Lipid peroxidation and antielastase activity in the lung under oxidant stress: role of antioxidant defenses VAHID MOHSENIN The John B. Pierce Laboratory and Department New Haven, Connecticut 06519 MOHSENIN, VAHID. Lipid peroxidation and antielastase activity in the lung under oxidant stress: role of antioxidant defenses. J. Appl. Physiol. 70(4): 1456-1462,1991.-The role of lipid peroxidation in the inactivation of qprotease inhibitor (q-PI) in the alveolar lining fluid of human subjects has been examined under oxidant stress. Exposure to nitrogen dioxide (NO,) at 4 ppm for 3 h resulted in a significant increase in the amount of lipid peroxidation products in the alveolar lining fluid, with conjugated dienes the predominant species. Four-week supplementation with vitamins C and E before NO, exposure markedly decreased the levels of conjugated dienes (control 804 k 103 pmollpg total phospholipids vs. vitamin-supplemented 369 t 58, P = 0.003). Malondialdehydes, although detectable in the lavage fluid, contributed little to the total amount of lipid peroxidation products, and the levels were comparable in both groups. NO, exposure in the absence of vitamin supplementation caused a significant decrease in the elastase inhibitory capacity (EIC) of the alveolar lining fluid in the control group but not in the vitamin-supplemented group [control 3.67 t 0.32 pg q-PIlpg porcine pancreatic elastase (PPE) vs. vitamin-supplemented 2.75 t 0.17, P < 0.031. The vitamin-supplemented group had a lower level of conjugated dienes and a higher EIC. Conversely, the control group had higher levels of conjugated dienes and a lower EIC in their lavage fluid. These studies demonstrate that lipid peroxidation occurs as an early event during oxidant exposure in the lungs of normal subjects. The appearance of lipid peroxidation products in the lavage fluid is associated with a decrease in the EIC of the alveolar lining fluid. Vitamins C and E diminish lipid peroxidation and preserve the EIC of the lower respiratory tract fluid during oxidant stress.
elastase inhibitory capacity; bronchoalveolar lavage; porcine pancreatic elastase; nitrogen dioxide; cr,-protease inhibitor; phospholipid
ALPHA,-protease inhibitor (q-PI), the major plasma and lung protease inhibitor of elastase, is important in protecting the lung from proteolytic damage, particularly from the elastase of neutrophils. It is generally accepted that lung destruction in emphysema occurs because of imbalance between proteases and antiproteases (19). A reduced level of elastase inhibitory activity in the lung can result from the partial oxidation of (w,PI through inhalation of oxidants present in tobacco smoke (10, 16) and nitrogen dioxide (NO,) (27)) as well as from oxidants released by lung macrophages and other phagocytic cells (25). There is increasing evidence that an imbalance of oxidants and antioxidants in the lower respiratory tract contributes to this process (47). Oxidants, such as superoxide anion radical (0,)) hydrogen peroxide (H,O,) ,-hy1456
0161-7567/91
$1.50 Copyright
of Medicine, Yale University School of Medicine,
droxyl radical (OH*), and hypochlorite (OCl-), are generated in the lower respiratory tract as a result of normal biochemical processes, activation of inflammatory cells, and inhaled toxic gases. We have previously reported inactivation of q-PI in the lower respiratory tract fluid of healthy nonsmoker subjects after exposure to NO, (27). Chronic exposure to NO, in animals leads to emphysematous changes in the lung (20). Under normal circumstances, the parenchymal cells are protected by intracellular antioxidants and membrane radical scavengers. If the oxidant burden overcomes these defenses, the parenchymal cells may be injured, the connective tissue matrix may be partially degraded, and the antiprotease defense that protects the lower respiratory tract from attack by neutrophil elastase may be rendered impotent. Because oxidant exposure decreasesthe elastase inhibitory capacity (EIC) of the q-PI, one possible therapeutic approach involves the augmentation of antioxidant protection at the level of alveolar structure. We have previously demonstrated that combination of a-tocopherol and ascorbic acid protected the q-PI in vitro from an oxidation attack of peroxidized lipids (28). The significant roles of cw-tocopherol and ascorbic acid in the protection of the lung from oxidants can be inferred from the levels of these vitamins in cigarette smokers. The alveolar fluid of smokers is deficient in vitamin E (33), and many studies have shown that, on average, the level of serum vitamin C in heavy smokers is lower than in nonsmokers (11, 41). Although cw-tocopherol and ascorbic acid have been extensively evaluated in animal models of oxidant lung injury and in in vitro experiments, little information is available about their role in human oxidant exposure. To test the hypothesis that lipid peroxidation products mediate the oxidant-induced inactivation of cu,-PI in vivo, nonsmokers were evaluated by bronchoalveolar lavage after a 4-wk course of vitamin C and vitamin E supplementation or placebo and after 3-h exposure to NO,. These studies provided evidence that the lower respiratory tract fluid of subjects exposed to NO, contains increased levels of lipid peroxidation products and partially inactivated q-PI, which reverse with antioxidant supplementation. METHODS
Subjects. Nineteen healthy subjects (10 male, 9 female) 21-33 yr of age participated in the study. Criteria for exclusion from the study were history of asthma, sea-
0 1991 the American
Physiological
Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
OXIDANT-TNDIJCED
LIPID
PEROXIDATION
