Deficiency of Alveolar Fluid Glutathione in Patients with Sepsis and the Adult Respiratory Distress Syndrome* Eric R. Pacht, M.D., F.C.C.R;t Anthony R Timerman, Ph.D.;* Michael G. Lykens, M.D., F.C.C.R;§ and A. John Merola, Ph.D.1I
The adult respiratory distress syndrome (ARDS) is a devastating clinical illness characterized by refractory hypoxemia and high-permeability pulmonary edema. Reactive oxygen species such as hydrogen peroxide and hypochlorous acid may playa key role in the pathogenesis of the acute lung injury. Glutathione (GSH) is a tripeptide that is able to react with and effectively neutralize oxidants such as hydrogen peroxide and hypochlorous acid. The present study found that the alveolar epithelial lining Ruid of patients with ARDS was deficient in total GSH compared to normal subjects (21.7 p.mol ± 7. 8p.mol vs 91.8p.mol± 14.5p.mol; p=0.OO2). In addition, if GSH was measured in unconcentrated bronchoalveolar lavage (HAL)
Ruid and indexed to total HAL protein, there was also a deficiency in patients with ARDS compared to nonnal subjects (0.004 ± 0.003 nmol of GSH per microgram of total protein vs 0.026±0.OO5 nmol of GSH per microgram of total protein; p=0.002). Since patients with ARDS are subjected to an increased burden of oxidants in the alveolar Ruid, principally released by recruited neutrophils, this deficiency of GSH may predispose these patients to enhanced lung cell injury. (Che.t 1991; 100:1397-1403)
The adult respiratory distress syndrome (ARDS) is a form of acute lung injury characterized by highpermeability, low-pressure pulmonary edema, refractory hypoxemia, and respiratory failure. 1.2 The pathogenesis of ARDS is complex and likely involves multiple mechanisms and mediators. 1.3.4 There is ample experimental evidence to suggest that reactive oxygen species or oxidants play an important role in the pathogenesis of this disorder. 4.5 Both morphologic and bronchoalveolar lavage (BAL) studies have demonstrated that neutrophils are recruited to the lower respiratory tract of patients with ARDS.6.7 These neutrophils, once activated, are capable of releasing toxic oxygen intermediates that can cause significant lung injury. Toxic oxygen intermediates, such as superoxide anion (02 -), hydrogen peroxide (H 20 2), hydroxyl radical (OH), and hypochlorous acid (HOCI), are capable ofcausing increased vascular permeability in endothelial cell monolayers, 8 in isolated perfused lungs,9 and in intact animals. 9.10 Oxidant-generating systems, infused into isolated rabbit lungs, caused
edema formation similar to that seen in ARDS.ll H 20 2-generating systems, instilled into the tracheas of experimental animals, caused an acute lung injury similar to ARDS which could be inhibited, in a dosedependent fashion, by catalase. 12 Other investigators have shown that oxidants are capable of causing direct cellular cytotoxicity to lung cells, which can be partially inhibited by antioxidants. 13 More direct evidence for the role of reactive oxygen species on human ARDS includes two recent studies which have measured H 20 2 in the expiratory condensate of patients with ARDS.14.15 In addition, Cochrane et al l6 demonstrated that increased concentrations of oxidized and thus inactivated ai-antitrypsin was found in the BAL fluid of patients with ARDS. Thus, it seems apparent that reactive oxygen species, primarily neutrophilderived H 20 2 and hypochlorous acid, play an important role in the pathogenesis of ARDS. Glutathione (GSH) is the most abundant nonprotein thiol in living organisms and is essential for a number of vital biologic functions, including synthesis of proteins and DNA, transport of amino acids, enzyme activity, metabolism, and protection of cells. 17.18 Most important, this tripeptide (L-\fJ-glutamyl-L-cysteinylglycine) can function as an antioxidant by acting as a reductant in the presence of glutathione peroxidase to reduce H 20 2 or other hydroperoxides to less toxic substances. 17-19 Recently, reduced GSH has been detected in high concentrations in the extracellular epithelial lining fluid (ELF) of the lower respiratory tract of normal subjects. 20 In additional studies, a deficiency of GSH was noted in the ELF of patients
