relaxed state (R-state) as in oxy-Hb [9,10]. To our knowledge, it has not been reported hitherto that coexistence of met-Hb and GSSG in intact erythrocytes induces formation of G-Hb. Both G-Hb and S-nitrosohemoglobin are bonded to cysteine 93 of b-Hb and both substances are reported to increase the oxygen affinity [25,29,30]. It is suggested that transnitrosation of S-nitrosohemoglobin to S-nitrosoglutathione relates to regulation of blood flow [9]; therefore, it may be possible that G-Hb has also some physiological roles. Possible relationship between G-Hb and S-nitrosohemoglobin is unclear. The reversed phase HPLC of globin chains showed an apparent increase in pre-b-globin despite the decrease in b-globin (Fig. 2). The increase in pre-b-globin was faster than the decrease in b-globin and stopped at 5 min (Fig. 4). Thus, a part of the increase in pre-b-globin was independent of the decrease in b-globin. The nature of pre-b-globin does not seem to be well defined yet, but it is reported that the amino acid composition of pre-b-globin is indistinguishable to that of b-globin [36]. It is thought that pre-b-globin consists of a posttranslationally modified protein, presumably containing a b-globin–glutathione adduct [36]. This study shows for the first time that pre-b-globin reacts with anti-glutathione antibody (Fig. 7B). It is of interest to note that both pre-b-globin and G-Hb react with anti-glutathione antibody. However, the time course of the changes in pre-b-globin during the oxidation of Hb by the hydroperoxide was different from that of G-Hb (Fig. 4), and the present study shows that b-globin in G-Hb, in contrast to pre-b-globin, is not detectable by the reversed phase HPLC. This study indicates that unpaired a-Hb increased in parallel with the formation of G-Hb, and the electrophoretic studies of the membranes showed time dependent increase of the membrane associated a-Hb (Figs. 8A and B). This situation is similar to b-thalassemias, where lack of sufficient b-Hb leads to membrane associated a-Hb [7,8]. However, the membrane bound a-Hb is indicated to be only a small part of the unpaired a-Hb, so that the total a-globin chain is not significantly changed (Figs. 2 and 3). Acknowledgments This work was supported in part by a grant for scientific research from Fukuoka Prefecture, Japan, and by funding from the Institute of Rheological Functions of Food, Fukuoka, Japan. References [1] L.M. Snyder, N.L. Fortier, J. Trainor, J. Jacob, B. Lubin, D. Chiu, S. Sohet, N. Mohandas, J. Clin. Invest. 76 (1985) 1971–1977.
[2] P. Jarolim, M. Lahave, S.C. Liu, J. Palek, Blood 76 (1990) 2125–2131. [3] T.P. Flynn, D.W. Allen, G.J. Johnson, J.G. White, J. Clin. Invest. 71 (1985) 1215–1223. [4] S.-C. Liu, J.Y. Scott, J.R. Mehta, P.E. Nichols, S.K. Balias, P.W. Yacono, D.E. Golan, J. Palek, J. Clin. Invest. 97 (1997) 29–36. [5] B.R. Rank, J. Carisson, R.P. Hebbel, J. Clin. Invest. 75 (1985) 1531–1537. [6] R. Advani, S. Sorenson, E. Shinar, W. Lande, E.A. Rachimilewitz, S.L. Shrier, Blood 79 (1992) 1058–1063. [7] M.D. Scott, J.J.M. van den Berg, T. Repka, J. Clin. Invest. 91 (1993) 1706–1712. [8] R.E. Schrier, N. Mohandas, Blood 74 (1989) 2194–2202. [9] L. Jia, C. Bonaventura, J. Bonaventura, J.S. Stamler, Nature 380 (1996) 221–226. [10] H. Kosaka, Biochim. Biophys. Acta 1411 (1999) 370–377. [11] T.H. Han, D.R. Hyduke, M.W. Vaughn, J.M. Fukuto, J.C. Liao, Proc. Nat. Acad. Sci. USA 99 (2002) 7763–7768. [12] C. Bonavenrura, C.H. Taboy, P.S. Low, R.D. Stevens, C. Lafonn, A.L. Crumbliss, J. Biol. Chem. 277 (2002) 14557–14563. [13] R.J. Trotta, S.G. Sullivan, A. Stern, Biochem. J. 212 (1983) 759–772. [14] G.A. Dise, D.B.P. Goodman, Biochim. Biophys. Acta 859 (1986) 23–69. [15] B. Deuticke, K.B. Heller, C.W.M. Haest, Biochim. Biophys. Acta 899 (1987) 113–124. [16] T.T. Rohn, T.R. Hinds, F.F. Vincenczi, Biochim. Biophys. Acta 1153 (1993) 67–76. [17] S. Mawatari, K. Murakami, Lipids 36 (2001) 57–65. [18] K.A. Evelyn, H.T. Malloy, J. Biol. Chem. 126 (1938) 655–662. [19] International Committee for Standardization in Haematology, Recommended methods for additional red cell enzyme (pyridine 50 -nucleotidae) assay and the determination of red cell adenosine50 -triphosphate, 2,3-diphosphoglycerate and reduced glutathione, Clin. Lab. Haematol. 