Elevated Levels of NAP-1/lnterleukin-8 Are Present in the Airspaces of Patients with the Adult Respiratory Distress Syndrome and Are Associated with Increased Mortality1-3
E. J. MILLER, A. B. COHEN, S. NAGAO, D. GRIFFITH, R. J. MAUNDER, T. R. MARTIN, J. P. WEINER-KRONISH, M. STICHERLING, E. CHRISTOPHERS, and M. A. MATTHAY
Introduction
T he adult respiratory distress syndrome (ARDS) is an acute deterioration of lung. function that occurs in association with severe clinical disorders including sepsis and aspiration, and following major trauma (1, 2). ARDS results from a severe injury to the pulmonary alveolar-capillary membrane which leads to flooding of the alveolar spaces. The alveolar flooding disrupts normal gas exchange, and severe hypoxemia results. Although ARDS can occur in patients who are neutropenic (3), several lines of evidence suggest that neutrophils may influence the initiation and/or severity of the acute lung injury in patients with normal supplies of neutrophils. A large number of neutrophils accumulate within the lung in the early phase of acute lung injury (4, 5). Neutrophils are sequestered in the lung capillaries (4, 6), causing a transient decrease in the number of circulating neutrophils in some patients (7). In most individuals with ARDS, there is a marked increase in neutrophils in the alveolar spaces, such that neutrophils often constitute 80070 or more of the total cells obtained by bronchoalveolar lavage (BAL) (8), compared with approximately 3070 in normal subjects. Neutrophil migration alone probably does not cause injury to the alveolar epithelium (9, 10). However, lung injury may result when neutrophils migrating into the alveoli are stimulated to secrete elastase and myeloperoxidase. There is considerable evidence linking neutrophils to the severity of the ARDS. Both the neutrophil elastase concentration (11) and the neutrophil concentration (5, 12) were reported to correlate with the severity of the gasexchange abnormality in the lungs in these patients. The activity of myeloperoxidase is high in the lung washes, and this activity is cytotoxic to normal lung parenchymal cells (5). In addition,
SUMMARY The adult respiratory distress syndrome (AROS) Is characterized by Increased neutrophlls within the airspaces of the lungs. In order to determine If neutrophil activating protein (NAP)1/Interleukln-& (NAP-1/IL-&) could be an Important cause of neutrophil Influx and activation In AROS, we examined fluid, which was either directly aspirated or lavaged with saline from the lungs of patients with AROS. NAP-1/IL-& was present In significantly higher concentrations In the fluids of patients with AROS compared with control subjects. There was a significant correlation between the percentage of neutrophllsln the lavage fluids and the NAP-1/IL-& concentration (r 2 = 0.74). Further· more, the NAP·1/IL-& concentration of the pulmonary edema fluid was equivalent to the optimal concentration required to Induce neutrophil chemotaxis In vitro. Although not all of the chemotactic activity of the edema fluid was removed by an antl·NAP·1/IL·& affinity column, the data established that NAP·1/IL·&Is an Important neutrophil chemotaxln In the airspaces of patients with AROS. In addition, those patients with very high concentrations of NAP·1/IL·&ln their bronchoalveolar lavage fluids had a higher mortality rate than those patients with lower concentrations of NAP·1/IL·&. The correlation between NAp·1/IL·& concentration and mortality Is not paralleled by total protein eonAM REV RESPIR DIS 1992; 146:427-432 centratlon and mortality.