*From the Division of Pulmonary and Critical Care, Department of Internal Medicine, and the Department of Physiological Chemistry, Ohio State University, Columbus Supported by grant 21571-55-00 from the American Lung Association and grant 8707 from the Bremer Foundation. tAssistant Professor of Medicine. :l:Research Technician. Currently with Department of Molecular Biology, Vanderbilt University, Nashville, Tenn. §Pulmonary Fello~ Currently with Section of Pulmonary and Critical Care Medicine, Winston-Salem, NC. I\Professor of Physiological Chemistry. Manuscript received December 26; revision accepted March 8. Reprint requests: Dr. lbcht, N-325 Means Hall, 1654 Upham Drive, Columbus, Ohio 43210
GSH = glutathione; GSSG = oxidized GSH; ELF = epithelial lining 8uid; IPF = idiopathic pulmonary fibrosis
CHEST I 100 I 5 I NOVEMBER, 1991
1397
with idiopathic pulmonary fibrosis (IPF)21 and asymptomatic HIV-positive patients. 22 In patients with IPF, high levels of toxic oxygen species are spontaneously released by neutrophils and other in8ammatory cells recruited to the lower respiratory tract. These oxidants can react with myeloperoxidase present in the alveolar fluid of these patients to form hypochlorous acid and other oxidants and to induce lung cell damage. 23 An analogous situation occurs in ARDS, as myeloperoxidase has also been found in high concentrations in the alveolar 8uid. 7 Glutathione is able to react with and scavenge H 20 2 and HOCI,24 perhaps the key oxidants in lung cell damage and injury, and thus may be important in lung cell protection. Thus, GSH in the ELF may constitute the first line of defense in the protection of lung cells from oxidant-mediated damage. Any defect or breach in that defense may be of critical importance in the protection of the lung parenchymal cells, particularly in ARDS. The present study addresses the hypothesis that the GSH in lung ELF may be deficient in ARDS. Similar to the situation seen in IPF, the lung in ARDS is subjected to an increased burden of oxidants, principally those released by neutrophils and other inflammatory cells recruited to the lower respiratory tract. This oxidant burden could produce a relative deficiency of GSH which could contribute to lung cell injury and the pathogenesis of ARDS. In this study the data indeed suggest that there is a deficiency of total GSH in the alveolar ELF of patients with ARDS. MATERIALS AND METHODS
Subjects and Protocol Young, healthy nonsmoking volunteers were recruited from the campus of Ohio State University. All of the subjects had normal 6ndings on history, physical examination, chest roentgenogram, and forced expiratory spirometry. The subjects were not receiving any medications or drugs. There were six male and four female subjects, with an average age of 25 ± 2 years. Ten patients with ARDS were recruited from the medical intensive care unit at Ohio State University Hospital. There were seven women and three men, with an age range of 21 to 59 years (average age of 38 ± 5 years). Standard diagnostic criteria for ARDS were employedi."S and included the following: (1) acute hypoxemic respiratory failure requiring mechanical ventilation; (2) diffuse bilateral alveolar in61trates on the chest roentgenogram; (3) refractory hypoxemia with a Flo2 of more than 0.60 to maintain a POI of more than 60 mm Hg; (4) recognized appropriate clinical setting or risk factor for the development of ARDS; and (5) evidence of highpermeability, low-pressure pulmonary edema. AU ten patients had either sepsis or the sepsis syndrome (Table 1). The lung injury score was determined on all patients with ARDS according to the method of Murray et a1.- BrieRy, the lung injury score assigns a value of 0 to 4 for increasing severity of alveolar in6ltrates seen on the chest roentgenogram, PaO/Flo2 ratio, amount of positive end-expiratory pressure (PEEP) required, and respiratory system compliance. The total score is divided by the number of parameters scored. A score of 0 is considered "no lung injury," while a score of 0.1 to 2.5 is considered mild to moderate lung injury. A lung injury score of greater than 2.5 is considered severe lung injury. All ten patients had all four parameters scored; and in seven of the ten subjects, the
1398
lung injury score was 2.5 or greater. All subjects were studied under an approved protocol of the Ohio State University Human Subjects Review Committee. Both normal volunteers and subjects with ARDS underwent simultaneous cOllection of BAL Ruid and serum (see subsequent explanation).