sonal allergic rhinitis, cigarette smoking, acute respiratory illness within the past 6 wk, or use of supplemental vitamins. All candidates underwent screening procedures including a history, physical examination, and determination of serum vitamin C and vitamin E levels. All subjects signed an informed consent previously approved by the Yale University Human Investigation Committee. Study design. The protocol was designed as a single blind placebo-controlled study comparing placebo with vitamin E (dL-cu-tocopherol) and vitamin C (L-ascorbic acid). Placebo and vitamin capsules were indistinguishable and were a generous gift of Hoffmann-La Roche (Nutley, NJ). Subjects were randomly divided into placebo or vitamin groups. Each subject took a 4-wk course of either placebo or vitamins (1,500 mg vitamin C/day and 1,200 IU vitamin E/day). After the 4wk period, subjects underwent a 3-h exposure to 4 ppm NO, as previously described (27). Before exposure serum samples were collected, and immediately after exposure bronchoalveolar lavage was performed. All samples were stored at -70°C until analysis. Bronchoalueolar lauage. Fiber-optic bronchoscopy and bronchoalveolar lavage (BAL) were performed as previously described (27). BAL was performed on room air, and subjects did not receive supplemental 0,. Briefly, after topical anesthesia with 2% lidocaine, a flexible fiber-optic bronchoscope (model 5 BF2, Olympus, New Hyde Park, NY) was advanced via mouth or nose into the right middle lobe until a wedge position was achieved. Lavage was performed with 300 ml of sterile 0.9% NaCl (room temperature) in 50-ml aliquots; the fluid was recovered by gentle aspiration through a syringe and kept on ice until further processing. The recovered BAL fluid was filtered through two layers of coarse gauze to remove mucus. Filtered lavage fluid was centrifuged at 900 g for 10 min at 4°C to separate cellular and noncellular components. The BAL supernatants were stored under N, at -70°C in lo-ml aliquots until further analysis. The cell pellets were washed once with Hanks’ balanced salt solution free of Mg2+ and Ca2+ (GIBCO, Grand Island, NY) and counted with a hemocytometer, and a small aliquot was used to prepare a cytocentrifuge preparation. The slides were air-dried and then stained with Diff-Quick (American Hospital Supply, McGaw Park, IL). Cell differentials were performed by counting 500 cells. Viability was assessedby measuring the ability of cells to exclude 0.4% trypan blue. Quantitation of tu,-PI. Concentration of cu,-PI in BAL fluid was measured in the concentrated BAL using radial immunodiffusion or rocket immunoelectrophoresis as previously described (27). Briefly, aliquots of BAL supernatant were concentrated l5- to 2O-fold by positive pressure ultrafiltration with nitrogen at 4°C using an Amicon YMlO membrane that excluded substances with molecular weights greater than 10,000 (Amicon, Lexington, MA). The amount of ti,-PI in concentrated BAL was measured in triplicate using human cu,-PI as the standard (Calbiochem, La Jolla, CA). Elastase inhibitory capacity of BAL. The EIC of BAL fluid was determined as previously described (27, 28). Briefly, serial dilutions of concentrated BAL were incu-
IN HUMAN
LUNG
1457
bated for 30 min at 25°C with constant amounts of chromatographically purified and active site-titrated porcine pancreatic elastase (PPE; Elastin Products, Pacific, MO). The residual elastase activity was measured spectrofluorometrically with methoxysuccinyl-Ala,-Pro-Valaminomethylcoumarin (Enzyme Systems Products, Livermore, CA) as substrate. The initial rate of increase in fluorescence due to the release of aminomethyl-coumarin was monitored kinetically at X,, of 370 and X,, of 460 nm. Immunoreactive cy,-PI concentration in each sample was used to construct an inhibition curve for each sample. Results are expressed as the amount of a,-PI in micrograms required for complete inhibition of 1.0 pg of elastase or as percentage of the theoretical EIC value of 2.12 pg cu,-PIlpg PPE based on an equimolar reaction between a pure and fully active PPE with fully active (Y,-PI (3). In practice, however, this theoretical value is rarely obtained in commercially available cu,-PI. Internal and external standards were run simultaneously using a purified cu,-PI for the construction of a standard curve. Lipid extraction of BAL. A portion of BAL fluid supernatant was used for lipid extraction according to the method of Bligh and Dyer (6). Briefly, 50 ml of BAL fluid supernatant were added to a mixture of methanol and chloroform with a ratio of 0.8:2.0:1.0 (vol/vol/vol) in a separating funnel. The mixture was shaken to form an emulsion, which was extracted with chloroform and water (l.O:l.O, vol/vol). After separation of phases, the lower phase, which now contained most of the lipid, was brought to dryness on a rotary evaporator under N,. The residue was washed three times with a solution of chloroform-methanol-water (1:16:15.7, vol/vol/vol). The washed and dried residue was kept at -7OOC under N, until further analysis. Determination of malondialdehyde. To determine the thiobarbituric acid reactants in the lipid extract of BAL, 0.5 ml of the sample in 0.15 M NaCl was added to an equal volume of 1:l:l (vol/vol/vol) solution of trichloroacetic acid (15%, wt/vol), thiobarbituric acid (0.37%, wt/ vol), and hydrochloric acid (0.25 N) and heated at 100°C for 25 min. After cooling and centrifugation at 1,500 g for 5 min, the absorbance of the purple color supernatant was measured at 531 nm against the solvent (5). The concentration of thiobarbituric acid reactants was calculated from a standard curve of malonaldehyde bisdimethylacetal. The results were expressed as picomoles per microgram phospholipid. Conjugated dienes. Lipid extract of BAL fluid was dis solved in spectrophotometric grade cyclohexane, and the absorbance was measured at 233 nm against cyclohexane blank. The concentration of conjugated dienes was calculated using an extinction coefficient of 28,000 M/ cm (35). BAL phospholipids. Phospholipid phosphorus was determined in the lipid extract of BAL fluid after sulfuric acid digestion, and the results were multiplied by 25 to obtain the amount of phospholipid (2, 14). Quantification of albumin in BAL. Albumin in concentrated BAL was measured by radial immunodiffusion (24) using human serum albumin as standard (Calbiothem, La Jolla, CA). Ascorbic acid assay. Serum samples taken just before
