British Journal of Haematology, 1999, 104, 504–512
Metabolic indicators of oxidative stress correlate with haemichrome attachment to membrane, band 3 aggregation and erythrophagocytosis in b-thalassaemia intermedia M A R I A D O M E N I CA C A PPE L L I N I , DA R I O TAVA Z Z I , L O R E NA D U CA , G I OVAN NA G R A Z I A DE I , F R A N CA M AN N U,* F RA N C O T U R R I N I ,* PAO L O A R E S E * A ND G E M I N O F I OR E L L I Hereditary Anaemia Centre, Ospedale Maggiore IRCCS, Department of Internal Medicine, University of Milano Medical School, Milano, and *Department of Genetics, Biology and Biochemistry, University of Torino Medical School, Torino, Italy Received 23 June 1998; accepted for publication 19 November 1998
Summary. Haematological data, genotype, transfusion requirements, metabolic indicators of oxidative stress (flux via hexose-monophosphate shunt (HMPS); steady state level of GSH and GSSG, NADPH and NADP; activity of anti-oxidant enzymes), parameters of membrane damage (aggregated band 3; membrane-bound haemichromes, autologous immunoglobulins (Igs) and C3 complement fragments) and erythrophagocytosis were measured in erythrocytes (RBC) of 15 b-thalassaemia intermedia patients (nine splenectomized) with low, if any, transfusion requirements. Patients presented increased aggregated band 3, bound haemichromes, Igs and C3 complement fragments, and increased erythrophagocytosis. Bound haemichromes strongly correlated with aggregated band 3. Anti-band 3 Igs were predominantly associated with aggregated band 3. Erythrophagocytosis positively correlated with aggregated band 3, haemichromes and Igs, suggesting the involvement of haemichrome-induced band 3 aggregation in phagocytic removal of b-thalassaemic RBC.
Splenectomized patients showed higher degrees of membrane damage and phagocytosis, significantly higher numbers of circulating RBC precursors, and tendentially higher numbers of reticulocytes. Basal flux via HMPS was increased twofold, but HMPS stimulation by methylene blue was decreased, as was the glucose flux via HMPS. GSH was remarkably decreased, whereas NADPH was increased. Except for unchanged catalase and glutathione reductase, anti-oxidant enzymes had increased activity. Negative correlation between HMPS stimulation by methylene blue and bound haemichromes indicated that the ability to enhance HMPS may counteract haemichrome precipitation and limit consequent membrane damage leading to erythrophagocytosis.
In b-thalassaemia erythrocytes (RBC), excess free ahaemoglobin chains due to decreased synthesis of b-globin induce oxidative damages to both integral and cytoskeletal proteins (see Shinar & Rachmilewitz, 1990, 1993, for reviews). Clusters containing aggregated membrane proteins and precipitated haemoglobin derivatives (i.e. haemichromes (HCR), free haem and non-haem iron were found tightly associated to the thalassaemic RBC membrane (Rouyer-Fessard et al, 1989; Jarolim et al, 1990; Repka et
al, 1993). Thiol oxidation and aggregation of band 3 were responsible for massive deposition of removal opsonins such as immunoglobulins (Igs) with anti-band 3 specificity and complement fragments (Yuan et al, 1992; Mannu et al, 1995). Not surprisingly, thalassaemia intermedia RBC were phagocytosed more intensely by mouse macrophages than normal controls (Knyszynski et al, 1979). Excess free ahaemoglobin chains also influence RBC redox metabolism in b-thalassaemia. Significantly increased activity of glucose-6phosphate dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (6PGD) was observed in homozygous and heterozygous b-thalassaemic patients (Vives Corrons et al, 1984; Ponnazhagan & Sarkar, 1992), whereas glucose consumption via glycolysis was remarkably increased in
Correspondence: Dr Maria Domenica Cappellini, Department of Internal Medicine, Ospedale Maggiore Policlinico, Padiglione Granelli, Via F. Sforza 35, 20122 Milano, Italy. e-mail: maria.
[email protected].