Bronchoscopy with BAL Bronchoalveolar lavage was performed as previously described.1:1 Subjects received an intramuscular injection of atropine (0.6 to 0.8 mg) 30 minutes prior to bronchoscopy. Some patients received midazolam (1 to 5 mg) intravenously prior to bronchoscopy. The mouth, pharynx, nose, and throat were anesthetized with 4 percent lidocaine (Xylocaine) in the normal volunteer or control group. All patients were monitored with continuous 6ngertip oximet~ Patients with ARDS received supplemental oxygen to maintain the oxygen saturation greater than 90 percent throughout the procedure. A 8exible 6beroptic bronchoscope (Olympus BF-4B2) was passed transnasally in the control group and via the endotracheal tube in the group with ARDS. The bronchoscope was passed into a subsegmental bronchus ofeither the right middle lobe or the lingula in aU subjects. Once the bronchoscope was wedged, 100 ml of sterile saline solution (5 x 2o-ml aliquot) was injected through the suction port into the distal alveolar spaces and immediately aspirated into 5O-ml suction traps under continuous low suction (SO mm Hg of negative pressure). The 8uid was immediately placed on ice for use in later assays. All patients with ARDS underwent BAL within 24 hours of the onset of respiratory failure.
BAL Processing The BAL 8uid was immediately 6ltered through coarse surgical gauze and centrifuged (500g for 15 minutes) to separate cellular and noncellular elements. The supernatant was stored at - SOOC in small-volume a1iquots until used in later assays. Once the supernatant was decanted, the cell pellet was suspended in 10 ml of Hanks' balanced salt solution without calcium or magnesium. The total cell count was determined by hemocytometer. A small aliquot was cytocentrifuged (35g for 10 min), air-dried, and stained by a modified Wright-Giemsa stain. A differential ceO count was performed on a minimum of 300 cells.
Measurement of BAL Proteins The total protein concentration in the BAL Ruid was determined by a modi6cation of a commercially available protein microassay procedure (Bio-Rad Laboratories). 28
Measurement of ELF Volumes Since BAL Ruid dilutes lung ELF, it was 6rst necessary to measure the volume of ELF recovered by the lavage procedure.