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
1458
OXIDANT-INDUCED
LIPID
PEROXIDATION
1. Characteristics of subjects and lung lavage fluid composition
TABLE
Placebo
9
n
Age, yr Males/females Fluid recovered, % Total cells, X106 Macrophages, % Lymphocytes, % Neutrophils, % Cell viability, %
24t2 514 69+3 14.Ok1.7 90.8k2.1 8.6k2.0 0.6fO.l 88+3
Vitamin Supplemented 10 26rtl 515 72+1 16.2k4.2 91.3f2.1 8.2+2.1 0.5fO.l 89+2
Values are means + SE.
each exposure were analyzed for ascorbic acid content calorimetrically using reduction of 2,6dichlorophenolindophenol as an indicator (32). Vitamin E assay. Vitamin E levels were determined by high-pressure liquid chromatography according to Craft et al. (12). Statistical methods. All data were expressed as means & SE for each experimental and control group. Statistical differences between groups were determined by Student’s t test. Differences were considered significant if P < 0.05. RESULTS
IN HUMAN
LUNG
significant effect on the antigenic cu,-PI, whether it was expressed per original lavage volume or albumin (Fig. 2). EIC of BAL fluid. To determine whether vitamin C and vitamin E supplementation could prevent the reduction in the EIC of the lavage fluid by NO,, concentrated BAL fluid samples were assayed for their ability to inhibit PPE. Exposure to 4 ppm NO, for 3 h decreased the EIC of lavage fluid by 42% in the placebo group (Fig. 3) in contrast to a 23% decrease in the vitamin-supplemented group; the difference between the two groups was highly significant (58 it 9 vs. 77 + 6%, P < 0.03). The EIC of the vitamin-supplemented group (2.75 + 0.17 pg a,-PIlpg PPE) was similar to that of the control air-exposed group reported previously (2.78 ~fr0.41)(27). Both of these values were significantly lower (i.e., higher EIC) than the placebo-treated NO,-exposed group (3.67 + 0.32, P < 0.03). Lipid peroxidation products in BAL fluid. To examine
the effects of lipid peroxidation products on the EIC of BAL fluid, portions of lung lavage fluid were extracted for lipids, and the lipid peroxidation products were measured in the lipid extract as malondialdehyde and conjugated dienes. The lipid peroxidation products were expressed per total lavage fluid phospholipids. The amount of malondialdehyde in the lipid fraction of the lavage fluid of the placebo group was 10.2 + 3.6 pmollpg phospholipid as opposed to 8.5 + 3.3 pmollpg in the vitaminsupplemented group. Although malondialdehyde concentration tended to be higher in the placebo group, the difference was not statistically significant. However, the levels of conjugated dienes in the NO,-exposed group without vitamin supplementation (placebo) was significantly higher than those in the vitamin-supplemented group (P = 0.003; Table 2). There were approximately SO- and 40-fold larger amounts of conjugated dienes than malondialdehyde in the lavage fluid of placebo and vitamin-supplemented groups, respectively. There was no
Bronchoalveolar lavage. Table 1 shows the characteristics of the subjects and the cellular composition of the BAL. There were no significant differences in the subject characteristics. The NO, exposure and the bronchoalveolar lavage were well tolerated by all subjects. There was no significant difference in the volume of recovered fluid between the placebo and vitamin-supplemented groups. NO, did not increase the total cell numbers or alter the differential cell count in either group. The percentage and the absolute number of neutrophils did not differ from our air-exposed control group reported previously (27). Vitamins C and E had no discernible effect on the total cell numbers or the differential cell counts. There was no significant difference in the cell viability between the two groups, and both were greater than 88%. Serum vitamin C and vitamin E levels. Serum levels of both vitamins C and E were significantly higher in the group with supplementation than in the placebo group. The placebo group had mean levels of 1.1 * 0.2 and 1.01 rfr 0.06 mg/dl for vitamin C and vitamin E, respectively, which were not significantly different from those reported for normal nonsmoking subjects (26). With supplementation, the ascorbic acid increased more than twofold, and vitamin E increased more than threefold (Fig. 1). Albumin and antigenic a,-PI. The concentration of albumin in the lavage fluid was determined as a measure of increased permeability of pulmonary vascular endothelium and alveolar epithelial lining. NO, exposure failed SUPP. SUPP. to increase the albumin concentration in the lavage fluid. FIG. 1. Serum levels of ascorbic acid and a-tocopherol Likewise, vitamin C and vitamin E supplementation had and vitamin C- and vitamin E-supplemented groups. Vitaminin placebo suppleno effect on the lavage albumin. Similarly, neither NO, mentation for 4 wk increased serum ascorbic acid more than twice and alone nor NO, with vitamin supplementation had any cu-tocopherol more than three times the levels in placebo group.