11 (1989) 131–138. [20] B. Masala, L. Manca, Methods Enzymol. 231 (1994) 21–44. [21] T.H.J. Huisman, J. Chromatogr. 418 (1987) 277–304. [22] A. Pastore, A.F. Mozzi, G. Tozzi, L.M. Gaeta, G. Federici, E. Bercini, A.L. Russo, L. Mannucci, F. Piemonte, Anal. Biochem. 312 (2003) 85–90. [23] B.P. Alter, S.C. Goff, G.D. Eferemov, M.E. Gravely, T.H.J. Huisman, Brit. J. Haematol. 44 (1980) 527–534. [24] S. Sorensen, E. Rubin, H. Polster, N. Mohandas, S. Schrier, Blood 75 (1990) 1333–1336. [25] M.-C. Carel, C. Domenget, J. Caburi-Martin, C. Prehu, F. Galacteros, Y. Beuzard, J. Biol. Chem. 261 (1986) 14704–14709. [26] C.T. Craescu, G. Poyart, C. Schaeffer, M.-C. Carel, J. Kister, Y. Beuzard, J. Biol. Chem. 261 (1986) 14710–14716. [27] A. Mansouri, K.H. Winterhalter, Biochemistry 12 (1973) 4949– 4976. [28] M. Tsuruga, A. Matsuoka, A. Hachimori, Y. Sugawara, K. Shikama, J. Biol. Chem. 273 (1998) 8607–8615. [29] R.P. Patel, N. Hog, N.Y. Spencer, B. Kalynaraman, S. Matalon, V.M. Darley-Usmar, J. Biol. Chem. 274 (1999) 15487–15492. [30] C. Bonaventura, G. Ferruzzi, S. Tesh, R.D. Steven, J. Biol. Chem. 274 (1999) 24742–24748. [31] T. Niwa, C. Naito, A.H. Mawjood, K. Imai, Clin. Chem. 46 (2000) 145–146. [32] C. Naito, T. Niwa, J. Chromatogr. B. 746 (2000) 91–94. [33] F. Takayama, S. Tsutsui, M. Horie, M. Shimokata, T. Niwa, Kidney Int. Suppl. 78 (2001) S155–S158. [34] F. Piemonte, A. Pastore, G. Tozzi, Eur. J. Clin. Invest. 31 (2001) 1007–1011. [35] S.E. Bursell, G.L. King, Clin. Chem. 46 (2000) 145–146. [36] W.A. Schroeder, J.V. Shelton, V. Huynh, J.R. Shelton, Hemoglobin 10 (1986) 239–251.
Oxidation of hemoglobin to methemoglobin in intact erythrocyte by a hydroperoxide induces formation of glutathionyl hemoglobin and binding of a-hemoglobin to membrane Kaori Murakami and Shiro Mawatari* Department of Nutrition and Health Science, Faculty of Human Environmental Science, Fukuoka WomenÕs University, 1-1-1 Kasumigaoka, Higashi-ku, Fukuoka, Japan Received 6 June 2003, and in revised form 22 July 2003
Abstract Biochemical consequences of oxidation of hemoglobin (Hb) in intact human erythrocytes were studied. The incubation of washed erythrocyte with 1 mM tert-butylhydroperoxide induced an increase in glutathionyl-Hb (G-Hb). The formation of G-Hb occurred linearly until 10 min in parallel with the formation of methemoglobin (metHb) after exhaustion of reduced glutathione. The results show that metHb, but not normal Hb, reacts with oxidized glutathione to form G-Hb. G-Hb was confirmed by immunoblotting with anti-glutathione antibody and the formation of G-Hb was accompanied by parallel decrease in b-globin detected with a reversed phase HPLC. Electrophoretic studies showed time-dependent increase in membrane-associated a-Hb until 10 min, indicating that a part of unpaired a-Hb bound to the membrane. Pre-b-globin increased despite the decrease in b-globin and a part of the increase was independent of the decrease in b-globin. Pre-b-globin reacted with anti-glutathione antibody, but it differs from G-Hb in many features. Ó 2003 Elsevier Inc. All rights reserved. Keywords: Intact erythrocytes; Peroxidation; Hydroperoxide; Hemoglobin; Methemoglobin; a-Hemoglobin; b-Hemoglobin; Pre-b-globin; Glutathionyl hemoglobin
Formation of methemoglobin (metHb)1 in intact erythrocytes causes dysfunction of erythrocytes including peroxidative changes in erythrocyte membranes [1,2] and is implicated in the pathology of unstable hemoglobin diseases [3–8]. Biological functions of nitric oxide may also relate to oxidation of Hb in intact erythrocytes [9–12]. However, biochemical consequences of formation of metHb in intact erythrocytes do not seem to be fully understood. tert-Butylhydroperoxide (tBHP), a cell membrane permeable hydroperoxide, is known to induce peroxi* Corresponding author. Fax: +92-671-7843. E-mail address: [email protected] (S. Mawatari). 1 Abbreviations used: G-Hb, glutathionyl-Hb; metHb, methemoglobin; tBHP, tert-Butylhydroperoxide; a-Toc, a-tocopherol; PL, phospholipids; HPLC, high-performance liquid chromatography; PBS, phosphate-buffered saline; DTNB, 5,50 -dithiobis(2-benzoic acid); SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel electrophoresis; PVDF, polyvinylidene difluoride.
