British Journal of Haematology, 1979,42, 215-223.
Oxidative Damage to Neutrophils in Glutathione Sy n t he t a se D eficiency STEPHEN P. SPIFI BERG, LAURENCE A. BOXER, JANET M. OLIVER, J O H N M. ALLEN AND J O ~ E P HD. SCHULMAN Section on Human Biochemical Genetics, National Institutes of Child Health and Human Development, Departments of Pediatrics and Pharmacology, johns Hopkins University School of Medicine, Department of Pediatrics, University ojlndiana School of Medicine, and Department of Physiology, University of Connecticut School of Medicine (Received 2 March 1978; accepted for publication 9 October 1978) Several episodes of neutropenia were observed in a child with glutathione synthetase deficiency (5-oxoprolinuria). Studies of the patient’s glutathione-deficient neutrophils were undertaken to examine the responses of the cells to oxidative stress associated with phagocytosis. The patient’s neutrophils contained ro-zo% of normal glutathione content. Circulating neutrophils in infection-free periods appeared less mature than normal by morphologic criteria, suggesting increased cell turnover. The cells ingested particles, responded to chemotactic stimuli, and oxidized 1-I4C glucose normally. However, following ingestion of particles, the cells accumulated excess hydrogen peroxide compared with normal cells, and showed impaired protein iodination and bacterial killing. Electron micrographs revealed damage to microtubules and membranous structures in the patient’s neutrophils during phagocytosis. The level of glutathione in the cells appears inadequate to protect against peroxide generated during normal cell function, and the cells are thus damaged and rendered less effective in bacterial killing. The data provide evidence for a protective role of glutathione in normal neutrophil function. SUMMARY.
One of the major functions of glutathione is the protection of cells against oxidative stress from endogenous production of peroxide as well as from drugs and their metabolites (Meister, 197s). The erythrocyte is most often thought of as a target cell in individuals with abnormalities in glutathione metabolism. Thus, haemolytic anaemia resulting from oxidative damage to haemoglobin and cell membranes is a hallmark of diseases such as glucose-6-phosphate dehydrogenase deficiency that are associated with decreased available reduced glutathione (Beutler, 1969). In leucocytes, increased production of peroxide accompanies the normal process of phagocytosis (Iyer et al, 1961). The peroxide is utilized in bacterial killing, excess peroxide being detoxified in large part via glutathione peroxidase (Mills, 1960). Reduced capacity to generate peroxide as in chronic granulomatous disease results in impaired polymorphonuclear leucocyte (PMN) bacterial killing (Holmes et al, 1967). Conversely, P M N defiCorrespondence: Dr J. D. Schulman, NIH Building 10/13N260, Bethesda, Md 20014, U.S.A. 0007-1048/79/0600-021 s$oz.oo 01979 Blackwell Scientific Publications 21s
216
Glutathione Deficiency Neutrophils
cient in glutathione might be at increased risk for toxicity from peroxide produced during phagocytosis. The observation of several episodes of neutropenia during the course of acute infections in a patient with glutathione synthetase deficiency with 5-oxoprolinuria prompted detailed investigation of PMN function in this disorder. The patient’s erythrocytes and leucocytes both havc 5 % of normal glutathione synthetase activity and 1 ~ 2 0 %of normal glutathione content (Spielberg et al, 1977).We have found that the patient’s PMN accumulate greater than normal amounts of peroxide during phagocytosis, exhibit impaired iodination and bacterial killing, and show electron microscopic evidence of damage to intracellular structures on phagocytic stimulation. The data suggest that glutathione deficient PMN cannot effectively detoxify peroxide, and thus have decreased bactericidal capacity and are destroyed more readily during normal cell function. CASE REPORT The patient (A.R.) presented as a newborn with acidosis and haemolytic anaemia (Spielberg et al, 1977). 