there are many animal models of ARDS that are totally or partially dependent on the presence of neutrophils (13-15). In order to identify a potential mechanism for the recruitment of neutrophils into the airspaces in patients with ARDS, we examined the role of the recently described neutrophil chemotaxin NAP-II interleukin-8 (NAP-I/IL-8) (16)in the attraction of neutrophils into the lungs during ARDS. NAP-I/IL-8 is a 72-aminoacid peptide that induces neutrophil chemotaxis (16) and releases enzymes from primed neutrophils (17). Because many of the cells of the alveolar environment have the capacity to produce NAP-I/IL-8 (18-20), we examined fluids from the airspaces of the lungs both for NAP-I/IL-8 content and NAP-I/IL-8related chemotactic activity. Weanalyzed the lung lining fluid directly using samples of pulmonary edema fluid (PEF) and indirectly using BAL fluids. Methods Protocols wereapproved by the reviewboards for human experimentation at each of the participating institutions. 1\\'0 different kinds of fluid were studied. BAL fluids from patients with AROS were
obtained in Seattle, Washington. BAL fluids from normal volunteers and patients with chronic obstructive pulmonary disease (COPO) were obtained from Tyler, Texas. PEF were obtained from San Francisco, California. Preliminary studies showed that shipping and storage did not alter the variables being examined. Pulmonary edema was classified as AROS if the patient had acute respiratory failure severeenough to require mechanical ventilation, (Received in originalform August 26, 1991 and in revisedform February 7, 1992) 1 From the Departments of Biochemistry and Medicine, University of Texas Health Center, Tyler, Texas, the Department of Medicine, Harborview Medical Center, Seattle, Washington, the Veterans.Affairs Medical Center, Seattle, Washington, the Cardiovascular Research Institute, University of California, San Francisco, California, and the Department of Dermatology, University of Kiel, Germany. 2 Supported by Grant Nos. ROI-HL43650, HL30542, HL40626, SCOR-HLI9155,andAI29103 from the National Heart, Lung, and Blood Institute, a grant from the American Heart Association, Texas Affiliate Inc., the Elizabeth Guggenheim Fund, and the Department of VeteransAffairs. 3 Correspondence and requests for reprints should be addressed to E. J. Miller, Department of Biochemistry, P.O. Box 2003, University of Texas Health Center, Tyler, TX 75710.
427
428
MILLER, COHEN, NAGAO, GRIFFITH, MAUNDER, MARrIN, WEINER-KRONISH, STICHERLlNG, CHRISTOPHERS, AND MATTHAY
a ratio of arterial to inspired Pao, ~ 0.2, bilateral diffuse infiltrates on chest roentgenogram, a pulmonary artery wedge pressure < 15mm Hg, and the absence of clinical congestive cardiac failure, or bacterial pneumonia. Pulmonary edema was considered to be cardiogenic if there was either a pulmonary artery wedge pressure> 18 mm Hg or a central venous pressure > 12 mm Hg, BAL fluids from patients with ARDS, normal volunteers, and patients with COPD were obtained. The samples, which were collected as previously described (8), werealiquoted and stored at -70 0 C. The samples were thawed and analyzed as a single batch. We analyzed BAL fluids from 17normal subjects, 19individuals with ARDS, and 20 subjects with COPD. Samples of greater than 2 ml of PEF were suetioned as previously described (21). A sample of peripheral blood was also collected at this time to determine the edema/plasma protein ratio. Heparinized plasma and PEF samples were centrifuged at 6000 x g for 10min. NAP-l/IL-8 was quantitated using a twosite enzyme-linked immunosorbent assay (Quantikine; R&D Systems, Minneapolis, MN). The assay was standardized using recombinant NAP-I/IL-8, which was quantified using amino-acid analysis. Murine monoclonal antibody 52E8, which is specific for NAP-l/IL-8 (22), was used to produce an affinity column for NAP-l/IL-8. The antibody was linked to Affi-geliO (BioRad, Richmond, CA), according to the manufacturer's instructions, to produce a I-ml affinity column. A control column was produced and utilized in an identical manner,
cells. The distance that the leading two cells had moved through the filter was measured for six fields on each filter. The measurements were made with two filters for each set of conditions.