0003-9861/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved. doi:10.1016/S0003-9861(03)00389-8
dative changes in intact human erythrocytes including peroxidation of Hb [13–17]. Our previous study showed that, by the incubation of erythrocytes with 1 mM tBHP, reduced glutathione (GSH) of the erythrocytes exhausted within 1 min followed by the changes in membrane a-tocopherol (a-Toc) and phospholipids (PL). a-Toc and PL decreased maximally within a few minutes, whereas the formation of metHb occurred linearly until 15 min [17]. Dithiothreitol (DTT), a sulfhydryl reductant, protected the erythrocytes completely against the formation of metHb by tBHP, indicating the importance of sulfhydryls in the erythrocytes [17]. Actually, the erythrocytes contain a large amount of GSH, which protects the erythrocytes from peroxidative changes. During the experiment of peroxidation of the intact human erythrocytes with tBHP, we observed time dependent decrease in b-globin until 10 min by a reversed phase high-performance liquid chromatography (HPLC). An ion exchange HPLC of Hb revealed that
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
the decrease in b-globin was associated with a parallel increase in glutathionyl-Hb (G-Hb). Electrophoretic studies showed time dependent increase in membranebound a-Hb until 10 min. These changes in Hb occurred in parallel with formation of metHb. Pre-b-globin increased despite the decrease of b-globin. Both G-Hb and pre-b-globin reacted with anti-glutathione antibody, however, pre-b-globin differed from G-Hb in many features. These biochemical consequences of peroxidation of Hb in intact erythrocytes by a hydroperoxide have not been reported hitherto, and the results indicate that metHb, but not normal Hb, reacts with oxidized glutathione (GSSG) to form G-Hb.
245
Measurement of metHb A portion (0.4 ml) of the erythrocyte suspension was hemolyzed with 5 ml of solution containing 100 mM phosphate buffer (pH 6.8) and 1% Triton X-100 (4:6, v/v) and the hemolysate was divided into two parts (2 ml each). Absorbance of the first part was read at 630 nm and then read again at 630 nm after addition of 0.02 ml of 5% potassium cyanide. Absorbance of the second part was read at 630 nm with 0.02 ml of 5% potassium ferricyanide and then read again at 630 nm after addition of 0.02 ml of 5% potassium cyanide [18]. The spectrophotometer used was Shimadzu BioSpec-1600 (Shimadzu, Kyoto, Japan). Measurement of erythrocyte GSH
Materials and methods
GSH, GSSG, tBHP, and 5,50 -dithiobis(2-benzoic acid) (DTNB) were obtained from Wako Pure Chem., Osaka, Japan. Rabbit anti-glutathione polyclonal antibody (Chemicon International Temecula, Ca) was purchased from Cosmo Bio, Tokyo, Japan. Other reagent grade chemicals and HPLC grade solvents were purchased from Wako Pure Chem., Osaka, Japan.
Erythrocyte GSH was measured by the recommended method by International Committee for Standardization in Haematology [19]. Briefly, 1 ml of the erythrocyte suspension was hemolyzed with 1 ml of water and then 3 ml of metaphosphoric acid solution, composed of 1.67 g of metaphosphoric acid, 0.2 g of EDTA-2Na, and 30 g of NaCl in 100 ml of water, was added. After filtration with a filter paper (No. 5B, Advantec, Tokyo, Japan), the filtrate (1 ml) was subjected to measurement of GSH using DTNB.