Table 1-Clinical Profile of Patients with ABDS Patient, Sex, Age (yr) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
F,59 M, 41 F,21 M,59 F,46 M,25 F,45 F,23 F,36 F,27
Risk Factor
Lung Injury Score
Outcome
Sepsis Sepsis syndrome Sepsis syndrome Sepsis Sepsis Sepsis syndrome Sepsis syndrome Sepsis Sepsis Sepsis syndrome
3.33 2.33 2.67 2.67 2.75 2.0 2.5 3.33 2.5 2.5
Survived Survived Survived Died Survived Survived Survived Died Died Survived
Deficiency of Alveolar Fluid Glutathione in ARCS (Pacht et aI)
Although Sf'gnwnlal BAL sampl,'s airways s",·rPlions. as w,'l1 as a1v....lar ELF. II... surfa,·,' area ..I' Ih.. alveoli is 1(1) limes grealer Ihan Ih.. surfael' art'a of II... airways distal to th.. tip ..I' the hroueho""p.,. Thus. tl... vasI majority of the r..,,,vered lowerrespiratory-trat'! fluid r..pr,·sents alv..olar ELF.~' TI... volullle of ELF was "slimalt'd hy us.. of tilt, lu..a dilutiou tet·hnillue.'"' Un'a
The adult respiratory distress syndrome (ARDS) is a devastating clinical illness characterized by refractory hypoxemia and high-permeability pulmonary edema. Reactive oxygen species such as hydrogen peroxide and hypochlorous acid may playa key role in the pathogenesis of the acute lung injury. Glutathione (GSH) is a tripeptide that is able to react with and effectively neutralize oxidants such as hydrogen peroxide and hypochlorous acid. The present study found that the alveolar epithelial lining Ruid of patients with ARDS was deficient in total GSH compared to normal subjects (21.7 p.mol ± 7. 8p.mol vs 91.8p.mol± 14.5p.mol; p=0.OO2). In addition, if GSH was measured in unconcentrated bronchoalveolar lavage (HAL)
Ruid and indexed to total HAL protein, there was also a deficiency in patients with ARDS compared to nonnal subjects (0.004 ± 0.003 nmol of GSH per microgram of total protein vs 0.026±0.OO5 nmol of GSH per microgram of total protein; p=0.002). Since patients with ARDS are subjected to an increased burden of oxidants in the alveolar Ruid, principally released by recruited neutrophils, this deficiency of GSH may predispose these patients to enhanced lung cell injury. (Che.t 1991; 100:1397-1403)
The adult respiratory distress syndrome (ARDS) is a form of acute lung injury characterized by highpermeability, low-pressure pulmonary edema, refractory hypoxemia, and respiratory failure. 1.2 The pathogenesis of ARDS is complex and likely involves multiple mechanisms and mediators. 1.3.4 There is ample experimental evidence to suggest that reactive oxygen species or oxidants play an important role in the pathogenesis of this disorder. 4.5 Both morphologic and bronchoalveolar lavage (BAL) studies have demonstrated that neutrophils are recruited to the lower respiratory tract of patients with ARDS.6.7 These neutrophils, once activated, are capable of releasing toxic oxygen intermediates that can cause significant lung injury. Toxic oxygen intermediates, such as superoxide anion (02 -), hydrogen peroxide (H 20 2), hydroxyl radical (OH), and hypochlorous acid (HOCI), are capable ofcausing increased vascular permeability in endothelial cell monolayers, 8 in isolated perfused lungs,9 and in intact animals. 9.10 Oxidant-generating systems, infused into isolated rabbit lungs, caused
edema formation similar to that seen in ARDS.ll H 20 2-generating systems, instilled into the tracheas of experimental animals, caused an acute lung injury similar to ARDS which could be inhibited, in a dosedependent fashion, by catalase. 12 Other investigators have shown that oxidants are capable of causing direct cellular cytotoxicity to lung cells, which can be partially inhibited by antioxidants. 13 More direct evidence for the role of reactive oxygen species on human ARDS includes two recent studies which have measured H 20 2 in the expiratory condensate of patients with ARDS.14.15 In addition, Cochrane et al l6 demonstrated that increased concentrations of oxidized and thus inactivated ai-antitrypsin was found in the BAL fluid of patients with ARDS. Thus, it seems apparent that reactive oxygen species, primarily neutrophilderived H 20 2 and hypochlorous acid, play an important role in the pathogenesis of ARDS. Glutathione (GSH) is the most abundant nonprotein thiol in living organisms and is essential for a number of vital biologic functions, including synthesis of proteins and DNA, transport of amino acids, enzyme activity, metabolism, and protection of cells. 17.18 Most important, this tripeptide (L-\fJ-glutamyl-L-cysteinylglycine) can function as an antioxidant by acting as a reductant in the presence of glutathione peroxidase to reduce H 20 2 or other hydroperoxides to less toxic substances. 17-19 Recently, reduced GSH has been detected in high concentrations in the extracellular epithelial lining fluid (ELF) of the lower respiratory tract of normal subjects. 20 In additional studies, a deficiency of GSH was noted in the ELF of patients