using an isotypic antibody, of undetermined specificity (MOPC 21; Sigma Chemical, St. Louis, MO). Nonspecific binding sites on the columns were blocked by treatment with 10 ml of 0.01M phosphate buffered saline (PBS), pH 7.4, containing 1070 (wt/vol) bovine serum albumin (BSA). The column, which was kept at 4 0 C throughout, was then washed with 5 ml of Hanks' balanced salt solution (HBSS). PEF were centrifuged to remove particulates. A l00-t.ll aliquot of edema fluid diluted 1:2 with HBSS was then added to the column. The column was eluted with 1.15 ml HBSS, and the eluate was collected and stored at -70 0 C until analyzed for chemotactic activity. The column was regenerated by washing with 0.2 M acetic acid, HBSS, and HBSS containing 1070 (wt/vol) BSA, successively. In preparation for chemotactic studies, human neutrophils were isolated from donors using a modification of the method of Boyum (23). The neutrophils, which were 80 to 90070 pure, were suspended in HBSS. Chemotaxis was performed using the leading front method as described by Zigmond and Hirsch (24). An aliquot of the test material (140 J.1I) was placed in the lowerwellof a Boydenchamber (25). A 5 J.1m pore size, 100 J.1m thick cellulose nitrate filter (Sartorius Filter, Inc., San Francisco, CAl was placed on the surface, and the chamber was then assembled. A 200-J.11 aliquot of the neutrophil preparation (1 X 106 cells/ml) in RPMI-1640 containing 1070 albumin was added to the top of the filter and incubated at 37 0 C for 30 min. The filter was then fixed and stained and mounted on a glass microscope slide. The leading front was determined by the position of the leading two
Statistical Analysis The data were analyzed for statistical significance using one of a series of tests. (1) Student's t test was used if the data were normally distributed and contained two data sets. (2) The nonparametric Mann-Whitney U test was used if the two data sets were not normally distributed. (3) If three independent random samples were tested to determine if they came from identical distributions, the Kruskal-Wallis test was used. All results are given as mean ± standard deviation unless otherwise specified. Results
Correlation between NAP-l/IL-8 Concentration and Disease State In the first series of studies, we measured NAP-l/IL-8 in BAL fluids of patients with AROS, normal volunteers and patients with COPO. The clinical characteristics of the patients groups are presented in tables 1-3. The BAL fluids of the AROS patients contained 58 ± 26070 neutrophils compared with the BAL fluids from the normal and COPO groups, which had 1.3 ± 1.5070 and 1 ± 1.1 070, respectively.The mean NAP-l/IL-8 concentrations in the patients with AROS was 521 ± 484 pg/ml, compared
TABLE 1 BAL FLUIDS: CLINICAL CHARACTERISTICS OF ARDS PATIENT GROUP Bronchoalveolar Lavage Fluid Risk
Age (yr)
ARDS Day·
Primary
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
68 20 43 29 38 38 62 19 19 32 44 60 48 62 52 30 20 29 56
1 11 4 4 2 1 3 3 3 7 4 2 6 8 7 7 2 7 6
Other Trauma Trauma Sepsis Trauma Trauma Sepsis Trauma Trauma Drug overdose Sepsis Trauma Sepsis Sepsis Sepsis Sepsis Trauma Trauma Transfusion
Mean SO
40 16
5 3
Patient No.
Secondary
Transfusion Transfusion Transfusion Transfusion Trauma
Transfusion
Trauma Aspiration
Aspiration
• AROS day = time from onset of respiratory failure. NO = not done. l = lived, 0 = died.
t
*
Total WBC (x 103/ml )
PMN (x 103/ml )
PMN (%)
Total Proteint (mg/d/)
Outcome;
1,771.7 304.5 395.7 173.9 98.7 1,423.8 169.7 478.1 1,521.2 314.0 2,483.9 302.9 2,060.9 313.9 202.1 468.2 618.0 721.4 277.0
1,275.6 243.6 87.1 66.1 59.2 1,267.2 125.6 186.5 1,217.0 12.6 2,136.2 133.3 1,813.6 244.8 28.3 215.4 457.3 295.8 193.9
72 80 22 38 60 89 74 39 80 4 86 44 88 78 14 46 74 41 70
86.3 43.1 42.3 143.3 100.4 159.0 44.0 145.3 583.2 NO 73.1 15.8 418.4 71.1 1.1 35.8 106.1 26.1 22.3
L 0 L L L L L L L 0 L L 0 0 L 0 0 L L
742.1 726.8
529.4 659.4
58 26
117.6 145.4
NAP-1/IL-8, NEUTROPHILS, AND MORTALITY IN ARDS
429
TABLE 2
1500
BAL FLUIDS: CLINICAL CHARACTERISTICS OF NORMAL VOLUNTEER GROUP
1250
Bronchoalveolar Lavage Fluid Patient No.