Preparation of erythrocytes
Preparation of Hb and erythrocyte membranes
Human venous blood from healthy adult volunteers who had fasted overnight, under informed consent, was drawn into tubes containing EDTA-2Na and processed immediately. Serum and buffy coats were removed after centrifugation at 1000g for 5 min, and the erythrocytes were washed three times in cold phosphate-buffered saline (PBS) at 1000g for 5 min. A small portion of the top layer was removed at each washing. The washed erythrocytes were re-suspended in PBS so that the final concentration of the erythrocytes was a hematocrit (Hct) of 10%. Hct was measured by using an automated blood counter (Sysmex F-520, Kobe, Japan). PBS, pH 7.4, was composed of 138 mM NaCl, 5 mM KCl, 6.1 mM Na2 HPO4 , 1.4 mM NaHPO4 , 1 mM MgCl2 , and 5 mM glucose.
The packed erythrocytes from 3 ml of the erythrocyte suspension were hemolyzed by 1.5 ml of cooled water and centrifuged at 20,000g for 15 min at 4 °C. To 0.9 ml of the supernatant, 0.3 ml of CCl4 was added [20]. The Hb solution was stored at )80 °C until use. Erythrocyte membranes (the sediment of 20,000g centrifugation after hemolyzation) were washed with 8 ml of 10 mM Tris– HCl, pH 7.4, by centrifugation at 20,000g for 15 min at 4 °C. The washing was repeated three or four times until the supernatant became clear. The membranes were stored at )80 °C until use.
Materials
Peroxidation of erythrocytes The total volume of the reaction mixture was brought up to 5 ml by the addition of PBS and the final concentration of the erythrocytes was 5% Hct. The final concentration of tBHP was 1 mM. After pre-incubation at 37 °C for 5 min, the reactions were started by addition of tBHP. The reactions were terminated putting the tubes in an ice bath. The erythrocytes were pelleted by centrifugation at 1000g for 5 min at 4 °C and washed twice with PBS. The erythrocytes were re-suspended in 5 ml PBS.
Separation of Hb chains Separation of hemoglobin chains (Hb-chains) was done using the reversed phase HPLC method [20,21]. The HPLC system consisted of a Hewlett Packard HP 1100 series (Yokogawa Analytic. Systems, Tokyo, Japan) equipped with a four solvent delivery system, a degassing unit, an automatic injector, a visual-ultraviolet detector, and a HP-Vectra computer with HPLC ChemStation software. The column was a Vydac C4, 250 4.6 mm (Agilent Technology, Tokyo, Japan). Mobile phase A was 20% acetonitrile in water containing 0.1% trifluoroacetic acid (TFA), and mobile phase B was 60% acetonitrile in water containing 0.1% TFA. Gradient elution was done with 52% B at 0 min, 59% B at 50 min, 92% B at 51 min,
246
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
92% B at 55 min, and 52% B at 60 min. The turnaround time was 70 min. The flow rate was 1 ml/min and the column temperature was 25 °C. Globins were detected at 220 nm. Some globin chain fractions were collected and concentrated under a stream of N2 gas followed by centrifugal ultrafiltration method using Centricon YM-10 (Millipore, Bedford, MA, USA). Determination of glutathionyl-Hb Separation of G-Hb was done with an ion-exchange HPLC [21,22]. The column was a TSK-GEL CM-2SW, 4.6 250 mm (Tosoh, Tokyo, Japan). Mobile phase A consisted of 30 mM Bis–Tris, 1.5 mM potassium cyanide, and 150 mM sodium acetate, pH 6.40, and mobile phase B was 30 mM Bis–Tris and 1.5 mM potassium cyanide, pH 6.4. The mobile phase A was increased linearly from 40 to 95% in 40 min. The flow rate was 0.8 ml/min and the column temperature was 30 °C. Hb was detected at 415 nm. For identification of G-Hb, GHb fraction was collected from the HPLC, concentrated by the centrifugal ultrafiltration method using Centriplus YM-10 (Millipore), and subjected to immunoblotting study with anti-glutathione antibody. Some of the Hb of the erythrocytes incubated with and without tBHP were incubated further with GSH or GSSG for 60 min at 37 °C. Immunoblotting of G-Hb and pre-b-globin The concentrated G-Hb and pre-b-globin were subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) with 14% gel in which 2mercaptoethanol was omitted from the sample buffer and electrically blotted on polyvinylidene difluoride (PVDF) membrane. PVDF membrane was blocked in PBS containing 5% skimmed milk and 0.1% Tween 20 and subsequently incubated overnight with anti-glutathione antibody at 5 °C. After washing, anti-rabbit IgG linked to horseradish peroxidase (Amersham Bioscience, Tokyo, Japan) was added and incubated for 1 h. Detection was performed using ECL Plus (Amersham Bioscience) according to the manufacturerÕs instructions.
used [19]. The gels were cast in a slab gel cell (AE-6401 M, ATTO, Tokyo, Japan), with gel size being 80 mm long, 90 mm wide, and 1 mm thick. One part by volume of the erythrocyte membrane solution (protein concentration was approximately 2 mg/ml) was dissolved in three parts by volume of sample buffer. The sample buffer consisted of 4 ml of 8 M urea, 0.4 ml of acetic acid, 0.4 ml of 2-mercaptoethanol, and 0.04 ml Triton X-100. Electrophoresis was done at 10 mA/gel for 4 h and Hb chains were stained with Coomassie brilliant blue.