T
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
OXIDANT-INDUCED A Albumin
LIPID
PEROXIDATION
IN HUMAN
B ojPl
LUNG
1459
C o ,PI / Albumin
50
50
40
40
30
30
20
20
10
10
E Y
0 L
0
FIG. 2. Effect of NO, exposure and vitamin C and E supplementation on alveolar lining fluid albumin and antigenic ol,-protease inhibitor (oc,-PI). Neither NO, exposure nor vitamin supplementation altered concentration of albumin (A) and antigenic a,-PI (B) in lavage fluid. oc,-PI-to-albumin ratio (C) remained unchanged and was comparable between groups.
significant difference in the recovered total phospholipid content of the lavage fluid between the groups. Relationship between lipid peroxidation products and EIC of BAL fluid. To determine the relationship between
lipid peroxidation and EIC, we used a least square regression analysis to examine the interaction of conjugated dienes and malondialdehyde with EIC of BAL fluid. There was a strong correlation between the levels of conjugated dienes and EIC of BAL fluid, r = 0.62, P = 0.008 (Fig. 4). Although there was an overlap in three data points, overall the analysis demonstrated that the higher the conjugated dienes the lower the EIC. In Fig. 4,
0 Placebo
Vitamin
C+E
3. EIC of alveolar lining fluid after NO, exposure in placebo and vitamin C- and E-supplemented groups. Concentrated lavage fluids with known levels of antigenic @,-PI were titrated against chromatographically purified and active site-titrated porcine pancreatic elastase (PPE). Residual elastase activity was measured by fluorogenic substrate as described under METHODS. Amount ofcu,-PI in pg required to fully inhibit 1.0 pg PPE was taken as the elastase inhibitory capacity (EIC) of alveolar lining fluid. NOz exposure decreased EIC of control group by 42%, whereas vitamin-supplemented group retained most of its EIC. EIC of vitamin-supplemented group was comparable to our previously reported data on air-exposed controls (27). Values are means rf: SE and are expressed as percentage of fully active CQ-PI with a FIG.
value
Of 2.12 (See METHODS).
the EIC of BAL fluid is expressed as a percentage of purified and fully active (w,-PI. Figure 4 also shows that vitamin-supplemented subjects as a group not only had lower levels of conjugated dienes but also had higher EIC. In contrast to conjugated dienes, malondialdehyde levels in the lavage fluid did not correlate with the EIC (data not shown). The concentrations of malondialdehyde were 1.3 and 2.3% of respective conjugated diene levels in placebo and vitamin-supplemented groups. DISCUSSION
The data presented in this study demonstrate that exposure of young healthy subjects to low levels of NO, for 3 h results in a significant diminution of EIC of the alveolar lining fluid. This decrease in antiprotease defense of the lower respiratory tract is associated with the appearance of peroxidation products in the surfactant lipids. The relationship between enhanced lipid peroxidation and low EIC of the alveolar lining fluid is further supported by the protective action of vitamins C and E in these subjects in whom vitamins C and E decreased lipid peroxidation and enhanced EIC of the alveolar lining fluid. NO,, a common air pollutant and a major component in cigarette smoke (31), is an oxidant that can cause emphysema in experimental animals after long-term exposure (20). We previously demonstrated that even shortterm exposure (3 h) of healthy subjects to NO, causes a significant reduction in the antiprotease defense of the alveolar lining fluid (27). However, exposure of (u,-PI to NO, in vitro did not result in a decrease of its EIC, even if the level of NO, was 10 times higher than the in vivo 2. Lung lavage phospholipids, malondialdehyde, and conjugated dienes after exposure to nitrogen dioxide
TABLE
Total Phospholipids, &nl
Placebo Vitamin supplemented
BAL
21.3e2.9 23.012.3
Malondialdehyde, pmol/ag
PL
10.2k3.6 F3.5r3.3
Conjugated Dienes, pmol/rg
PL
304+103 369+58*
Values are means + SE. BAL, bronchoalveolar lavage; PL, phospholipids. * Significantly different from placebo group, P = 0.003.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
1460
OXIDANT-INDUCED
LIPID
PEROXIDATION
r = 0.62 P= 0.008 0
a l2
a.
-
.4
L
-
.