*From the Division of Pulmonary and Critical Care, Department of Internal Medicine, and the Department of Physiological Chemistry, Ohio State University, Columbus Supported by grant 21571-55-00 from the American Lung Association and grant 8707 from the Bremer Foundation. tAssistant Professor of Medicine. :l:Research Technician. Currently with Department of Molecular Biology, Vanderbilt University, Nashville, Tenn. §Pulmonary Fello~ Currently with Section of Pulmonary and Critical Care Medicine, Winston-Salem, NC. I\Professor of Physiological Chemistry. Manuscript received December 26; revision accepted March 8. Reprint requests: Dr. lbcht, N-325 Means Hall, 1654 Upham Drive, Columbus, Ohio 43210
GSH = glutathione; GSSG = oxidized GSH; ELF = epithelial lining 8uid; IPF = idiopathic pulmonary fibrosis
CHEST I 100 I 5 I NOVEMBER, 1991
1397
with idiopathic pulmonary fibrosis (IPF)21 and asymptomatic HIV-positive patients. 22 In patients with IPF, high levels of toxic oxygen species are spontaneously released by neutrophils and other in8ammatory cells recruited to the lower respiratory tract. These oxidants can react with myeloperoxidase present in the alveolar fluid of these patients to form hypochlorous acid and other oxidants and to induce lung cell damage. 23 An analogous situation occurs in ARDS, as myeloperoxidase has also been found in high concentrations in the alveolar 8uid. 7 Glutathione is able to react with and scavenge H 20 2 and HOCI,24 perhaps the key oxidants in lung cell damage and injury, and thus may be important in lung cell protection. Thus, GSH in the ELF may constitute the first line of defense in the protection of lung cells from oxidant-mediated damage. Any defect or breach in that defense may be of critical importance in the protection of the lung parenchymal cells, particularly in ARDS. The present study addresses the hypothesis that the GSH in lung ELF may be deficient in ARDS. Similar to the situation seen in IPF, the lung in ARDS is subjected to an increased burden of oxidants, principally those released by neutrophils and other inflammatory cells recruited to the lower respiratory tract. This oxidant burden could produce a relative deficiency of GSH which could contribute to lung cell injury and the pathogenesis of ARDS. In this study the data indeed suggest that there is a deficiency of total GSH in the alveolar ELF of patients with ARDS. MATERIALS AND METHODS
Subjects and Protocol Young, healthy nonsmoking volunteers were recruited from the campus of Ohio State University. All of the subjects had normal 6ndings on history, physical examination, chest roentgenogram, and forced expiratory spirometry. The subjects were not receiving any medications or drugs. There were six male and four female subjects, with an average age of 25 ± 2 years. Ten patients with ARDS were recruited from the medical intensive care unit at Ohio State University Hospital. There were seven women and three men, with an age range of 21 to 59 years (average age of 38 ± 5 years). Standard diagnostic criteria for ARDS were employedi."S and included the following: (1) acute hypoxemic respiratory failure requiring mechanical ventilation; (2) diffuse bilateral alveolar in61trates on the chest roentgenogram; (3) refractory hypoxemia with a Flo2 of more than 0.60 to maintain a POI of more than 60 mm Hg; (4) recognized appropriate clinical setting or risk factor for the development of ARDS; and (5) evidence of highpermeability, low-pressure pulmonary edema. AU ten patients had either sepsis or the sepsis syndrome (Table 1). The lung injury score was determined on all patients with ARDS according to the method of Murray et a1.- BrieRy, the lung injury score assigns a value of 0 to 4 for increasing severity of alveolar in6ltrates seen on the chest roentgenogram, PaO/Flo2 ratio, amount of positive end-expiratory pressure (PEEP) required, and respiratory system compliance. The total score is divided by the number of parameters scored. A score of 0 is considered "no lung injury," while a score of 0.1 to 2.5 is considered mild to moderate lung injury. A lung injury score of greater than 2.5 is considered severe lung injury. All ten patients had all four parameters scored; and in seven of the ten subjects, the
1398
lung injury score was 2.5 or greater. All subjects were studied under an approved protocol of the Ohio State University Human Subjects Review Committee. Both normal volunteers and subjects with ARDS underwent simultaneous cOllection of BAL Ruid and serum (see subsequent explanation).