Age (yr)
PMN (x 103/ml)
Total WBC (x 103/m/)
1000
PMN (%)
Total Protein (mg/d/)
S
E. J. MILLER, A. B. COHEN, S. NAGAO, D. GRIFFITH, R. J. MAUNDER, T. R. MARTIN, J. P. WEINER-KRONISH, M. STICHERLING, E. CHRISTOPHERS, and M. A. MATTHAY
Introduction
T he adult respiratory distress syndrome (ARDS) is an acute deterioration of lung. function that occurs in association with severe clinical disorders including sepsis and aspiration, and following major trauma (1, 2). ARDS results from a severe injury to the pulmonary alveolar-capillary membrane which leads to flooding of the alveolar spaces. The alveolar flooding disrupts normal gas exchange, and severe hypoxemia results. Although ARDS can occur in patients who are neutropenic (3), several lines of evidence suggest that neutrophils may influence the initiation and/or severity of the acute lung injury in patients with normal supplies of neutrophils. A large number of neutrophils accumulate within the lung in the early phase of acute lung injury (4, 5). Neutrophils are sequestered in the lung capillaries (4, 6), causing a transient decrease in the number of circulating neutrophils in some patients (7). In most individuals with ARDS, there is a marked increase in neutrophils in the alveolar spaces, such that neutrophils often constitute 80070 or more of the total cells obtained by bronchoalveolar lavage (BAL) (8), compared with approximately 3070 in normal subjects. Neutrophil migration alone probably does not cause injury to the alveolar epithelium (9, 10). However, lung injury may result when neutrophils migrating into the alveoli are stimulated to secrete elastase and myeloperoxidase. There is considerable evidence linking neutrophils to the severity of the ARDS. Both the neutrophil elastase concentration (11) and the neutrophil concentration (5, 12) were reported to correlate with the severity of the gasexchange abnormality in the lungs in these patients. The activity of myeloperoxidase is high in the lung washes, and this activity is cytotoxic to normal lung parenchymal cells (5). In addition,
SUMMARY The adult respiratory distress syndrome (AROS) Is characterized by Increased neutrophlls within the airspaces of the lungs. In order to determine If neutrophil activating protein (NAP)1/Interleukln-& (NAP-1/IL-&) could be an Important cause of neutrophil Influx and activation In AROS, we examined fluid, which was either directly aspirated or lavaged with saline from the lungs of patients with AROS. NAP-1/IL-& was present In significantly higher concentrations In the fluids of patients with AROS compared with control subjects. There was a significant correlation between the percentage of neutrophllsln the lavage fluids and the NAP-1/IL-& concentration (r 2 = 0.74). Further· more, the NAP·1/IL-& concentration of the pulmonary edema fluid was equivalent to the optimal concentration required to Induce neutrophil chemotaxis In vitro. Although not all of the chemotactic activity of the edema fluid was removed by an antl·NAP·1/IL·& affinity column, the data established that NAP·1/IL·&Is an Important neutrophil chemotaxln In the airspaces of patients with AROS. In addition, those patients with very high concentrations of NAP·1/IL·&ln their bronchoalveolar lavage fluids had a higher mortality rate than those patients with lower concentrations of NAP·1/IL·&. The correlation between NAp·1/IL·& concentration and mortality Is not paralleled by total protein eonAM REV RESPIR DIS 1992; 146:427-432 centratlon and mortality.