Results Formation of metHb and no hemolysis The incubation of the erythrocytes with 1 mM tBHP induced a lineal increase in metHb until 10 min (Fig. 1). The control incubation without tBHP for 15 min showed no formation of metHb. None of the supernatants in the reaction mixtures after the incubation of the erythrocytes with tBHP showed coloration by monitoring at 540 nm, which indicated that no hemolysis had occurred under any of the conditions created in the present study. Rapid exhaustion of erythrocyte GSH GSH levels in the control reaction mixture containing 5% erythrocytes from six healthy adult volunteers were 90–160 lM. GSH was exhausted within 1 min after the incubation with 1 mM tBHP (Fig. 1).
Electrophoresis of erythrocyte membrane proteins The membranes were heated at 100 °C for 3 min in the sample buffer containing 0.125 M Tris–HCl, pH, 6.8, 10% sucrose, 4% SDS, and 10% of 2-mercaptoethanol, and an aliquot of the membrane solution was applied to SDS–PAGE with 12% gel. The protein bands were stained with Coomassie brilliant blue. To identify the Hb chains bound to the membranes, Triton acid urea gel electrophoresis was applied to the membrane [23,24]. Polyacrylamide gel (12%) containing 6 M urea and 2% Triton X-100 in 5% acetic acid was
Fig. 1. Time courses of the formation of metHb and the change in GSH after the incubation of erythrocytes with 1 mM tBHP. MetHb increased almost linearly until 10 min. GSH was exhausted within 1 min and began to recover after 10 min. Data are means SD of triplicate determinations of the erythrocytes from the same volunteer.
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
247
Decrease in b-globin and formation of G-Hb Each globin was clearly separated by the reversed phase HPLC within 50 min (Fig. 2). b-Globin decreased linearly from a 46% to a 35% per total globin (sum of pre-b-, b-, a-, and other minor globins) during the initial 10 min of the incubation of the erythrocytes with 1 mM tBHP (Figs. 2 and 3). The decrease in b-globin was about 10% of the initial level (Fig. 3). The relative composition of pre-b-globin increased despite the decrease in b-globin (Fig. 3), and a part of the increase of pre-b-globin was independent of the decrease in b-globin (Figs. 2 and 3). The ion exchange HPLC of Hb revealed an increase in G-Hb after the incubation of erythrocyte with tBHP (Fig. 4A). The increase in G-Hb was almost linear until 10 min (Fig. 6) and was associated with parallel decrease Fig. 3. Time courses of the changes in each Hb chain after the incubation with 1 mM tBHP. Hb chains were separated by a reversed phase HPLC as shown in Fig. 2. The increase in a-globin was relative to the decrease in b-globin; however, a part of the increase in pre-b-globin was independent of the decrease in b-globin. Each data point represents the mean SD of the erythrocytes from three different volunteers.
in HbA0 (data not shown). The amount of G-Hb at 10 min was about 9% of total Hb peaks (sum of HbA1c , G-Hb, HbA1d HbA0 , and HbA2 ) (Fig. 6). The peak of G-Hb decreased by the further incubation of Hb with 1 mM DTT for 60 min (Fig. 5). The Hb of the erythrocytes incubated without tBHP formed only trace amount of G-Hb by the further incubation with 1 mM GSSG (Fig. 4B-1). On the other hand, the Hb of the erythrocytes incubated with tBHP showed marked increase in G-Hb by the further incubation with 1 mM GSSG (Fig. 4B-2), but the Hb did not increase G-Hb by the incubation with GSH (data are not shown). SDS–PAGE of the concentrated G-Hb peak showed about 16 kDa protein and the electrophoretic mobility of G-Hb was almost same as that of b-globin (Fig. 7A). Immunoblotting study using anti-glutathione antibody confirmed the presence of glutathione in G-Hb (Fig. 7B). Pre-b-globin also reacted with anti-glutathione antibody (Fig. 7B).