5-Oxoprolinuria was diagnosed by gas-liquid chromatography-mass spectroscopy, and biochemical studies revealed a deficiency of glutathione synthetase activity (s-ro’h of normal) and intracellular glutathione content (ro-zo% or normal) in erythocytes, leucocytes, and cultured skin fibroblasts. Growth has been normal to 2 years of age. The patient’s only medication is sodium citrate for correction of acidosis. There is no history of skin abscesses, pneumonia, fungal infections, meningitis, or sepsis. However, during the second year of life, the patient had six episodes of acute otitis media. Neutropenia was noted during two such infections. White blood cell counts fell to 3 .o x 109/l with one neutrophil and one band on one occasion, and to 4.6 x ro9/l with six neutrophils and six bands on the other. (The patient typically maintains a normal white count and differential.) Haemoglobin and reticulocyte and platelet counts remained unchanged in the patient. The neutropenia occurred prior to initiation of any medication and counts returned to normal in 4-5 d. Repeated blood cultures were negative and the patient responded well to oral ampicillin. Also noted were elevations of SGOT (244 iu/l), SGPT (304 iu/l), LDH (452 iu/l), with alkaline phosphatase 180 iu/l, bilirubin 0.4 mg/dl, and HAA negative. Enzymes returned to normal in approximately I week. There was no hepatosplenomegaly nor hepatic tenderness. Elevations of hepatic enzymes had been noted during several other febrile illnesses. METHODS
Cells. Human leucocytes were isolated from peripheral blood by dextran sedimentation of heparinized blood (Boxer et al, 1977). Purified neutrophils were further isolated using a Ficoll-Hypaque gradient. Red cells were removed by brief hypotonic lysis and leucocytes resuspended in Krebs ringer phosphate (KRP). The patient’s and normal purified neutrophils were 9 8 9 9 % mature and I-2% bands by light microscopy. Bacterial killing test and paraffin oil phagocytosis. Purified neutrophils ( 107/ml) were incubated a t 37°C for varying periods of time up to roo min with equal numbers of Staphylococcus aureus 502A opsonized with AB negative human serum (Quie et al, 1967). After lysis of the
Glutathione Dejiciency Neutrophils
217
neutrophils, the viable bacteria were enumerated on agar pour plates. Ingestion of lipopolysaccharide-coated paraffin-oil droplets and reduction of nitroblue tetrazolium (NBT) by mixed leucocytes were analysed as previously described (Stossel, 1973). Chemotaxis. Leucocyte migration in uitro was measured with the modified Boyden chamber technique (Hill et al, 197s). 1.5 x 1 0 6 neutrophils were deposited with a cytocentrifuge (Shandon Scientific Co.) on a premoistened 5 pm Millipore filter (Millipore Corporation). The filter was then placed in the Boyden chamber and incubated for 3 h at 37°C in a 10%COz, 90% air incubator. A bacterial chemotatic factor was prepared from a culture filtrate of Escherirhia coli. The chemotatic index was determined from the number of P M N migrating through the filter in 10 high-power objective fields. Metabolic studies. The rate of 14C02 evolution from [ ~ - ~ ~ C ] g l u c owas s e determined by a method previously described (Boxer & Stossel, 1974). The availability of H 2 0 2 following stimulation was quantified by a modification of the technique of Root et a1 (1975). Purified neutrophils (2.5 x ro6/ml) were suspended in KRP to which I mM sodium azide had been added. The cells were incubated at 37°C for 20 min to determine resting H 2 0 2 levels and in the presence of 0.I ml of either latex (1000particles/cell) or Concanavalin A (100 pg/l) for 5 min a t 37°C to determine stimulated activity. Following centrifugation at 400 g, I ml of the supernatant was added to 0.4 ml of scopoletin (final concentration 20 nM) and 0.2 ml of KRP, pH 7.4. H 2 0 2 was measured by reduction in scopoletin fluorescence using a Perkin-Elmer spectrofluorimeter. The rate of iodination was determined by the method of Hakim et al (1975).
Electron microscopy. Cell pellets were fixed for 3 0 min at room temperature in i % glutaraldehyde in 0.1 M sodium cacodylate buffer pH 7.4, washed with buffer, and processed for transmission electron microscopy as previously described (Albertini et al, 1977; Oliver et al, 1976). Thin (silver) sections through neutrophil centrioles were examined in a Philips 3 0 0 electron microscope for the presence or absence of microtubules. Other random sections that were positively identified as neutrophils from nuclear morphology were photographed for observation of nuclear condensation, the presence or absence of mitochondria and examination of general ultrastructural integrity.