there are many animal models of ARDS that are totally or partially dependent on the presence of neutrophils (13-15). In order to identify a potential mechanism for the recruitment of neutrophils into the airspaces in patients with ARDS, we examined the role of the recently described neutrophil chemotaxin NAP-II interleukin-8 (NAP-I/IL-8) (16)in the attraction of neutrophils into the lungs during ARDS. NAP-I/IL-8 is a 72-aminoacid peptide that induces neutrophil chemotaxis (16) and releases enzymes from primed neutrophils (17). Because many of the cells of the alveolar environment have the capacity to produce NAP-I/IL-8 (18-20), we examined fluids from the airspaces of the lungs both for NAP-I/IL-8 content and NAP-I/IL-8related chemotactic activity. Weanalyzed the lung lining fluid directly using samples of pulmonary edema fluid (PEF) and indirectly using BAL fluids. Methods Protocols wereapproved by the reviewboards for human experimentation at each of the participating institutions. 1\\'0 different kinds of fluid were studied. BAL fluids from patients with AROS were
obtained in Seattle, Washington. BAL fluids from normal volunteers and patients with chronic obstructive pulmonary disease (COPO) were obtained from Tyler, Texas. PEF were obtained from San Francisco, California. Preliminary studies showed that shipping and storage did not alter the variables being examined. Pulmonary edema was classified as AROS if the patient had acute respiratory failure severeenough to require mechanical ventilation, (Received in originalform August 26, 1991 and in revisedform February 7, 1992) 1 From the Departments of Biochemistry and Medicine, University of Texas Health Center, Tyler, Texas, the Department of Medicine, Harborview Medical Center, Seattle, Washington, the Veterans.Affairs Medical Center, Seattle, Washington, the Cardiovascular Research Institute, University of California, San Francisco, California, and the Department of Dermatology, University of Kiel, Germany. 2 Supported by Grant Nos. ROI-HL43650, HL30542, HL40626, SCOR-HLI9155,andAI29103 from the National Heart, Lung, and Blood Institute, a grant from the American Heart Association, Texas Affiliate Inc., the Elizabeth Guggenheim Fund, and the Department of VeteransAffairs. 3 Correspondence and requests for reprints should be addressed to E. J. Miller, Department of Biochemistry, P.O. Box 2003, University of Texas Health Center, Tyler, TX 75710.
427
428
MILLER, COHEN, NAGAO, GRIFFITH, MAUNDER, MARrIN, WEINER-KRONISH, STICHERLlNG, CHRISTOPHERS, AND MATTHAY
a ratio of arterial to inspired Pao, ~ 0.2, bilateral diffuse infiltrates on chest roentgenogram, a pulmonary artery wedge pressure < 15mm Hg, and the absence of clinical congestive cardiac failure, or bacterial pneumonia. Pulmonary edema was considered to be cardiogenic if there was either a pulmonary artery wedge pressure> 18 mm Hg or a central venous pressure > 12 mm Hg, BAL fluids from patients with ARDS, normal volunteers, and patients with COPD were obtained. The samples, which were collected as previously described (8), werealiquoted and stored at -70 0 C. The samples were thawed and analyzed as a single batch. We analyzed BAL fluids from 17normal subjects, 19individuals with ARDS, and 20 subjects with COPD. Samples of greater than 2 ml of PEF were suetioned as previously described (21). A sample of peripheral blood was also collected at this time to determine the edema/plasma protein ratio. Heparinized plasma and PEF samples were centrifuged at 6000 x g for 10min. NAP-l/IL-8 was quantitated using a twosite enzyme-linked immunosorbent assay (Quantikine; R&D Systems, Minneapolis, MN). The assay was standardized using recombinant NAP-I/IL-8, which was quantified using amino-acid analysis. Murine monoclonal antibody 52E8, which is specific for NAP-l/IL-8 (22), was used to produce an affinity column for NAP-l/IL-8. The antibody was linked to Affi-geliO (BioRad, Richmond, CA), according to the manufacturer's instructions, to produce a I-ml affinity column. A control column was produced and utilized in an identical manner,
cells. The distance that the leading two cells had moved through the filter was measured for six fields on each filter. The measurements were made with two filters for each set of conditions.