Fig. 2. Representative HPLC chromatograms of Hb chains after the incubation of the erythrocytes with tBHP. (A) The incubation of the erythrocytes without tBHP. (B and C) b-Globin decreased time dependently after the incubation of the erythrocytes with tBHP for 5 min and 10 min, respectively. The chromatograms also show an increase in pre-b-globin at 5 min after the incubation with tBHP (B), which is independent of the decrease in b-globin. All of the chromatograms show the results from the same batch of erythrocytes for which incubations were done simultaneously. Heme, heme; a-gl, a-globin; b-gl, b-globin; and Pre-b-gl, and pre-b-globin.
Time dependent increase in binding of a-Hb to erythrocyte membranes The results of SDS–PAGE of the erythrocyte membranes show a time dependent increase in monomeric Hb until 10 min by the incubation of the erythrocytes with tBHP (Fig. 8A). The Triton acid urea gel electrophoresis of the membranes at 10 min incubation of the erythrocyte with 1 mM tBHP indicates that most of the Hb-chains bound to the membrane are a-Hb (Fig. 8B).
248
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
Fig. 4. Representative HPLC chromatograms of hemoglobin. (A-1) Control incubation of the erythrocyte without tBHP showing only trace of glutathionyl-Hb (G-Hb). (A-2) G-Hb increased after the incubation of erythrocyte with tBHP for 10 min. (B-1) The Hb of the erythrocytes of the control incubation formed only a little amount of G-Hb by the further incubation with 1 mM GSSG for 60 min (compare A-1 and B-1). (B-2) The Hb of the erythrocytes incubated with tBHP markedly increased G-Hb by the further incubation with 1 mM GSSG for 60 min (compare A-2 and B-2). All of the chromatograms show the results from the same batch of erythrocytes.
Discussion It is reported that HbA can be reacted with glutathione with disulfide bond formation between the cysteine 93 of b-Hb and the cysteine of glutathione [25,26] and that G-Hb is not present in significant amounts in normal erythrocytes because GSSG, which could form a mixed disulfide with Hb, is present at a very low con-
Fig. 5. G-Hb of the erythrocytes after the incubation with tBHP for 10 min (A) decreased by the further incubation with 1 mM DTT for 60 min (B).
centration [25]. In this study, the erythrocyte GSH was exhausted within 1 min by the incubation of intact erythrocytes with tBHP indicating rapid increase of GSSG (Fig. 1), but the formation of G-Hb occurred linearly until 10 min in parallel with the formation of metHb (Figs. 1 and 6). The results indicate that metHb, but not normal Hb (oxy-Hb and deoxy-Hb), reacts with GSSG to form G-Hb. Actually, the Hb of the erythrocytes incubated without tBHP formed only trace amount of
Fig. 6. Time dependent increase in glutathionyl-Hb (G-Hb) by the incubation of the erythrocytes with tBHP. G-Hb was calculated on the total Hb peaks (sum of HbA1c , G-Hb, HbA1d , HbA0 , and HbA2 ). Each data point represents the mean of two determinations from two different erythrocytes taken from two volunteers.
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
G-Hb by the further incubation with 1 mM GSSG for 60 min (Fig. 4B-1), whereas the Hb of the erythrocytes incubated with tBHP showed marked increase in G-Hb by the further incubation with GSSG (Fig. 4B-2). The G-Hb peak decreased by the incubation of the Hb with 1 mM DTT, indicating that metHb and GSSG formed disulfide bond (Fig. 5). The decrease in b-globin detected by the reversed phase HPLC is almost symmetrical to the increase in G-Hb detected by the ion exchange HPLC, indicating that the decrease in b-globin was due to the formation of G-Hb (Figs. 3 and 6). The results also indicate that disulfide bonded b-globin with glutathione is not detectable by the reversed phase HPLC method using 0.1% TFA, which has commonly been used for the separation of Hb chains [20,21].
249
The detection of G-Hb by a cation exchange HPLC was recently reported [22]. The mobile phases of HPLC used in this study were different from those in the reported HPLC; however, the appearance of G-Hb on the chromatograph was identical. We collected the G-Hb peak from the HPLC and confirmed for the first time that G-Hb reacts with anti-glutathione antibody (Fig. 7B). Recently, an increase in G-Hb has been reported in patients with diabetes mellitus, hyperlipidemia, uremia, and FriedreichÕs ataxia [31–34], and G-Hb is suggested to be a possible clinical marker of oxidative stress [35]. This study clearly shows that metHb in intact erythrocytes induces formation of G-Hb by reacting with GSSG. In human oxy-Hb, a-Hb is oxidized more rapidly than b-Hb [27,28]. Nitric oxide (NO) preferentially
Fig. 7. (A) SDS–PAGE of glutathionyl-Hb, pre-b-globin, and b-globin stained with Coomassie brilliant blue, indicating that the molecular weight of the three proteins is almost the same. Fourteen percent gel was used. 2-Mercaptoethanol was omitted from the sample buffer. (B) Immunoblotting study of the proteins separated by SDS–PAGE shows that glutathionyl-Hb and pre-b-globin react with anti-glutahione antibody. Each protein was collected repeatedly from HPLC of Hb from three different volunteers and concentrated. For details, see Materials and methods. G-Hb, glutathionylHb; Pre-b-gl, pre-b-globin; and b-gl, and b-globin.