RESULTS Despite the marked deficiency of glutathione content of the patient’s neutrophils, ingestion of opsonized lipopolysaccharide paraffin-oil droplets, NBT reduction, and directed cell movement in the Boyden chamber by these cells were normal (Table I). Similarly, the patient’s neutrophils were able to increase glucose oxidation (a measure of hexose monophosphate shunt activity) normally in response to phagocytosis of opsonized zymosan (Table 11). Two glutathione oxidants, diamide and t-butyl hydroperoxide, both of which increase the cell’s requirement for NADPH to reduce the generated oxidized glutathione, also stimulated glucose oxidation normally in the patient’s cells. In comparison to normal neutrophils, however, the patient’s cells showed an abnormally high accumulation of peroxide upon ingestion of opsonized zymosan (Table III), and their bactericidal capacity was impaired (Fig I ) . PMN iodination was also depressed in the glutathione deficient cells (Fig 2). Iodination is dependent on the production of peroxide, myeloper-
Stephen P. Spielberg et a1
21 8
TABLE 1. Phagocytic rates and chemotaxis of glutathione deficient neutrophils Ingestion*
Normal Patient
0.037
NBT redurriont C h e m o t u x i a 0.148 0.259
0.03 I
213+24 233f35
* Initial rate of ingestion of E. coli lipopolysaccharide-coated paraffinail droplets opsonized with fresh homologous serum as mg paraffin oil ingested per 10’ neutrophils/min (mean, two determinations). t Initial rate of NB T reductions following ingestion of opsonized paraffin-oil droplets as mg formazan per 10’ neutrophilslmin (mean, two determinations). $ Chemotactic index: Number of neutrophils in 10 random fieldslnumber of neutrophils ( x I O ~ in ) 0.4 ml delivered to the starting side of the filter, mean fSD. TABLE 11. [ ~-‘~C]Glucose oxidation by glutathione deficient neutrophils ‘‘C02/io6 c e l l s / ~ omin
cpm
Normal
Patient
~~
Resting Zymosan Diamide 5 nM Diamide 50 nM Diamide 200 nM Butylhydroperoxide 5 nM Butylhydroperoxide 50 nM Butylhydroperoxide 200 nM
97+ 7 64f 22 6 3 1 f 1 6 650+129 I75k 33 869+ 69 665+ 3 881k 70 238+ 4 180f 1 5 367+ 2 302k 47 676frz9 211k25
688+50
~
Cells (2.0x Io6/assay) were suspended in 0.9 ml of KRP, pH 7.4, containing glucose and 0.5 pCi [ ~-‘~C]glucose. Test solutions (0.I ml) were added and the suspensions were incubated at 37°C for 30 min. Values expressed as mean fSD.
oxidase activity, and the cells’ ability to deliver myeloperoxidase from lysosomes to the phagosome. The patient’s iodination was markedly decreased despite elevated peroxide content and normal myeloperoxidase activity [ 0 . D . 4 4 0 n m = z . ~ / mprotein/min; g normal 3 . 8 & I . I (mean of three+ SD) by a nonautomated modification of the 0-toluidine method of Baggiolini et al (1y6y)], suggesting a n impairment in lysosomal fusion in the patient’s cells. Diamide decreased iodination in the normal and nearly abolished the reaction in glutathione deficient neutrophils, further suggesting a vital role of reduced glutathione in iodination.