-
a
.6 .8 1 Conjugated Dienes (pmoleslmg Phospholipids)
-
I
1.2
FIG. 4. Correlation between conjugated dienes and EIC of bronchoalveolar lavage fluid in placebo (0) and vitamin-supplemented groups (0) after NO, exposure. Placebo group had high levels of conjugated dienes and decreased EIC. Conversely, the vitamin-supplemented group had low levels of conjugated dienes and high levels of EIC.
exposure (28). On the basis of our in vitro experiments, we postulated that NO, exerts i ts inhibitory action against cu,-PI th rough .peroxidation of surfactant and cell membrane lipids (28). Although the major component of surfactant phospholipids is disaturated phosphatidylcholine, a significant portion of phosphatidylcholine contains mono- and polyunsaturated fatty acyl moieties (34,37). The process of lipid peroxidation starts with the initial step of hydrogen abstraction from an unsaturated fatty acid, which forms conjugated dienes or lipid radicals. Lipid radicals subsequently form lipid peroxides in the presence of molecular 0, and eventually lipid endoperoxides. The further reactions lead to the formation of malonaldehydes or malondialdehydes and lipid hydroperoxides. In the current study, conjugated dienes were the predominant species, and thiobarbituric acid reactants (primarily malondialdehydes) constituted
elastase inhibitory capacity; bronchoalveolar lavage; porcine pancreatic elastase; nitrogen dioxide; cr,-protease inhibitor; phospholipid
ALPHA,-protease inhibitor (q-PI), the major plasma and lung protease inhibitor of elastase, is important in protecting the lung from proteolytic damage, particularly from the elastase of neutrophils. It is generally accepted that lung destruction in emphysema occurs because of imbalance between proteases and antiproteases (19). A reduced level of elastase inhibitory activity in the lung can result from the partial oxidation of (w,PI through inhalation of oxidants present in tobacco smoke (10, 16) and nitrogen dioxide (NO,) (27)) as well as from oxidants released by lung macrophages and other phagocytic cells (25). There is increasing evidence that an imbalance of oxidants and antioxidants in the lower respiratory tract contributes to this process (47). Oxidants, such as superoxide anion radical (0,)) hydrogen peroxide (H,O,) ,-hy1456
0161-7567/91
$1.50 Copyright
of Medicine, Yale University School of Medicine,
droxyl radical (OH*), and hypochlorite (OCl-), are generated in the lower respiratory tract as a result of normal biochemical processes, activation of inflammatory cells, and inhaled toxic gases. We have previously reported inactivation of q-PI in the lower respiratory tract fluid of healthy nonsmoker subjects after exposure to NO, (27). Chronic exposure to NO, in animals leads to emphysematous changes in the lung (20). Under normal circumstances, the parenchymal cells are protected by intracellular antioxidants and membrane radical scavengers. If the oxidant burden overcomes these defenses, the parenchymal cells may be injured, the connective tissue matrix may be partially degraded, and the antiprotease defense that protects the lower respiratory tract from attack by neutrophil elastase may be rendered impotent. Because oxidant exposure decreasesthe elastase inhibitory capacity (EIC) of the q-PI, one possible therapeutic approach involves the augmentation of antioxidant protection at the level of alveolar structure. We have previously demonstrated that combination of a-tocopherol and ascorbic acid protected the q-PI in vitro from an oxidation attack of peroxidized lipids (28). The significant roles of cw-tocopherol and ascorbic acid in the protection of the lung from oxidants can be inferred from the levels of these vitamins in cigarette smokers. The alveolar fluid of smokers is deficient in vitamin E (33), and many studies have shown that, on average, the level of serum vitamin C in heavy smokers is lower than in nonsmokers (11, 41). Although cw-tocopherol and ascorbic acid have been extensively evaluated in animal models of oxidant lung injury and in in vitro experiments, little information is available about their role in human oxidant exposure. To test the hypothesis that lipid peroxidation products mediate the oxidant-induced inactivation of cu,-PI in vivo, nonsmokers were evaluated by bronchoalveolar lavage after a 4-wk course of vitamin C and vitamin E supplementation or placebo and after 3-h exposure to NO,. These studies provided evidence that the lower respiratory tract fluid of subjects exposed to NO, contains increased levels of lipid peroxidation products and partially inactivated q-PI, which reverse with antioxidant supplementation. METHODS
Subjects. Nineteen healthy subjects (10 male, 9 female) 21-33 yr of age participated in the study. Criteria for exclusion from the study were history of asthma, sea-
0 1991 the American
Physiological
Society
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
OXIDANT-TNDIJCED
LIPID
PEROXIDATION
sonal allergic rhinitis, cigarette smoking, acute respiratory illness within the past 6 wk, or use of supplemental vitamins. All candidates underwent screening procedures including a history, physical examination, and determination of serum vitamin C and vitamin E levels. All subjects signed an informed consent previously approved by the Yale University Human Investigation Committee. Study design. The protocol was designed as a single blind placebo-controlled study comparing placebo with vitamin E (dL-cu-tocopherol) and vitamin C (L-ascorbic acid). Placebo and vitamin capsules were indistinguishable and were a generous gift of Hoffmann-La Roche (Nutley, NJ). Subjects were randomly divided into placebo or vitamin groups. Each subject took a 4-wk course of either placebo or vitamins (1,500 mg vitamin C/day and 1,200 IU vitamin E/day). After the 4wk period, subjects underwent a 3-h exposure to 4 ppm NO, as previously described (27). Before exposure serum samples were