Bronchoscopy with BAL Bronchoalveolar lavage was performed as previously described.1:1 Subjects received an intramuscular injection of atropine (0.6 to 0.8 mg) 30 minutes prior to bronchoscopy. Some patients received midazolam (1 to 5 mg) intravenously prior to bronchoscopy. The mouth, pharynx, nose, and throat were anesthetized with 4 percent lidocaine (Xylocaine) in the normal volunteer or control group. All patients were monitored with continuous 6ngertip oximet~ Patients with ARDS received supplemental oxygen to maintain the oxygen saturation greater than 90 percent throughout the procedure. A 8exible 6beroptic bronchoscope (Olympus BF-4B2) was passed transnasally in the control group and via the endotracheal tube in the group with ARDS. The bronchoscope was passed into a subsegmental bronchus ofeither the right middle lobe or the lingula in aU subjects. Once the bronchoscope was wedged, 100 ml of sterile saline solution (5 x 2o-ml aliquot) was injected through the suction port into the distal alveolar spaces and immediately aspirated into 5O-ml suction traps under continuous low suction (SO mm Hg of negative pressure). The 8uid was immediately placed on ice for use in later assays. All patients with ARDS underwent BAL within 24 hours of the onset of respiratory failure.
BAL Processing The BAL 8uid was immediately 6ltered through coarse surgical gauze and centrifuged (500g for 15 minutes) to separate cellular and noncellular elements. The supernatant was stored at - SOOC in small-volume a1iquots until used in later assays. Once the supernatant was decanted, the cell pellet was suspended in 10 ml of Hanks' balanced salt solution without calcium or magnesium. The total cell count was determined by hemocytometer. A small aliquot was cytocentrifuged (35g for 10 min), air-dried, and stained by a modified Wright-Giemsa stain. A differential ceO count was performed on a minimum of 300 cells.
Measurement of BAL Proteins The total protein concentration in the BAL Ruid was determined by a modi6cation of a commercially available protein microassay procedure (Bio-Rad Laboratories). 28
Measurement of ELF Volumes Since BAL Ruid dilutes lung ELF, it was 6rst necessary to measure the volume of ELF recovered by the lavage procedure.
Table 1-Clinical Profile of Patients with ABDS Patient, Sex, Age (yr) 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
F,59 M, 41 F,21 M,59 F,46 M,25 F,45 F,23 F,36 F,27
Risk Factor
Lung Injury Score
Outcome
Sepsis Sepsis syndrome Sepsis syndrome Sepsis Sepsis Sepsis syndrome Sepsis syndrome Sepsis Sepsis Sepsis syndrome
3.33 2.33 2.67 2.67 2.75 2.0 2.5 3.33 2.5 2.5
Survived Survived Survived Died Survived Survived Survived Died Died Survived
Deficiency of Alveolar Fluid Glutathione in ARCS (Pacht et aI)
Although Sf'gnwnlal BAL sampl,'s airways s",·rPlions. as w,'l1 as a1v....lar ELF. II... surfa,·,' area ..I' Ih.. alveoli is 1(1) limes grealer Ihan Ih.. surfael' art'a of II... airways distal to th.. tip ..I' the hroueho""p.,. Thus. tl... vasI majority of the r..,,,vered lowerrespiratory-trat'! fluid r..pr,·sents alv..olar ELF.~' TI... volullle of ELF was "slimalt'd hy us.. of tilt, lu..a dilutiou tet·hnillue.'"' Un'a