Clutathione Deficiency Neutrophils
219
TABLE Ill. Available H 2 0 2 in glutathione deficient neutrophils*
Resting Zymosan phagocytosis
Normal
Pallent
O.I+_O.l
O.I+_O.l
2 . 1 +o.r
3.4+0.2t
* Means+SD are indicated. Values are expressed as nmoles H 2 0 2 / ~ . x5 lo6 neutrophils/min. Different from normal, P
Oxidative Damage to Neutrophils in Glutathione Sy n t he t a se D eficiency STEPHEN P. SPIFI BERG, LAURENCE A. BOXER, JANET M. OLIVER, J O H N M. ALLEN AND J O ~ E P HD. SCHULMAN Section on Human Biochemical Genetics, National Institutes of Child Health and Human Development, Departments of Pediatrics and Pharmacology, johns Hopkins University School of Medicine, Department of Pediatrics, University ojlndiana School of Medicine, and Department of Physiology, University of Connecticut School of Medicine (Received 2 March 1978; accepted for publication 9 October 1978) Several episodes of neutropenia were observed in a child with glutathione synthetase deficiency (5-oxoprolinuria). Studies of the patient’s glutathione-deficient neutrophils were undertaken to examine the responses of the cells to oxidative stress associated with phagocytosis. The patient’s neutrophils contained ro-zo% of normal glutathione content. Circulating neutrophils in infection-free periods appeared less mature than normal by morphologic criteria, suggesting increased cell turnover. The cells ingested particles, responded to chemotactic stimuli, and oxidized 1-I4C glucose normally. However, following ingestion of particles, the cells accumulated excess hydrogen peroxide compared with normal cells, and showed impaired protein iodination and bacterial killing. Electron micrographs revealed damage to microtubules and membranous structures in the patient’s neutrophils during phagocytosis. The level of glutathione in the cells appears inadequate to protect against peroxide generated during normal cell function, and the cells are thus damaged and rendered less effective in bacterial killing. The data provide evidence for a protective role of glutathione in normal neutrophil function. SUMMARY.
One of the major functions of glutathione is the protection of cells against oxidative stress from endogenous production of peroxide as well as from drugs and their metabolites (Meister, 197s). The erythrocyte is most often thought of as a target cell in individuals with abnormalities in glutathione metabolism. Thus, haemolytic anaemia resulting from oxidative damage to haemoglobin and cell membranes is a hallmark of diseases such as glucose-6-phosphate dehydrogenase deficiency that are associated with decreased available reduced glutathione (Beutler, 1969). In leucocytes, increased production of peroxide accompanies the normal process of phagocytosis (Iyer et al, 1961). The peroxide is utilized in bacterial killing, excess peroxide being detoxified in large part via glutathione peroxidase (Mills, 1960). Reduced capacity to generate peroxide as in chronic granulomatous disease results in impaired polymorphonuclear leucocyte (PMN) bacterial killing (Holmes et al, 1967). Conversely, P M N defiCorrespondence: Dr J. D. Schulman, NIH Building 10/13N260, Bethesda, Md 20014, U.S.A. 0007-1048/79/0600-021 s$oz.oo 01979 Blackwell Scientific Publications 21s
216
Glutathione Deficiency Neutrophils
cient in glutathione might be at increased risk for toxicity from peroxide produced during phagocytosis. The observation of several episodes of neutropenia during the course of acute infections in a patient with glutathione synthetase deficiency with 5-oxoprolinuria prompted detailed investigation of PMN function in this disorder. The patient’s erythrocytes and leucocytes both havc 5 % of normal glutathione synthetase activity and 1 ~ 2 0 %of normal glutathione content (Spielberg et al, 1977).We have found that the patient’s PMN accumulate greater than normal amounts of peroxide during phagocytosis, exhibit impaired iodination and bacterial killing, and show electron microscopic evidence of damage to intracellular structures on phagocytic stimulation. The data suggest that glutathione deficient PMN cannot effectively detoxify peroxide, and thus have decreased bactericidal capacity and are destroyed more readily during normal cell function. CASE REPORT The patient (A.R.) presented as a newborn with acidosis and haemolytic anaemia (Spielberg et al, 1977). 