collected, and immediately after exposure bronchoalveolar lavage was performed. All samples were stored at -70°C until analysis. Bronchoalueolar lauage. Fiber-optic bronchoscopy and bronchoalveolar lavage (BAL) were performed as previously described (27). BAL was performed on room air, and subjects did not receive supplemental 0,. Briefly, after topical anesthesia with 2% lidocaine, a flexible fiber-optic bronchoscope (model 5 BF2, Olympus, New Hyde Park, NY) was advanced via mouth or nose into the right middle lobe until a wedge position was achieved. Lavage was performed with 300 ml of sterile 0.9% NaCl (room temperature) in 50-ml aliquots; the fluid was recovered by gentle aspiration through a syringe and kept on ice until further processing. The recovered BAL fluid was filtered through two layers of coarse gauze to remove mucus. Filtered lavage fluid was centrifuged at 900 g for 10 min at 4°C to separate cellular and noncellular components. The BAL supernatants were stored under N, at -70°C in lo-ml aliquots until further analysis. The cell pellets were washed once with Hanks’ balanced salt solution free of Mg2+ and Ca2+ (GIBCO, Grand Island, NY) and counted with a hemocytometer, and a small aliquot was used to prepare a cytocentrifuge preparation. The slides were air-dried and then stained with Diff-Quick (American Hospital Supply, McGaw Park, IL). Cell differentials were performed by counting 500 cells. Viability was assessedby measuring the ability of cells to exclude 0.4% trypan blue. Quantitation of tu,-PI. Concentration of cu,-PI in BAL fluid was measured in the concentrated BAL using radial immunodiffusion or rocket immunoelectrophoresis as previously described (27). Briefly, aliquots of BAL supernatant were concentrated l5- to 2O-fold by positive pressure ultrafiltration with nitrogen at 4°C using an Amicon YMlO membrane that excluded substances with molecular weights greater than 10,000 (Amicon, Lexington, MA). The amount of ti,-PI in concentrated BAL was measured in triplicate using human cu,-PI as the standard (Calbiochem, La Jolla, CA). Elastase inhibitory capacity of BAL. The EIC of BAL fluid was determined as previously described (27, 28). Briefly, serial dilutions of concentrated BAL were incu-
IN HUMAN
LUNG
1457
bated for 30 min at 25°C with constant amounts of chromatographically purified and active site-titrated porcine pancreatic elastase (PPE; Elastin Products, Pacific, MO). The residual elastase activity was measured spectrofluorometrically with methoxysuccinyl-Ala,-Pro-Valaminomethylcoumarin (Enzyme Systems Products, Livermore, CA) as substrate. The initial rate of increase in fluorescence due to the release of aminomethyl-coumarin was monitored kinetically at X,, of 370 and X,, of 460 nm. Immunoreactive cy,-PI concentration in each sample was used to construct an inhibition curve for each sample. Results are expressed as the amount of a,-PI in micrograms required for complete inhibition of 1.0 pg of elastase or as percentage of the theoretical EIC value of 2.12 pg cu,-PIlpg PPE based on an equimolar reaction between a pure and fully active PPE with fully active (Y,-PI (3). In practice, however, this theoretical value is rarely obtained in commercially available cu,-PI. Internal and external standards were run simultaneously using a purified cu,-PI for the construction of a standard curve. Lipid extraction of BAL. A portion of BAL fluid supernatant was used for lipid extraction according to the method of Bligh and Dyer (6). Briefly, 50 ml of BAL fluid supernatant were added to a mixture of methanol and chloroform with a ratio of 0.8:2.0:1.0 (vol/vol/vol) in a separating funnel. The mixture was shaken to form an emulsion, which was extracted with chloroform and water (l.O:l.O, vol/vol). After separation of phases, the lower phase, which now contained most of the lipid, was brought to dryness on a rotary evaporator under N,. The residue was washed three times with a solution of chloroform-methanol-water (1:16:15.7, vol/vol/vol). The washed and dried residue was kept at -7OOC under N, until further analysis. Determination of malondialdehyde. To determine the thiobarbituric acid reactants in the lipid extract of BAL, 0.5 ml of the sample in 0.15 M NaCl was added to an equal volume of 1:l:l (vol/vol/vol) solution of trichloroacetic acid (15%, wt/vol), thiobarbituric acid (0.37%, wt/ vol), and hydrochloric acid (0.25 N) and heated at 100°C for 25 min. After cooling and centrifugation at 1,500 g for 5 min, the absorbance of the purple color supernatant was measured at 531 nm against the solvent (5). The concentration of thiobarbituric acid reactants was calculated from a standard curve of malonaldehyde bisdimethylacetal. The results were expressed as picomoles per microgram phospholipid. Conjugated dienes. Lipid extract of BAL fluid was dis solved in spectrophotometric grade cyclohexane, and the absorbance was measured at 233 nm against cyclohexane blank. The concentration of conjugated dienes was calculated using an extinction coefficient of 28,000 M/ cm (35). BAL phospholipids. Phospholipid phosphorus was determined in the lipid extract of BAL fluid after sulfuric acid digestion, and the results were multiplied by 25 to obtain the amount of phospholipid (2, 14). Quantification of albumin in BAL. Albumin in concentrated BAL was measured by radial immunodiffusion (24) using human serum albumin as standard (Calbiothem, La Jolla, CA). Ascorbic acid assay. Serum samples taken just before
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
1458
OXIDANT-INDUCED
LIPID
PEROXIDATION
1. Characteristics of subjects and lung lavage fluid composition
TABLE
Placebo
9
n
Age, yr Males/females Fluid recovered, % Total cells, X106 Macrophages, % Lymphocytes, % Neutrophils, % Cell viability, %
24t2 514 69+3 14.Ok1.7 90.8k2.1 8.6k2.0 0.6fO.l 88+3
Vitamin Supplemented 10 26rtl 515 72+1 16.2k4.2 91.3f2.1 8.2+2.1 0.5fO.l 89+2
Values are means + SE.