using an isotypic antibody, of undetermined specificity (MOPC 21; Sigma Chemical, St. Louis, MO). Nonspecific binding sites on the columns were blocked by treatment with 10 ml of 0.01M phosphate buffered saline (PBS), pH 7.4, containing 1070 (wt/vol) bovine serum albumin (BSA). The column, which was kept at 4 0 C throughout, was then washed with 5 ml of Hanks' balanced salt solution (HBSS). PEF were centrifuged to remove particulates. A l00-t.ll aliquot of edema fluid diluted 1:2 with HBSS was then added to the column. The column was eluted with 1.15 ml HBSS, and the eluate was collected and stored at -70 0 C until analyzed for chemotactic activity. The column was regenerated by washing with 0.2 M acetic acid, HBSS, and HBSS containing 1070 (wt/vol) BSA, successively. In preparation for chemotactic studies, human neutrophils were isolated from donors using a modification of the method of Boyum (23). The neutrophils, which were 80 to 90070 pure, were suspended in HBSS. Chemotaxis was performed using the leading front method as described by Zigmond and Hirsch (24). An aliquot of the test material (140 J.1I) was placed in the lowerwellof a Boydenchamber (25). A 5 J.1m pore size, 100 J.1m thick cellulose nitrate filter (Sartorius Filter, Inc., San Francisco, CAl was placed on the surface, and the chamber was then assembled. A 200-J.11 aliquot of the neutrophil preparation (1 X 106 cells/ml) in RPMI-1640 containing 1070 albumin was added to the top of the filter and incubated at 37 0 C for 30 min. The filter was then fixed and stained and mounted on a glass microscope slide. The leading front was determined by the position of the leading two
Statistical Analysis The data were analyzed for statistical significance using one of a series of tests. (1) Student's t test was used if the data were normally distributed and contained two data sets. (2) The nonparametric Mann-Whitney U test was used if the two data sets were not normally distributed. (3) If three independent random samples were tested to determine if they came from identical distributions, the Kruskal-Wallis test was used. All results are given as mean ± standard deviation unless otherwise specified. Results
Correlation between NAP-l/IL-8 Concentration and Disease State In the first series of studies, we measured NAP-l/IL-8 in BAL fluids of patients with AROS, normal volunteers and patients with COPO. The clinical characteristics of the patients groups are presented in tables 1-3. The BAL fluids of the AROS patients contained 58 ± 26070 neutrophils compared with the BAL fluids from the normal and COPO groups, which had 1.3 ± 1.5070 and 1 ± 1.1 070, respectively.The mean NAP-l/IL-8 concentrations in the patients with AROS was 521 ± 484 pg/ml, compared
TABLE 1 BAL FLUIDS: CLINICAL CHARACTERISTICS OF ARDS PATIENT GROUP Bronchoalveolar Lavage Fluid Risk
Age (yr)
ARDS Day·
Primary
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
68 20 43 29 38 38 62 19 19 32 44 60 48 62 52 30 20 29 56
1 11 4 4 2 1 3 3 3 7 4 2 6 8 7 7 2 7 6
Other Trauma Trauma Sepsis Trauma Trauma Sepsis Trauma Trauma Drug overdose Sepsis Trauma Sepsis Sepsis Sepsis Sepsis Trauma Trauma Transfusion
Mean SO
40 16
5 3
Patient No.
Secondary
Transfusion Transfusion Transfusion Transfusion Trauma
Transfusion
Trauma Aspiration
Aspiration
• AROS day = time from onset of respiratory failure. NO = not done. l = lived, 0 = died.
t
*
Total WBC (x 103/ml )
PMN (x 103/ml )
PMN (%)
Total Proteint (mg/d/)
Outcome;
1,771.7 304.5 395.7 173.9 98.7 1,423.8 169.7 478.1 1,521.2 314.0 2,483.9 302.9 2,060.9 313.9 202.1 468.2 618.0 721.4 277.0
1,275.6 243.6 87.1 66.1 59.2 1,267.2 125.6 186.5 1,217.0 12.6 2,136.2 133.3 1,813.6 244.8 28.3 215.4 457.3 295.8 193.9
72 80 22 38 60 89 74 39 80 4 86 44 88 78 14 46 74 41 70
86.3 43.1 42.3 143.3 100.4 159.0 44.0 145.3 583.2 NO 73.1 15.8 418.4 71.1 1.1 35.8 106.1 26.1 22.3
L 0 L L L L L L L 0 L L 0 0 L 0 0 L L
742.1 726.8
529.4 659.4
58 26
117.6 145.4
NAP-1/IL-8, NEUTROPHILS, AND MORTALITY IN ARDS
429
TABLE 2
1500
BAL FLUIDS: CLINICAL CHARACTERISTICS OF NORMAL VOLUNTEER GROUP
1250
Bronchoalveolar Lavage Fluid Patient No.
Age (yr)
PMN (x 103/ml)
Total WBC (x 103/m/)
1000
PMN (%)
Total Protein (mg/d/)
S