5-Oxoprolinuria was diagnosed by gas-liquid chromatography-mass spectroscopy, and biochemical studies revealed a deficiency of glutathione synthetase activity (s-ro’h of normal) and intracellular glutathione content (ro-zo% or normal) in erythocytes, leucocytes, and cultured skin fibroblasts. Growth has been normal to 2 years of age. The patient’s only medication is sodium citrate for correction of acidosis. There is no history of skin abscesses, pneumonia, fungal infections, meningitis, or sepsis. However, during the second year of life, the patient had six episodes of acute otitis media. Neutropenia was noted during two such infections. White blood cell counts fell to 3 .o x 109/l with one neutrophil and one band on one occasion, and to 4.6 x ro9/l with six neutrophils and six bands on the other. (The patient typically maintains a normal white count and differential.) Haemoglobin and reticulocyte and platelet counts remained unchanged in the patient. The neutropenia occurred prior to initiation of any medication and counts returned to normal in 4-5 d. Repeated blood cultures were negative and the patient responded well to oral ampicillin. Also noted were elevations of SGOT (244 iu/l), SGPT (304 iu/l), LDH (452 iu/l), with alkaline phosphatase 180 iu/l, bilirubin 0.4 mg/dl, and HAA negative. Enzymes returned to normal in approximately I week. There was no hepatosplenomegaly nor hepatic tenderness. Elevations of hepatic enzymes had been noted during several other febrile illnesses. METHODS
Cells. Human leucocytes were isolated from peripheral blood by dextran sedimentation of heparinized blood (Boxer et al, 1977). Purified neutrophils were further isolated using a Ficoll-Hypaque gradient. Red cells were removed by brief hypotonic lysis and leucocytes resuspended in Krebs ringer phosphate (KRP). The patient’s and normal purified neutrophils were 9 8 9 9 % mature and I-2% bands by light microscopy. Bacterial killing test and paraffin oil phagocytosis. Purified neutrophils ( 107/ml) were incubated a t 37°C for varying periods of time up to roo min with equal numbers of Staphylococcus aureus 502A opsonized with AB negative human serum (Quie et al, 1967). After lysis of the
Glutathione Dejiciency Neutrophils
217
neutrophils, the viable bacteria were enumerated on agar pour plates. Ingestion of lipopolysaccharide-coated paraffin-oil droplets and reduction of nitroblue tetrazolium (NBT) by mixed leucocytes were analysed as previously described (Stossel, 1973). Chemotaxis. Leucocyte migration in uitro was measured with the modified Boyden chamber technique (Hill et al, 197s). 1.5 x 1 0 6 neutrophils were deposited with a cytocentrifuge (Shandon Scientific Co.) on a premoistened 5 pm Millipore filter (Millipore Corporation). The filter was then placed in the Boyden chamber and incubated for 3 h at 37°C in a 10%COz, 90% air incubator. A bacterial chemotatic factor was prepared from a culture filtrate of Escherirhia coli. The chemotatic index was determined from the number of P M N migrating through the filter in 10 high-power objective fields. Metabolic studies. The rate of 14C02 evolution from [ ~ - ~ ~ C ] g l u c owas s e determined by a method previously described (Boxer & Stossel, 1974). The availability of H 2 0 2 following stimulation was quantified by a modification of the technique of Root et a1 (1975). Purified neutrophils (2.5 x ro6/ml) were suspended in KRP to which I mM sodium azide had been added. The cells were incubated at 37°C for 20 min to determine resting H 2 0 2 levels and in the presence of 0.I ml of either latex (1000particles/cell) or Concanavalin A (100 pg/l) for 5 min a t 37°C to determine stimulated activity. Following centrifugation at 400 g, I ml of the supernatant was added to 0.4 ml of scopoletin (final concentration 20 nM) and 0.2 ml of KRP, pH 7.4. H 2 0 2 was measured by reduction in scopoletin fluorescence using a Perkin-Elmer spectrofluorimeter. The rate of iodination was determined by the method of Hakim et al (1975).
Electron microscopy. Cell pellets were fixed for 3 0 min at room temperature in i % glutaraldehyde in 0.1 M sodium cacodylate buffer pH 7.4, washed with buffer, and processed for transmission electron microscopy as previously described (Albertini et al, 1977; Oliver et al, 1976). Thin (silver) sections through neutrophil centrioles were examined in a Philips 3 0 0 electron microscope for the presence or absence of microtubules. Other random sections that were positively identified as neutrophils from nuclear morphology were photographed for observation of nuclear condensation, the presence or absence of mitochondria and examination of general ultrastructural integrity.