each exposure were analyzed for ascorbic acid content calorimetrically using reduction of 2,6dichlorophenolindophenol as an indicator (32). Vitamin E assay. Vitamin E levels were determined by high-pressure liquid chromatography according to Craft et al. (12). Statistical methods. All data were expressed as means & SE for each experimental and control group. Statistical differences between groups were determined by Student’s t test. Differences were considered significant if P < 0.05. RESULTS
IN HUMAN
LUNG
significant effect on the antigenic cu,-PI, whether it was expressed per original lavage volume or albumin (Fig. 2). EIC of BAL fluid. To determine whether vitamin C and vitamin E supplementation could prevent the reduction in the EIC of the lavage fluid by NO,, concentrated BAL fluid samples were assayed for their ability to inhibit PPE. Exposure to 4 ppm NO, for 3 h decreased the EIC of lavage fluid by 42% in the placebo group (Fig. 3) in contrast to a 23% decrease in the vitamin-supplemented group; the difference between the two groups was highly significant (58 it 9 vs. 77 + 6%, P < 0.03). The EIC of the vitamin-supplemented group (2.75 + 0.17 pg a,-PIlpg PPE) was similar to that of the control air-exposed group reported previously (2.78 ~fr0.41)(27). Both of these values were significantly lower (i.e., higher EIC) than the placebo-treated NO,-exposed group (3.67 + 0.32, P < 0.03). Lipid peroxidation products in BAL fluid. To examine
the effects of lipid peroxidation products on the EIC of BAL fluid, portions of lung lavage fluid were extracted for lipids, and the lipid peroxidation products were measured in the lipid extract as malondialdehyde and conjugated dienes. The lipid peroxidation products were expressed per total lavage fluid phospholipids. The amount of malondialdehyde in the lipid fraction of the lavage fluid of the placebo group was 10.2 + 3.6 pmollpg phospholipid as opposed to 8.5 + 3.3 pmollpg in the vitaminsupplemented group. Although malondialdehyde concentration tended to be higher in the placebo group, the difference was not statistically significant. However, the levels of conjugated dienes in the NO,-exposed group without vitamin supplementation (placebo) was significantly higher than those in the vitamin-supplemented group (P = 0.003; Table 2). There were approximately SO- and 40-fold larger amounts of conjugated dienes than malondialdehyde in the lavage fluid of placebo and vitamin-supplemented groups, respectively. There was no
Bronchoalveolar lavage. Table 1 shows the characteristics of the subjects and the cellular composition of the BAL. There were no significant differences in the subject characteristics. The NO, exposure and the bronchoalveolar lavage were well tolerated by all subjects. There was no significant difference in the volume of recovered fluid between the placebo and vitamin-supplemented groups. NO, did not increase the total cell numbers or alter the differential cell count in either group. The percentage and the absolute number of neutrophils did not differ from our air-exposed control group reported previously (27). Vitamins C and E had no discernible effect on the total cell numbers or the differential cell counts. There was no significant difference in the cell viability between the two groups, and both were greater than 88%. Serum vitamin C and vitamin E levels. Serum levels of both vitamins C and E were significantly higher in the group with supplementation than in the placebo group. The placebo group had mean levels of 1.1 * 0.2 and 1.01 rfr 0.06 mg/dl for vitamin C and vitamin E, respectively, which were not significantly different from those reported for normal nonsmoking subjects (26). With supplementation, the ascorbic acid increased more than twofold, and vitamin E increased more than threefold (Fig. 1). Albumin and antigenic a,-PI. The concentration of albumin in the lavage fluid was determined as a measure of increased permeability of pulmonary vascular endothelium and alveolar epithelial lining. NO, exposure failed SUPP. SUPP. to increase the albumin concentration in the lavage fluid. FIG. 1. Serum levels of ascorbic acid and a-tocopherol Likewise, vitamin C and vitamin E supplementation had and vitamin C- and vitamin E-supplemented groups. Vitaminin placebo suppleno effect on the lavage albumin. Similarly, neither NO, mentation for 4 wk increased serum ascorbic acid more than twice and alone nor NO, with vitamin supplementation had any cu-tocopherol more than three times the levels in placebo group.