Fig. 8. (A) SDS–PAGE of the membrane proteins shows time dependent increase in membrane associated Hb monomer until 10 min by the incubation of erythrocytes with tBHP. Twelve percent gel was used. (B) Triton acid urea gel electrophoresis of the membrane after the incubation with tBHP shows preferential binding of a-Hb to the membrane. These results were confirmed with erythrocytes from three different volunteers. *, Arrow indicates dimeric hemoglobin; **, arrow indicates monomeric hemoglobin; St, standard proteins; Cont Hb, the Hb of the erythrocytes of the control incubation; tBHP, the membranes of the erythrocytes incubated with tBHP for 10 min; b-Hb, b-hemoglobin; and a-Hb, a-hemoglobin. The numbers below lanes indicate the incubation time (min).
250
K. Murakami, S. Mawatari / Archives of Biochemistry and Biophysics 417 (2003) 244–250
binds to a-Hb in a low affinity tense state (T-state) as in deoxy-Hb, and NO binds preferentially to cysteine 93 on b-Hb in a relaxed state (R-state) as in oxy-Hb [9,10]. To our knowledge, it has not been reported hitherto that coexistence of met-Hb and GSSG in intact erythrocytes induces formation of G-Hb. Both G-Hb and S-nitrosohemoglobin are bonded to cysteine 93 of b-Hb and both substances are reported to increase the oxygen affinity [25,29,30]. It is suggested that transnitrosation of S-nitrosohemoglobin to S-nitrosoglutathione relates to regulation of blood flow [9]; therefore, it may be possible that G-Hb has also some physiological roles. Possible relationship between G-Hb and S-nitrosohemoglobin is unclear. The reversed phase HPLC of globin chains showed an apparent increase in pre-b-globin despite the decrease in b-globin (Fig. 2). The increase in pre-b-globin was faster than the decrease in b-globin and stopped at 5 min (Fig. 4). Thus, a part of the increase in pre-b-globin was independent of the decrease in b-globin. The nature of pre-b-globin does not seem to be well defined yet, but it is reported that the amino acid composition of pre-b-globin is indistinguishable to that of b-globin [36]. It is thought that pre-b-globin consists of a posttranslationally modified protein, presumably containing a b-globin–glutathione adduct [36]. This study shows for the first time that pre-b-globin reacts with anti-glutathione antibody (Fig. 7B). It is of interest to note that both pre-b-globin and G-Hb react with anti-glutathione antibody. However, the time course of the changes in pre-b-globin during the oxidation of Hb by the hydroperoxide was different from that of G-Hb (Fig. 4), and the present study shows that b-globin in G-Hb, in contrast to pre-b-globin, is not detectable by the reversed phase HPLC. This study indicates that unpaired a-Hb increased in parallel with the formation of G-Hb, and the electrophoretic studies of the membranes showed time dependent increase of the membrane associated a-Hb (Figs. 8A and B). This situation is similar to b-thalassemias, where lack of sufficient b-Hb leads to membrane associated a-Hb [7,8]. However, the membrane bound a-Hb is indicated to be only a small part of the unpaired a-Hb, so that the total a-globin chain is not significantly changed (Figs. 2 and 3). Acknowledgments This work was supported in part by a grant for scientific research from Fukuoka Prefecture, Japan, and by funding from the Institute of Rheological Functions of Food, Fukuoka, Japan. References [1] L.M. Snyder, N.L. Fortier, J. Trainor, J. Jacob, B. Lubin, D. Chiu, S. Sohet, N. Mohandas, J. Clin. Invest. 76 (1985) 1971–1977.