504
Keywords: beta-thalassaemia intermedia, membrane alterations, phagocytosis, hexose-monophosphate shunt, anti-oxidant defence.
q 1999 Blackwell Science Ltd
Membrane Damage and Stress Indicators in b-thalassaemia intermedia b-thalassaemia intermedia RBC (Ting et al, 1994). However, no data are available on correlations between membrane damage, alteration of redox metabolism and induction of phagocytosis in this condition. The aim of the present work was to examine a larger number of indicators of oxidative stress and membrane damage, and RBC phagocytosis by monocytes, in a random sample of 15 splenectomized or non-splenectomized bthalassaemia intermedia patients. Our results show significant changes and a number of correlations between membrane damage, deposition of removal opsonins, erythrophagocytosis and metabolic indicators of oxidative stress. In general, the transfusion regimen was higher and indicators of membrane damage were more severely altered in splenectomized patients, whereas RBC numbers and haemoglobin levels were similar between the two groups. MATERIALS AND METHODS Materials. Rabbit anti-human immunoglobulins (Igs), rabbit anti-human C3c, mouse anti-rabbit second antibodies conjugated to alkaline phosphatase, EDTA, N-ethyl maleimide (NEM), phenylmethylsulphonyl fluoride (PMSF) and sodium dodecyl sulphate (SDS) were from Sigma (St Louis, Mo.). Octaethylene glycol mono-n-dodecyl ether (C12E8, polydocanol) was from Nikkei Chemical Co. (Tokyo, Japan). Eosin-5-maleimide (EOS) was from Molecular Probes (Eugene, Ore.). Sepharose CL-6B and Percoll were from Pharmacia Biotech (Uppsala, Sweden). [1-14C]glucose (1·85 GBq/mmol) was from Amersham Life Science (Little Chalfont, U.K.). Pico-Fluor 40 liquid scintillation cocktail was from Packard Instrument Co. (Meriden, Conn.). Dowex AG 1-X8 ion-exchange resin was from BioRad (Hercules, Calif.). Sterile plastics were from Costar (Cambridge, Mass.). All other reagents were purchased from common commercial sources. Patients. Eight male and seven female (aged 20–44 years) b-thalassaemia intermedia patients were studied. Diagnosis was based on clinical criteria, no or low transfusion requirement and genetic analysis. The b-thalassaemic genotype was determined in all patients by amplification refractory mutation system as previously reported (Camaschella et al, 1995). Nine patients were splenectomized and none of the patients had received blood transfusions in the preceding 3 months. Normal controls were sex- and age-matched apparently healthy laboratory personnel. All blood samples were drawn after informed consent according to protocols approved by University of Milano Ethical Committee. Samples of venous blood (8– 10 ml) were collected in heparin or K3EDTA, or without additive. Samples were kept at 48C after withdrawal and processed immediately (metabolite levels) or within 1 h (hexose monophosphate shunt [HMPS] flux, enzyme activities, membrane studies) to avoid artefacts due to blood storage. Representative haematological data obtained by standard techniques and analysis of the haemoglobin bchain mutation of the 15 patients are summarized in Table I. Preparation of hypotonic membranes. Standard hypotonic membranes were prepared at 08C by haemolysis in
505
haemolysis buffer (5 mM sodium phosphate, 1 mM EDTA pH 8·0) and two washes. To minimize handling artefacts, preparation of hypotonic membranes including three 2 min centrifugations in a refrigerated Eppendorf microfuge was performed 4000 kD molecular weight) and equilibrated with a solution containing 50 mM NaCl, 10 mM Hepes, 0·1% (vol/vol) polydocanol, pH 7·4. 1 ml of the polydocanol extract prepared from hypotonic membranes was loaded onto the column and separated at a flow rate of 1 ml/min. The effluent was collected in 0·8 ml fractions by an automated device (Waters 650 Advanced Protein Purification System, Millipore, Mass.). Assay of membrane-bound haem-containing compounds. Haem-containing compounds were quantified in polydocanol extracts, and identified as haemichromes (HCR), methaemoglobin (MetHb) or free haem according to their spectral characteristics (Winterbourn, 1985). Haemichromes were quantified by measuring absorbance at 560, 577 and 630 nm in the excluded volume fractions, using millimolar absorptivity values of 8·6, 6·8 and 0·92, respectively (Winterbourn, 1985). Due to the great variability in protein content of membrane in thalassaemic RBC, membrane volume was selected as a reference and haemcontaining compounds were expressed as nanomol haem/ml membrane. Membrane volume showed low inter-subject variability and roughly corresponded to the original RBC volume in both normal and thalassaemic RBC. Assay of aggregated band 3. RBC were washed in Hepesbuffered saline (HBS) (140 mM NaCl, 10 mM Hepes, 1 mM EDTA, 10 mM glucose, pH 7·4) and incubated in the dark for 20 min at room temperature in the same buffer containing 5 mM eosin-5-maleimide (EOS) (Cobb & Beth, 1990). Excess unbound EOS was removed by washing the cells in HBS containing 0·2% (wt/vol) bovine serum albumin followed by two washes with HBS alone. EOS-labelled RBC were treated as described above for membrane preparation and gelfiltration analysis of membrane-bound protein aggregates. EOS-labelled band 3 in eluted fractions was assayed by fluorometry (excitation wavelength at 522 nm, emission wavelength at 550 nm) and the fluorescence value measured in the excluded volume fractions was normalized to the total fluorescence measured in all fractions (Turrini et al, 1994). Assay of membrane-bound Igs and complement C3c. Membrane-bound Igs were measured after labelling RBC with rabbit anti-human Igs or rabbit anti-human C3c as described (Turrini et al, 1994). Briefly, washed RBC (10% haematocrit) were incubated for 1 h at 48C in HBS buffer containing 2%
q 1999 Blackwell Science Ltd, British Journal of Haematology 104: 504–512
506
Maria Domenica Cappellini et al
(wt/vol) bovine serum albumin and rabbit anti-human Igs or anti-human C3c (both diluted 1:1000). RBC were then washed five times in HBS buffer containing 1% (wt/vol) bovine serum albumin. After Igs labelling with second antibodies conjugated to alkaline phosphatase, hypotonic membranes were prepared, extracted by polydocanol (see above) and separated by gel-filtration chromatography as described before. Alkaline phosphatase activity was measured photometrically in the fractions and was typically localized in the excluded volume (>4000 kD molecular weight). Total activity found in the excluded volume was normalized to the volume of hypotonic membranes loaded on the column. Membrane-bound Igs were expressed as alkaline phosphatase activity, as the increment in absorbance at 405 nm/min/ml membrane. Assay of phagocytosis. Phagocytosis of RBC by adherent human monocytes was performed as described (Schwarzer et al, 1994). Monocytes were separated from fresh buffy coats discarded from apparently healthy normal donors, plated and utilized as described previously (Schwarzer et al, 1994). Erythrophagocytosis was expressed as the number of ingested RBC/monocyte. Assay of HMPS flux. HMPS flux was measured in leucocyte-free RBC as production of 14C-CO2 from D[1-14C]glucose as indicated (Gaetani et al, 1974). Assays were performed without stimulation (basal HMPS flux) or with stimulation by 150 mM methylene blue (MB) (MBstimulated HMPS). HMPS flux was expressed as mmol CO2 produced/h/1010 RBC at 378C. HMPS stimulation index was the ratio between MB-stimulated and basal HMPS flux. Glucose via HMPS was the percentage of glucose metabolized via HMPS. Glucose via HMPS stimulation index was the ratio between glucose via HMPS with and without MB stimulation. Basal and MB-stimulated glucose via HMPS was measured as indicated (Gaetani et al, 1974). Briefly, acidquenched, neutralized RBC extracts obtained at time 0 and after incubation with D-[1-14C]glucose were passed through a Dowex AG 1-X8 ion-exchange column in order to isolate nonmetabolized glucose and the terminal glycolytic products, lactate and pyruvate. After determination of the specific radioactivity of glucose and determination of the radioactivity of terminal glycolytic products, glucose via HMPS was calculated as indicated (Gaetani et al, 1974). Assay of enzyme activity. Activity of glutathione reductase (GR), glutathione peroxidase (GPX), catalase (CAT), glucose6-phosphate dehydrogenase (G6PD) and 6-phosphogluconate dehydrogenase (6PGD) was determined by standard spectrophotometric techniques, as recommended by the International Committee for Standardization in Haematology (ICSH, 1979). GR was assayed with or without 1 mM flavin adenine dinucleotide (FAD) to assess full activation of the apoenzyme. Enzyme activites were expressed as mmol substrate transformed/min/1010 RBC at 378C. Assay of GSH, GSSG, NADPH and NADP and creatine. GSH and GSSG were assayed in neutralized acid extracts obtained from heparinized peripheral blood (Neuschwander-Tetri & Roll, 1989). Glutathione was derivatized with orthophthalaldehyde (OPA) and separated by high-performance liquid chromatography (HPLC) in reversed-phase conditions. OPA