RESULTS Despite the marked deficiency of glutathione content of the patient’s neutrophils, ingestion of opsonized lipopolysaccharide paraffin-oil droplets, NBT reduction, and directed cell movement in the Boyden chamber by these cells were normal (Table I). Similarly, the patient’s neutrophils were able to increase glucose oxidation (a measure of hexose monophosphate shunt activity) normally in response to phagocytosis of opsonized zymosan (Table 11). Two glutathione oxidants, diamide and t-butyl hydroperoxide, both of which increase the cell’s requirement for NADPH to reduce the generated oxidized glutathione, also stimulated glucose oxidation normally in the patient’s cells. In comparison to normal neutrophils, however, the patient’s cells showed an abnormally high accumulation of peroxide upon ingestion of opsonized zymosan (Table III), and their bactericidal capacity was impaired (Fig I ) . PMN iodination was also depressed in the glutathione deficient cells (Fig 2). Iodination is dependent on the production of peroxide, myeloper-
Stephen P. Spielberg et a1
21 8
TABLE 1. Phagocytic rates and chemotaxis of glutathione deficient neutrophils Ingestion*
Normal Patient
0.037
NBT redurriont C h e m o t u x i a 0.148 0.259
0.03 I
213+24 233f35
* Initial rate of ingestion of E. coli lipopolysaccharide-coated paraffinail droplets opsonized with fresh homologous serum as mg paraffin oil ingested per 10’ neutrophils/min (mean, two determinations). t Initial rate of NB T reductions following ingestion of opsonized paraffin-oil droplets as mg formazan per 10’ neutrophilslmin (mean, two determinations). $ Chemotactic index: Number of neutrophils in 10 random fieldslnumber of neutrophils ( x I O ~ in ) 0.4 ml delivered to the starting side of the filter, mean fSD. TABLE 11. [ ~-‘~C]Glucose oxidation by glutathione deficient neutrophils ‘‘C02/io6 c e l l s / ~ omin
cpm
Normal
Patient
~~
Resting Zymosan Diamide 5 nM Diamide 50 nM Diamide 200 nM Butylhydroperoxide 5 nM Butylhydroperoxide 50 nM Butylhydroperoxide 200 nM
97+ 7 64f 22 6 3 1 f 1 6 650+129 I75k 33 869+ 69 665+ 3 881k 70 238+ 4 180f 1 5 367+ 2 302k 47 676frz9 211k25
688+50
~
Cells (2.0x Io6/assay) were suspended in 0.9 ml of KRP, pH 7.4, containing glucose and 0.5 pCi [ ~-‘~C]glucose. Test solutions (0.I ml) were added and the suspensions were incubated at 37°C for 30 min. Values expressed as mean fSD.
oxidase activity, and the cells’ ability to deliver myeloperoxidase from lysosomes to the phagosome. The patient’s iodination was markedly decreased despite elevated peroxide content and normal myeloperoxidase activity [ 0 . D . 4 4 0 n m = z . ~ / mprotein/min; g normal 3 . 8 & I . I (mean of three+ SD) by a nonautomated modification of the 0-toluidine method of Baggiolini et al (1y6y)], suggesting a n impairment in lysosomal fusion in the patient’s cells. Diamide decreased iodination in the normal and nearly abolished the reaction in glutathione deficient neutrophils, further suggesting a vital role of reduced glutathione in iodination.
Clutathione Deficiency Neutrophils
219
TABLE Ill. Available H 2 0 2 in glutathione deficient neutrophils*
Resting Zymosan phagocytosis
Normal
Pallent
O.I+_O.l
O.I+_O.l
2 . 1 +o.r
3.4+0.2t
* Means+SD are indicated. Values are expressed as nmoles H 2 0 2 / ~ . x5 lo6 neutrophils/min. Different from normal, P