[2] P. Jarolim, M. Lahave, S.C. Liu, J. Palek, Blood 76 (1990) 2125–2131. [3] T.P. Flynn, D.W. Allen, G.J. Johnson, J.G. White, J. Clin. Invest. 71 (1985) 1215–1223. [4] S.-C. Liu, J.Y. Scott, J.R. Mehta, P.E. Nichols, S.K. Balias, P.W. Yacono, D.E. Golan, J. Palek, J. Clin. Invest. 97 (1997) 29–36. [5] B.R. Rank, J. Carisson, R.P. Hebbel, J. Clin. Invest. 75 (1985) 1531–1537. [6] R. Advani, S. Sorenson, E. Shinar, W. Lande, E.A. Rachimilewitz, S.L. Shrier, Blood 79 (1992) 1058–1063. [7] M.D. Scott, J.J.M. van den Berg, T. Repka, J. Clin. Invest. 91 (1993) 1706–1712. [8] R.E. Schrier, N. Mohandas, Blood 74 (1989) 2194–2202. [9] L. Jia, C. Bonaventura, J. Bonaventura, J.S. Stamler, Nature 380 (1996) 221–226. [10] H. Kosaka, Biochim. Biophys. Acta 1411 (1999) 370–377. [11] T.H. Han, D.R. Hyduke, M.W. Vaughn, J.M. Fukuto, J.C. Liao, Proc. Nat. Acad. Sci. USA 99 (2002) 7763–7768. [12] C. Bonavenrura, C.H. Taboy, P.S. Low, R.D. Stevens, C. Lafonn, A.L. Crumbliss, J. Biol. Chem. 277 (2002) 14557–14563. [13] R.J. Trotta, S.G. Sullivan, A. Stern, Biochem. J. 212 (1983) 759–772. [14] G.A. Dise, D.B.P. Goodman, Biochim. Biophys. Acta 859 (1986) 23–69. [15] B. Deuticke, K.B. Heller, C.W.M. Haest, Biochim. Biophys. Acta 899 (1987) 113–124. [16] T.T. Rohn, T.R. Hinds, F.F. Vincenczi, Biochim. Biophys. Acta 1153 (1993) 67–76. [17] S. Mawatari, K. Murakami, Lipids 36 (2001) 57–65. [18] K.A. Evelyn, H.T. Malloy, J. Biol. Chem. 126 (1938) 655–662. [19] International Committee for Standardization in Haematology, Recommended methods for additional red cell enzyme (pyridine 50 -nucleotidae) assay and the determination of red cell adenosine50 -triphosphate, 2,3-diphosphoglycerate and reduced glutathione, Clin. Lab. Haematol. 11 (1989) 131–138. [20] B. Masala, L. Manca, Methods Enzymol. 231 (1994) 21–44. [21] T.H.J. Huisman, J. Chromatogr. 418 (1987) 277–304. [22] A. Pastore, A.F. Mozzi, G. Tozzi, L.M. Gaeta, G. Federici, E. Bercini, A.L. Russo, L. Mannucci, F. Piemonte, Anal. Biochem. 312 (2003) 85–90. [23] B.P. Alter, S.C. Goff, G.D. Eferemov, M.E. Gravely, T.H.J. Huisman, Brit. J. Haematol. 44 (1980) 527–534. [24] S. Sorensen, E. Rubin, H. Polster, N. Mohandas, S. Schrier, Blood 75 (1990) 1333–1336. [25] M.-C. Carel, C. Domenget, J. Caburi-Martin, C. Prehu, F. Galacteros, Y. Beuzard, J. Biol. Chem. 261 (1986) 14704–14709. [26] C.T. Craescu, G. Poyart, C. Schaeffer, M.-C. Carel, J. Kister, Y. Beuzard, J. Biol. Chem. 261 (1986) 14710–14716. [27] A. Mansouri, K.H. Winterhalter, Biochemistry 12 (1973) 4949– 4976. [28] M. Tsuruga, A. Matsuoka, A. Hachimori, Y. Sugawara, K. Shikama, J. Biol. Chem. 273 (1998) 8607–8615. [29] R.P. Patel, N. Hog, N.Y. Spencer, B. Kalynaraman, S. Matalon, V.M. Darley-Usmar, J. Biol. Chem. 274 (1999) 15487–15492. [30] C. Bonaventura, G. Ferruzzi, S. Tesh, R.D. Steven, J. Biol. Chem. 274 (1999) 24742–24748. [31] T. Niwa, C. Naito, A.H. Mawjood, K. Imai, Clin. Chem. 46 (2000) 145–146. [32] C. Naito, T. Niwa, J. Chromatogr. B. 746 (2000) 91–94. [33] F. Takayama, S. Tsutsui, M. Horie, M. Shimokata, T. Niwa, Kidney Int. Suppl. 78 (2001) S155–S158. [34] F. Piemonte, A. Pastore, G. Tozzi, Eur. J. Clin. Invest. 31 (2001) 1007–1011. [35] S.E. Bursell, G.L. King, Clin. Chem. 46 (2000) 145–146. [36] W.A. Schroeder, J.V. Shelton, V. Huynh, J.R. Shelton, Hemoglobin 10 (1986) 239–251.