was assayed by fluorometry (excitation wavelength at 340 nm, emission wavelength at 420 nm). NADPH and NADP were determined in alkaline extracts from heparinized peripheral blood (Stocchi et al, 1987). Assay was performed by reversed-phase HPLC on a Novapak-C18 column (Waters) by monitoring nucleotide elution profiles at 254 nm. Creatine was measured by a standard colourimetric method (Griffiths & Fitzpatrick, 1967). Steady-state metabolite levels were expressed as mmol/1010 RBC. RBC separation by Percoll discontinuous density gradient. Leucocyte-free RBC obtained from a patient with bthalassaemia intermedia chosen at random were separated on a Percoll discontinuous density gradient (60–63–66– 69% segments) prepared according to the manufacturer’s indications (Pharmacia Biotech, Uppsala, Sweden). The top fraction containing less dense reticulocytes and young RBC, and the bottom fraction containing most dense old RBC, were collected and centrifuged at 2700 rpm for 10 min at 48C. Sedimented RBC were washed three times with ice-cold PBS to remove Percoll. Creatine levels were measured in RBC fractions as an age-related parameter to monitor RBC separation (Griffiths & Fitzpatrick, 1967). Statistical analysis. Data were analysed by the nonparametric Mann-Whitney test. RESULTS Haematological data, genotype and transfusion requirements of patients Table I shows the basic haematological data and genotype of the 15 patients, according to the spleen status of the patients. Mean values 6SD and intra-group significance between splenectomized and non-splenectomized patients are indicated. Although gross indicators of RBC homeostasis such as haemoglobin level and number of RBC are very similar, distinctly higher numbers of circulating erythroblasts were observed in the splenectomized patients. 8/15 patients had no record of previous transfusions, 5/12 patients (patients 5, 7, 9, 11 and 13) had received occasional transfusions during infancy, one patient (no. 1) received 1 unit per year and only one patient (no. 2) needed 10–12 units per year. The two patients receiving transfusions and three of the occasionally transfused patients belonged to the splenectomized group. None of the non-splenectomized patients had received transfusions in the last 10 years. Indicators of membrane damage: aggregated band 3, membrane-bound haem-containing compounds, autologous Igs and complement C3 fragments In b-thalassaemic RBC, high-molecular-weight aggregates were isolated by gel-filtration chromatography in the excluded volume fractions where those aggregates constantly contained aggregated band 3, haem-containing compounds, and almost the totality of membrane-bound Igs and complement C3 fragments. Aggregated band 3 ranged from 1·42 to 17·50% of the total extracted band 3 (average 7·63 6 4·69, n¼15). HCR ranged from 0·21 to 12·8 nmol/ml membrane (average 4·45 6 4·07, n¼15), MetHb from 1·45 to 76·6 nmol/ml membrane (average
q 1999 Blackwell Science Ltd, British Journal of Haematology 104: 504–512
Membrane Damage and Stress Indicators in b-thalassaemia intermedia
507
Table I. Haematological and genetic data of b-thalassaemia intermedia patients.
Patient Sex
Age
Splenectomized 1 M 25 2 F 27 3 F 24 4 F 21 5 M 28 6 M 26 7 F 20 8 M 33 9 F 44 Mean
Haemoglobin MCV (g/dl) (fl)
MCHC (g/dl)
Reticulocytes RBC (%) (×1012/l)
Erythroblasts Bilirubin (%) (mmol/l)
Genotype
6·3 7·4 7·6 9·2 8·8 10·0 8·9 12·1 7·7
30·7 34·2 31·7 31·6 32·1 30·3 33·1 33·7 32·2
5·0 5·5 9·0 6·8 1·6 2·1 3·3 4·5 5·1
2390 2730 2660 3530 3540 4440 3570 4110 2820
86 70 91 95 60 89 90 63 83
54·7 44·5 80·4 42·8 78·7 61·6 54·7 53·1 44·5
Cod39/Cod39 Cod39/IVS 1-110 IVS 1-6/IVS 1-110 Cod39/IVS 1-110 db/IVS 1-110 IVS 1-6/IVS 1-110 db/IVS 1-110 Cod39/Cod39 IVS 1/IVS 1-61
3331 6 700
80 6 13
57·2 6 14·1
4650 2980 4300 3380 3750 3550
2 4 14 9 2 2
63·3 54·7 63·3 56·5 42·8 42·9
3770 6 610
565
53·9 6 9·3
0·233