T
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
OXIDANT-INDUCED A Albumin
LIPID
PEROXIDATION
IN HUMAN
B ojPl
LUNG
1459
C o ,PI / Albumin
50
50
40
40
30
30
20
20
10
10
E Y
0 L
0
FIG. 2. Effect of NO, exposure and vitamin C and E supplementation on alveolar lining fluid albumin and antigenic ol,-protease inhibitor (oc,-PI). Neither NO, exposure nor vitamin supplementation altered concentration of albumin (A) and antigenic a,-PI (B) in lavage fluid. oc,-PI-to-albumin ratio (C) remained unchanged and was comparable between groups.
significant difference in the recovered total phospholipid content of the lavage fluid between the groups. Relationship between lipid peroxidation products and EIC of BAL fluid. To determine the relationship between
lipid peroxidation and EIC, we used a least square regression analysis to examine the interaction of conjugated dienes and malondialdehyde with EIC of BAL fluid. There was a strong correlation between the levels of conjugated dienes and EIC of BAL fluid, r = 0.62, P = 0.008 (Fig. 4). Although there was an overlap in three data points, overall the analysis demonstrated that the higher the conjugated dienes the lower the EIC. In Fig. 4,
0 Placebo
Vitamin
C+E
3. EIC of alveolar lining fluid after NO, exposure in placebo and vitamin C- and E-supplemented groups. Concentrated lavage fluids with known levels of antigenic @,-PI were titrated against chromatographically purified and active site-titrated porcine pancreatic elastase (PPE). Residual elastase activity was measured by fluorogenic substrate as described under METHODS. Amount ofcu,-PI in pg required to fully inhibit 1.0 pg PPE was taken as the elastase inhibitory capacity (EIC) of alveolar lining fluid. NOz exposure decreased EIC of control group by 42%, whereas vitamin-supplemented group retained most of its EIC. EIC of vitamin-supplemented group was comparable to our previously reported data on air-exposed controls (27). Values are means rf: SE and are expressed as percentage of fully active CQ-PI with a FIG.
value
Of 2.12 (See METHODS).
the EIC of BAL fluid is expressed as a percentage of purified and fully active (w,-PI. Figure 4 also shows that vitamin-supplemented subjects as a group not only had lower levels of conjugated dienes but also had higher EIC. In contrast to conjugated dienes, malondialdehyde levels in the lavage fluid did not correlate with the EIC (data not shown). The concentrations of malondialdehyde were 1.3 and 2.3% of respective conjugated diene levels in placebo and vitamin-supplemented groups. DISCUSSION
The data presented in this study demonstrate that exposure of young healthy subjects to low levels of NO, for 3 h results in a significant diminution of EIC of the alveolar lining fluid. This decrease in antiprotease defense of the lower respiratory tract is associated with the appearance of peroxidation products in the surfactant lipids. The relationship between enhanced lipid peroxidation and low EIC of the alveolar lining fluid is further supported by the protective action of vitamins C and E in these subjects in whom vitamins C and E decreased lipid peroxidation and enhanced EIC of the alveolar lining fluid. NO,, a common air pollutant and a major component in cigarette smoke (31), is an oxidant that can cause emphysema in experimental animals after long-term exposure (20). We previously demonstrated that even shortterm exposure (3 h) of healthy subjects to NO, causes a significant reduction in the antiprotease defense of the alveolar lining fluid (27). However, exposure of (u,-PI to NO, in vitro did not result in a decrease of its EIC, even if the level of NO, was 10 times higher than the in vivo 2. Lung lavage phospholipids, malondialdehyde, and conjugated dienes after exposure to nitrogen dioxide
TABLE
Total Phospholipids, &nl
Placebo Vitamin supplemented
BAL
21.3e2.9 23.012.3
Malondialdehyde, pmol/ag
PL
10.2k3.6 F3.5r3.3
Conjugated Dienes, pmol/rg
PL
304+103 369+58*
Values are means + SE. BAL, bronchoalveolar lavage; PL, phospholipids. * Significantly different from placebo group, P = 0.003.
Downloaded from www.physiology.org/journal/jappl by ${individualUser.givenNames} ${individualUser.surname} (129.059.095.115) on January 13, 2019.
1460
OXIDANT-INDUCED
LIPID
PEROXIDATION
r = 0.62 P= 0.008 0
a l2
a.
-
.4
L
-
.
-
a
.6 .8 1 Conjugated Dienes (pmoleslmg Phospholipids)
-
I
1.2
FIG. 4. Correlation between conjugated dienes and EIC of bronchoalveolar lavage fluid in placebo (0) and vitamin-supplemented groups (0) after NO, exposure. Placebo group had high levels of conjugated dienes and decreased EIC. Conversely, the vitamin-supplemented group had low levels of conjugated dienes and high levels of EIC.
exposure (28). On the basis of our in vitro experiments, we postulated that NO, exerts i ts inhibitory action against cu,-PI th rough .peroxidation of surfactant and cell membrane lipids (28). Although the major component of surfactant phospholipids is disaturated phosphatidylcholine, a significant portion of phosphatidylcholine contains mono- and polyunsaturated fatty acyl moieties (34,37). The process of lipid peroxidation starts with the initial step of hydrogen abstraction from an unsaturated fatty acid, which forms conjugated dienes or lipid radicals. Lipid radicals subsequently form lipid peroxides in the presence of molecular 0, and eventually lipid endoperoxides. The further reactions lead to the formation of malonaldehydes or malondialdehydes and lipid hydroperoxides. In the current study, conjugated dienes were the predominant species, and thiobarbituric acid reactants (primarily malondialdehydes) constituted
